Colorado Child Welfare Evaluation by yaofenjin

VIEWS: 3 PAGES: 308

									    Colorado
Child Welfare
  Evaluation

  Second Interim
  Implementation
   Status Report




      June 30, 2001
Colorado Child Welfare
Evaluation
Second Interim Implementation Status
Report



June 30, 2001




                           William M. Mercer, Incorporated
                           3131 E. Camelback Road
                           Suite 300
                           Phoenix, AZ. 85016

                           602 522 6500
Contents
Contents .......................................................................................................................... i

Executive Summary........................................................................................................ 1
§ Process Findings...........................................................................................................................2
§ Outcome Findings ........................................................................................................................3
     Trends for the Prevention of Abuse ........................................................................................ 3
     Trends for Family Connection ................................................................................................ 3
     Trends for Staying in a Permanent Home............................................................................... 3
§ Cost Findings................................................................................................................................4
§ Title IV-E Demonstration Findings..............................................................................................5
     Practice Findings..................................................................................................................... 5
     Outcome Measure Findings .................................................................................................... 5
        Family Environment Findings ............................................................................................ 5
        Parental Capabilities Findings ............................................................................................ 5
        Family Interaction Findings................................................................................................ 5
        Family Safety Findings ....................................................................................................... 6
        Child Well-Being Findings ................................................................................................. 6
        NCFAS Change Scores....................................................................................................... 6
     Cost Findings .......................................................................................................................... 6
§ Recommendations ........................................................................................................................7
Introduction .................................................................................................................... 1
§ Memorandum of Understanding ..................................................................................................2
    Performance Expectations ...................................................................................................... 2
    Financial Incentives ................................................................................................................ 3
    Other Key Aspects .................................................................................................................. 3
§ Major Goals of the Managed Care Initiative................................................................................4
§ Evaluation of the Managed Care Initiative...................................................................................5
    Overall Evaluation Design ...................................................................................................... 5
    Evaluation Components .......................................................................................................... 6
       Process Evaluation .............................................................................................................. 6
       Practice Evaluation ............................................................................................................. 7
       Outcome Evaluation............................................................................................................ 8
       Cost Evaluation................................................................................................................. 10
Arapahoe County Findings ........................................................................................... 11
§ Process Findings.........................................................................................................................12
    Progress in the Introduction of Managed Care Processes..................................................... 12
    Process Flow ......................................................................................................................... 14
       Flow of People — Pre-Managed Care — Spring 1998 .................................................... 15
       Flow of People—Post-Managed Care—Spring 2001....................................................... 18
       Flow of Funds – Pre-Managed Care – Spring 1998 ......................................................... 25
       Flow of Funds—Post-Managed Care—Spring 2001........................................................ 28
    Fidelity of Process Implementation ...................................................................................... 34

William M. Mercer, Incorporated                                        i                  Colorado Child Welfare Evaluation
                                                                                 Second Interim Implementation Status Report
       Funding Flexibility............................................................................................................ 34
       Interagency Coordination.................................................................................................. 34
       Integration of Informal Networks and Non-Traditional Services..................................... 35
       Implementation of Early Intervention and Prevention Programs ..................................... 35
§ Outcome Findings ......................................................................................................................36
    Annual Outcome Trends ....................................................................................................... 36
       Indicator 1: Recurrence of Abuse and Neglect ................................................................. 36
       Indicator 2: Substantiated Abuse and Neglect .................................................................. 37
       Indicator 3: Referrals and Investigations .......................................................................... 37
       Indicator 4: Discharge to Reunification............................................................................ 38
       Indicator 5: Children in Out-of-Home Placement ............................................................ 39
       Indicator 6: Length of Out-of-Home Stay......................................................................... 39
       Indicator 7: Percentage in Positive Placements ................................................................ 40
       Indicator 8: Number of Adoptions.................................................................................... 40
       Indicator 9: Time to Adoption .......................................................................................... 41
    Outcome Comparisons.......................................................................................................... 42
    Summary of Findings............................................................................................................ 43
§ Cost Findings..............................................................................................................................44
Boulder County Findings .............................................................................................. 49
§ Process Findings.........................................................................................................................50
    Progress in the Introduction of Managed Care Processes..................................................... 50
    Process Flow ......................................................................................................................... 52
       Flow of People—Pre-Managed Care—Spring 1997 ........................................................ 54
       Flow of People—Post-Managed Care—Spring 2001....................................................... 56
       Flow of Funds—Pre-Managed Care—Spring 1997 ......................................................... 63
       Flow of Funds—Post-Managed Care—Spring 2001........................................................ 66
    Fidelity of Process Implementation ...................................................................................... 72
       Funding Flexibility............................................................................................................ 72
       Interagency Coordination.................................................................................................. 72
       Integration of Informal Networks and Non-Traditional Services..................................... 72
       Implementation of Early Intervention and Prevention Programs ..................................... 72
§ Outcome Findings ......................................................................................................................73
    Annual Outcome Trends ....................................................................................................... 74
       Indicator 1: Recurrence of Abuse and Neglect ................................................................. 74
       Indicator 2: Substantiated Abuse and Neglect .................................................................. 74
       Indicator 3: Referrals and Investigations .......................................................................... 75
       Indicator 4: Discharge to Reunification............................................................................ 76
       Indicator 5: Children in Out-of-Home Placement ............................................................ 77
       Indicator 6: Length of Out-of-Home Stay......................................................................... 77
       Indicator 7: Percentage in Positive Placements ................................................................ 78
       Indicator 8: Number of Adoptions.................................................................................... 78
       Indicator 9: Time to Adoption .......................................................................................... 79
    Outcome Comparisons.......................................................................................................... 80
    Summary of Findings............................................................................................................ 81
§ Cost Findings..............................................................................................................................82


William M. Mercer, Incorporated                                       ii                  Colorado Child Welfare Evaluation
                                                                                 Second Interim Implementation Status Report
El Paso County Findings ............................................................................................... 87
§ Process Findings.........................................................................................................................88
    Progress in the Introduction of Managed Care Processes..................................................... 88
    Process Flow ......................................................................................................................... 90
       Flow of People—Pre-Managed Care—Spring 1997 ........................................................ 92
       Flow of People—Post-Managed Care—Spring 2001....................................................... 95
       Flow of Funds—Pre-Managed Care—Spring 1997 ....................................................... 103
       Flow of Funds—Post-Managed Care—Spring 2001...................................................... 106
    Fidelity of Process Implementation .................................................................................... 113
       Funding Flexibility.......................................................................................................... 113
       Interagency Coordination................................................................................................ 113
       Integration of Informal Networks and Non-Traditional Services................................... 113
       Implementation of Early Intervention and Prevention Programs ................................... 113
§ Outcome Findings ....................................................................................................................114
    Annual Outcome Trends ..................................................................................................... 114
       Indicator 1: Recurrence of Abuse and Neglect ............................................................... 114
       Indicator 2: Substantiated Abuse and Neglect ................................................................ 115
       Indicator 3: Referrals and Investigations ........................................................................ 115
       Indicator 4: Discharge to Reunification.......................................................................... 116
       Indicator 5: Children in Out-of-Home Placement .......................................................... 117
       Indicator 6: Length of Out-of-Home Stay....................................................................... 117
       Indicator 7: Percentage in Positive Placements .............................................................. 118
       Indicator 8: Number of Adoptions.................................................................................. 118
       Indicator 9: Time to Adoption ........................................................................................ 119
    Outcome Comparisons........................................................................................................ 120
    Summary of Findings.......................................................................................................... 121
§ Cost Findings............................................................................................................................122
Jefferson County Findings.......................................................................................... 127
§ Process Findings.......................................................................................................................128
    Progress in the Introduction of Managed Care Processes................................................... 128
    Process Flow ....................................................................................................................... 130
       Flow of People—Pre-Managed Care—Spring 1997 ...................................................... 132
       Flow of People—Post-Managed Care—Spring 2001..................................................... 135
       Flow of Funds—Pre-Managed Care—Spring 1997 ....................................................... 142
       Flow of Funds—Post-Managed Care—Spring 2001...................................................... 145
    Fidelity of Process Implementation .................................................................................... 151
       Funding Flexibility.......................................................................................................... 151
       Interagency Coordination................................................................................................ 151
       Integration of Informal Networks and Non-Traditional Services................................... 151
       Implementation of Early Intervention and Prevention Programs ................................... 152
§ Outcome Findings ....................................................................................................................153
    Annual Outcome Trends ..................................................................................................... 153
       Indicator 1: Recurrence of Abuse and Neglect ............................................................... 153
       Indicator 2: Substantiated Abuse and Neglect ................................................................ 154
       Indicator 3: Referrals and Investigations ........................................................................ 154
       Indicator 4: Discharge to Reunification.......................................................................... 155

William M. Mercer, Incorporated                                      iii                 Colorado Child Welfare Evaluation
                                                                                Second Interim Implementation Status Report
       Indicator 5: Children in Out-of-Home Placement .......................................................... 156
       Indicator 6: Length of Out-of-Home Stay....................................................................... 156
       Indicator 7: Percentage in Positive Placements .............................................................. 157
       Indicator 8: Number of Adoptions.................................................................................. 157
       Indicator 9: Time to Adoption ........................................................................................ 158
    Outcome Comparisons........................................................................................................ 159
    Summary of Findings.......................................................................................................... 160
§ Cost Findings............................................................................................................................161
Mesa County Findings ................................................................................................ 166
§ Process Findings.......................................................................................................................167
    Progress in the Introduction of Managed Care Processes................................................... 167
§ Outcome Findings ....................................................................................................................169
    Annual Outcome Trends ..................................................................................................... 169
       Indicator 1: Recurrence of Abuse and Neglect ............................................................... 169
       Indicator 2: Substantiated Abuse and Neglect ................................................................ 170
       Indicator 3: Referrals and Investigations ........................................................................ 170
       Indicator 4: Discharge to Reunification.......................................................................... 171
       Indicator 5: Children in Out-of-Home Placement .......................................................... 172
       Indicator 6: Length of Out-of-Home Stay....................................................................... 173
       Indicator 7: Percentage in Positive Placements .............................................................. 173
       Indicator 8: Number of Adoptions.................................................................................. 174
       Indicator 9: Time to Adoption ........................................................................................ 174
    Outcome Comparisons........................................................................................................ 175
    Summary of Findings.......................................................................................................... 176
§ Cost Findings............................................................................................................................177
Pueblo County Findings .............................................................................................. 182
§ Process Findings.......................................................................................................................183
    Progress in the Introduction of Managed Care Processes................................................... 183
    Process Flow ....................................................................................................................... 185
       Flow of People—Pre-Managed Care—Spring 1998 ...................................................... 187
       Flow of People—Post-Managed Care—Spring 2001..................................................... 189
       Flow of Funds—Pre-Managed Care—Spring 1998 ....................................................... 196
       Flow of Funds—Post-Managed Care—Spring 2001...................................................... 199
    Fidelity of Process Implementation .................................................................................... 205
       Funding Flexibility.......................................................................................................... 205
       Interagency Coordination................................................................................................ 205
       Integration of Informal Networks and Non-Traditional Services................................... 205
       Implementation of Early Intervention and Prevention Programs ................................... 206
§ Outcome Findings ....................................................................................................................207
    Annual Outcome Trends ..................................................................................................... 207
       Indicator 1: Recurrence of Abuse and Neglect ............................................................... 207
       Indicator 2: Substantiated Abuse and Neglect ................................................................ 208
       Indicator 3: Referrals and Investigations ........................................................................ 208
       Indicator 4: Discharge to Reunification.......................................................................... 209
       Indicator 5: Children in Out-of-Home Placement .......................................................... 210

William M. Mercer, Incorporated                                      iv                  Colorado Child Welfare Evaluation
                                                                                Second Interim Implementation Status Report
       Indicator 6: Length of Out-of-Home Stay....................................................................... 211
       Indicator 7: Percentage in Positive Placements .............................................................. 211
       Indicator 8: Number of Adoptions.................................................................................. 212
       Indicator 9: Time to Adoption ........................................................................................ 212
    Outcome Comparisons........................................................................................................ 213
    Summary of Findings.......................................................................................................... 214
§ Cost Findings............................................................................................................................216
Multiple County Findings ............................................................................................ 221
§ Process Findings.......................................................................................................................221
    Introduction of Managed Care Principles ........................................................................... 221
       Financial Relationship with State ................................................................................... 221
       Management Information System................................................................................... 222
       Utilization Management.................................................................................................. 223
       Quality Improvement ...................................................................................................... 226
       Provider Network............................................................................................................ 227
       Provider Financial Incentives ......................................................................................... 228
    Fidelity of Process Implementation .................................................................................... 229
§ Outcome Findings ....................................................................................................................232
       Outcome Indicator Results for the Prevention of Abuse ................................................ 253
       Outcome Indicator Results for Family Connection ........................................................ 253
       Outcome Indicator Results for Staying in a Permanent Home ....................................... 254
§ Cost Findings............................................................................................................................255
Title IV-E Waiver Demonstration ................................................................................ 272
§ Practice Findings ......................................................................................................................274
    Family-Centered Practice Findings..................................................................................... 275
    Strengths-based Practice Findings ...................................................................................... 276
    Access and Quality of Services Findings............................................................................ 276
    Collaboration with Community Support Systems and Other Agencies Findings............... 277
    Conclusion .......................................................................................................................... 277
§ Outcome Measure Findings......................................................................................................282
    Family Environment Findings ............................................................................................ 284
    Parental Capabilities Findings ............................................................................................ 284
    Family Interaction Findings................................................................................................ 285
    Family Safety Findings ....................................................................................................... 285
    Note on the Averaging of NCFAS Scores .......................................................................... 287
    NCFAS Change Scores....................................................................................................... 287
§ Cost Findings............................................................................................................................289
Discussion and Recommendations ............................................................................ 291




William M. Mercer, Incorporated                                       v                  Colorado Child Welfare Evaluation
                                                                                Second Interim Implementation Status Report
Executive Summary
In response to sharp increases in child welfare costs, Colorado entered the managed care arena in
1997 with Senate Bill 97-218. This bill increased the focus on performance and outcomes while
providing counties with a much greater degree of flexibility in the use of funds than they had in
the past. It also established three managed care pilot counties. These counties were given the
ability to retain unspent general funds resulting from improvements in performance and use them
to add to their array of services. In addition, they were permitted to waive certain rules and
regulations, and negotiate a limited number of rates and payment methods with providers.

The number of managed care pilot counties was increased to six in 1998, with the passage of
Senate Bill 98-165. This bill also allowed other counties to enter into managed care agreements
with the Colorado Department of Human Services (CDHS) beginning in State Fiscal Year (SFY)
2001.

In 1998, the CDHS took another significant step in increasing the ability of counties to use
funds in a flexible manner by applying for a federal Title IV-E waiver to allow it to use
federal funds targeted to out-of-home placements for alternative services. The waiver, which
was awarded late in 1999, complements both bills, extending the managed care initiative to
another source of funds. Preparation for the implementation of the waiver began toward the
end of SFY 2000.

Boulder, Jefferson, and Mesa counties were designated pilot counties in SFY 1998 and
Arapahoe, Pueblo, and El Paso counties were designated in SFY 1999. Arapahoe County has
become the Title IV-E waiver demonstration site.

This report provides findings regarding changes in those counties, individually and as a group,
in the areas of process, outcomes, and cost. It also describes initial findings from the Title IV-
E waiver demonstration project. It concludes with a set of recommendations.




William M. Mercer, Incorporated                  1                Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Process Findings


The managed care counties were evaluated to determine their progress in implementing six key
managed care processes.

1. Utilization management. All counties except El Paso have implemented some level of prior
   authorization while four counties—Arapahoe, Boulder, Jefferson, and Mesa—perform
   concurrent review on out-of-home placements. All managed care counties have forums for
   multi-agency planning.
2. Quality improvement. Although none of the managed care counties have a full quality
   improvement system, three have implemented quality improvement activities. El Paso has a
   formal quality improvement plan and program tied to agency-wide and program-level goals,
   and has a major focus on using data to improve services over time. Boulder and Mesa
   counties have specific quality-oriented initiatives of a more limited scope.
3. Adequate provider network. Although none of the counties have determined their required
   network capacity, all have moved to address readily perceived gaps in their networks. At the
   level of network structure, Boulder County administers its own coordinated network.
   Arapahoe and El Paso counties have focused upon broad-level partnerships with provider
   networks managed by subcontractors. All managed care counties, except Arapahoe and
   Pueblo, have allocated dedicated resources to track the capacity of certain providers.
4. Performance indicators. Arapahoe and El Paso counties have built performance indicators
   into some contracts.
5. Provider financial incentives. Boulder and El Paso counties have taken the most steps in this
   direction. Both counties have formal contractual agreements in place that set capped funding
   levels and share risk between the county and other agencies. Boulder, El Paso, Jefferson,
   Mesa, and Pueblo counties also have contractual risk-sharing agreements for CPA treatment
   services that place the full risk for treatment costs on the local MHASA and the full risk for
   room and board costs on the county child welfare department.
6. Appropriate management information system. All counties need to improve their information
   system as the one that was in place for most of the period of the evaluation, which is operated
   by the CDHS on behalf of the counties, is not adequate for administering a managed care
   program. The CDHS has just implemented a new system, TRAILS, which has the potential
   to yield more useful and timely information.

The extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations was assessed in four domains – funding flexibility, interagency
coordination, integration of informal networks and non-traditional services, and implementation
of early intervention and prevention programs. Boulder and El Paso counties each achieved
mature levels of implementation fidelity in all but one fidelity domain. The other three counties
included in the analysis experienced less consistent levels in their system process fidelity ratings.
Across the four domains, the five counties, as a group, received the highest fidelity ratings in the
interagency collaboration domain.




William M. Mercer, Incorporated                  2                Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Outcome Findings


Trends in the nine outcome indicators that were used in the evaluation often were consistent
between the managed care counties and a comparison group of non-managed care counties.
Where the trends were different, the differences were slight. On the basis of outcomes, it appears
that managed care has not yet had a discernable effect on outcomes, as measured in this
evaluation, for children and families in Colorado.

Benchmarking results generally revealed favorable performance for all Colorado large counties,
relative to national averages and comparison state averages. For example, there is a lower rate of
confirmed abuse and neglect in Colorado and children are less likely to be in out-of-home
placement during the year. Furthermore, Colorado children spend less time in out-of-home
placement. On this basis, it appears that managed care has not impacted, either positively or
negatively, Colorado’s apparent superior performance on outcomes in comparison to other states.

Trends for the Prevention of Abuse
There has been a slight rise in the rate of abuse and neglect per 1,000 children in the state for
both the group of managed care counties and the comparison group of counties, a trend that has
also been seen nationally. At the same time, the rates of referrals and investigations for abuse and
neglect have declined. There was little or no discernable rise in the rates of recurrence of
confirmed abuse and neglect of children among the county groups.

Trends for Family Connection
There has been a decrease in the percentages of children residing with parents at case closure in
both the managed care counties and the comparison counties. There was an increase in the
number of children placed out of the home at some point during the fiscal year for both groups of
counties. The mean length of stay in out-of-home care for children has generally increased or
stayed approximately the same.

Trends for Staying in a Permanent Home
Results in this domain were mixed. Managed care counties, as a group, did not show
significantly superior performance in rates of positive placements and adoptions when compared
to non-managed care counties. However, they were clearly better in time to adoption.




William M. Mercer, Incorporated                  3               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Cost Findings


When each managed care county’s allocation at the beginning of SFY 2000 was compared to its
actual expenditures at the end of the year all except Mesa spent more of their child welfare block
than their original allocations. After close out, Arapahoe, Jefferson, and Mesa had either
balanced their budgets or shown a savings while the other three counties showed a deficit.

The ratio of CORE service expenditures to selected out-of-home placement expenditures was
used as a rough index of community service effort. The amount of variation was striking, ranging
from Boulder at 55 percent (highest index of community service effort) to Pueblo at 15 percent
(lowest index of community service effort). El Paso’s showed a significant increase in SFY
2000.

In the area of out-of-home placement, the large increases in utilization that were occurring prior
to the advent of managed care have, in general, diminished. For example, the rate of increase in
residential treatment center (RTC) utilization has decreased in five counties and remained the
same in the other. Increases did occur in relative foster care and subsidized adoption but these
are positive findings. The cost of RTC showed a high rate of growth across almost all counties
after the introduction of managed care.

The utilization and cost of RTC per child in poverty in each county was used as an index of
efficiency in the use of this costly form of care. Managed care counties expended less funds on
RTC per child in poverty than non-managed care counties for several years. They also use less
days each year per child in poverty. It appears that managed care counties are performing better
on this index than non-managed care counties.

Both managed care and non-managed care counties have increased their expenditures for
subsidized adoption, with El Paso having the largest increase. Associated with this is El Paso’s
lead among managed care counties in adoptions per 1,000 children.

In summary, cost results are mixed. On the one hand, it appears that managed care counties have
not contained costs, as expenditures have exceeded allocations for the managed care counties as
a group. On the other hand, the degree of increases in cost for different types of residential care
appears to have been reduced from the period prior to the advent of managed care. Managed care
counties also appear to perform better than non-managed care counties on an index based on
expenditures and days of RTC compared to the population of children in poverty.




William M. Mercer, Incorporated                  4               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Title IV-E Demonstration Findings


There are two phases to the evaluation of the Title IV-E demonstration in Arapahoe County. The
first phase, which started in SFY 2001, involves the collection of information on a pilot group of
children presently in residential care and their families. The areas studied are practice, outcomes,
and cost.

Practice Findings

In the areas of family-centered practice, it appears that Arapahoe County caseworkers and
service providers have adopted a family-centered approach in their work with parents and
caregivers who have a child in residential services. The results are less clear about the use of a
strengths-based approach with the families in the study than they were about family-centered
practices. On the one hand, most families stated that they were satisfied or very satisfied that the
services they received built on their families’ strengths. On the other hand, most of the study
participants indicated that they were not told what they were doing well in caring for their
children. In regard to access to services, about half of the caregivers were pleased with the
degree of access they experienced.

Outcome Measure Findings
Family Environment Findings
The environment of the family is a relative strength for the children in the pilot group. According
to ratings on the North Carolina Family Assessment Scale (NCFAS), the most difficult challenge
facing families in this domain appears to be disturbances in the neighborhood that may,
occasionally, prevent children from spending time outside in the community.

Parental Capabilities Findings
The parents of the children in the pilot group experience some mild problems overall in their
capacity to properly care for their children, according to the NCFAS ratings. Parents’ physical
health is adequate on average but caseworkers rated parents as having problems in disciplinary
practices, child supervision, enrichment opportunities, and in the area of mental health.

Family Interaction Findings
Mild problems are also evident among the pilot families in the domain of family interactions.
Some problems were noted in the amount of mutual support offered by family members, the
degree of bonding between parents and children, the clarity and appropriateness of the parents’
expectations of the children, and the relationship between adult caregivers.

Family Safety Findings
Family safety is a problem for families in the pilot group. Problems have been identified in the
areas of physical abuse, sexual abuse, neglect, and domestic violence. Emotional abuse is a
particular area of concern for these families.


William M. Mercer, Incorporated                  5                Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Child Well-Being Findings
The well-being of children in the pilot group is the most significant problem identified on the
NCFAS. The children, as a group, were rated by caseworkers as experiencing emotional
difficulties, handling stress poorly, being uncooperative and oppositional, fighting often with
peers, being influenced negatively by peers, having poor relationships with their parents, and
having problems with law enforcement. The ratings indicated they have problems to a lesser
degree with school performance and sibling relationships. Caseworkers did not rate children in
the pilot group as having a large amount of motivation to change their behaviors and to remain
with their families.

NCFAS Change Scores
Two children in the pilot group were discharged from the RTC during the course of the pilot
study. Both showed improved ratings on the NCFAS at discharge.

Cost Findings

Arapahoe County has been using a network fee with a provider network since March 2000 to
expedite admissions of children to RTC. Results indicate that Arapahoe County reduced its
average monthly utilization more than any other managed care county, compared to a baseline
period, after the beginning of the network fee. During the same period, the average monthly cost
increased more than any other county except Jefferson.




William M. Mercer, Incorporated                 6               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Recommendations


Three major managed care recommendations are made.

§     Revisit and revise Colorado’s vision for managed care in child welfare and develop a
    strategic plan and annual work plan. The CDHS and counties should do this in
    partnership through the Managed Care Implementation Team (MCIT). The MCIT
    should also be used as the primary vehicle for guiding the systematic implementation
    of the work plan. The MCIT is one of the most valuable tools in Colorado for the
    implementation of managed care as long as it has a clear focus and a logical sequence
    of tasks.
§     Clarify the financial risk arrangement between the CDHS and the counties. The first
    step is to rationalize the child welfare allocation. The original financial mechanism of
    allowing counties to retain unspent funds for service development in return for
    foregoing participation in any year-end surplus distribution requires refining. The stop-
    gap measure this past year of allowing counties to be treated as non-managed care
    counties was important in helping them remain managed care counties despite
    financing difficulties. This all or nothing approach does not seem to make sense over
    the longer term. Evolving this financing arrangement into a risk corridor arrangement
    that could create a financial incentive for containing costs while retaining some
    measure of protection through a common risk pool (e.g., the surplus distribution)
    should be a major goal of next year’s MOU development process.
§     Identify technical assistance and support needs in key managerial process areas and
    arrange for their provision. While all six process areas noted above (management
    information systems, utilization management, local financial incentives, performance
    indicators, quality improvement, and provider network capacity) are areas that could
    benefit from technical assistance, we recommend three target areas in the next year.
     Management information systems. Managed care requires a strong financial data
     management component and most counties do not currently possess this. Instead,
     most counties rely upon reporting from the state that was never envisioned as
     sufficient for a managed care environment with its premium upon real-time, tailored
     and flexible reporting. The CDHS has planned for TRAILS to provide some
     functionality in this area, but it is recommended that the CDHS set a strategic goal of
     helping each county to develop its own, independent financial tracking and reporting
     capacity and allocate resources to support counties in this area.
     Utilization management. All managed care counties have tried various strategies to
     impact the use of expensive out-of-home placements. The CDHS and the counties are
     now in the position of being able to look across the various approaches that have been
     tried, pick those that have been most successful, refine them, and begin to standardize
     expectations and strategies across counties. While differences will remain, areas of
     consensus (e.g., review elements to include in the prior authorization process for out-
     of-home care) are emerging and should be disseminated and supported. It is
     recommended that the CDHS set a strategic goal of developing cross-county


William M. Mercer, Incorporated                   7                Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
   utilization management standards and allocate resources to support counties in
   achieving them.
   Contract management functions that combine financial incentives and performance
   indicators. Counties such as El Paso and Boulder have developed contract
   management approaches and capacity that begin to move this area. Pilot counties
   such as Jefferson, which belong to the Northern Consortium, have begun to address
   these matters in their relationships with CPAs. It is recommended that the CDHS set a
   strategic goal to support contract approaches predicated on clear financial incentives
   and performance indicators and allocate resources to support counties in developing
   and managing them over time.

This evaluation report should be used as a tool in the coming year at both the state and
county level. The strengths as well as the areas needing improvement in each county are
fairly evident in this report. Where outcomes in a particular county are considerably
better than those in other counties, the processes related to those outcomes should be
examined for possible adoption in other counties, especially in counties with the lowest
level of performance in an area. Harnessing the evaluation as a major vehicle for
supporting the strategic goals of Colorado’s initiative should be a primary focus during
the next year.




William M. Mercer, Incorporated                 8               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Introduction

In the latter half of the 1990s, state and county child welfare services throughout the country
began experimenting with managed care principles to increase quality and reduce cost. The
programs that have been introduced have ranged from broad initiatives, such as Florida’s
program to privatize most child welfare services, to narrow ones, such as Illinois’ program to
incentivize adoptions. The extent of the use of managed care principles has also varied
tremendously, from relatively simple programs, such as South Carolina’s adoption program,
which offers bonus payments for adoptions, to Massachusetts’ Commonworks, which features a
case rate, private administrative services organization, and bonus payments. In 1998, there were
47 initiatives planned or implemented in 29 different states.

In response to sharp increases in child welfare costs, Colorado entered the managed care arena in
1997 with Senate Bill 97-218. This bill increased the focus on performance and outcomes, while
providing counties with a much greater degree of flexibility in the use of funds than they had in
the past. It also established three managed care pilot counties. These counties were given the
ability to retain general funds that were saved through improvements in performance and used
them to add to their array of services. In addition, they were permitted to waive certain rules and
regulations and negotiate a limited number of rates and payment methods with providers.

The number of managed care pilot counties was increased to six in 1998, with the passage of
Senate Bill 98-165. This bill also allowed other counties to enter into managed care agreements
with the CDHS beginning in SFY 2001. It required that an evaluation be conducted to assess the
results achieved by the six managed care pilot counties.

In 1998, the CDHS took another significant step in increasing the ability of counties to use funds
in a flexible manner. The CDHS applied for a federal Title IV-E waiver to allow it to use federal
funds targeted to out-of-home placements for alternative services. The waiver, which was
awarded late in 1999, complements both bills, extending the managed care initiative to another
source of funds. Preparation for the implementation of the waiver began toward the end of SFY
2000 in Arapahoe County.




William M. Mercer, Incorporated                  1               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Memorandum of Understanding

The CDHS has implemented Senate Bills 97-218 and 98-165 through the use of memorandums
of understanding (MOUs) with each of the six managed care counties. As originally
implemented, MOUs describe a model of managed care in which counties are expected to
achieve certain outcomes and, if they do, can retain unspent funds from the year. If counties have
a deficit, they are expected to use local funds to balance their budgets. Counties are permitted to
apply for waivers from certain CDHS rules and regulations so they can operate with more
flexibility. MOUs share basic elements in common but are tailored to each pilot county through a
yearly process of negotiation.

In 2001, in response to the managed care counties’ concerns about having to give up eligibility
for statewide surplus distributions, the CDHS offered them a choice between keeping their own
unspent funds or participating in the distribution of unspent funds from other counties at the end
of the year. Offered part way through the year, this gave counties the ability to predict how their
finances were likely to come out at the end of the fiscal year and to choose the option in their
best interest.

As the most recent quantifiable data in this report are from SFY 2000, this change in the MOU is
not a factor in the results. This report will discuss results within the context of the conditions of
the MOU that existed prior to the CDHS offering this option.

Performance Expectations

The MOUs contain basic outcomes that managed care counties are expected to achieve. A
county that has baseline performance on a particular outcome that is lower than the state average
is expected to improve performance. A county that has baseline performance on an outcome that
is equal to or greater than the state average is expected to not fall below the state average.
Finally, a county that has baseline performance on an outcome that is higher than the state
average may negotiate with the CDHS over the level of achievement that is expected.

There are two domains of outcomes. The first is safety, defined as freedom from abuse and
neglect, regardless of where a child lives. The second is permanency, defined as living in a safe,
permanent home. For each domain, there are several outcome indicators.




William M. Mercer, Incorporated                   2               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Financial Incentives

A managed care county that meets all its performance expectations and underspends the general
fund portion of its allocation is required to use the funds for additional child welfare services. A
county may elect to use those funds, in an amount that cannot exceed five percent of the general
fund portion of the county’s allocation, to offset the county share. If a county does not meet its
performance expectations, unspent funds must be used to bring it into compliance based on a
plan developed with the CDHS.

Managed care counties, as well as all other Colorado counties, are responsible for any overages
from budgeted expenditures. In addition, managed care counties are able to retain any unspent
funds. Managed care counties, however, have to allow non-managed care counties to use unspent
funds from other areas of the state to cover their overages before being eligible for these.
Managed care counties are still eligible for excess Title IV-E distributions they help earn.

Other Key Aspects

There are several other aspects of the MOUs that bear mention.

§     Counties signing the MOUs may seek waivers of certain rules and regulations to enable them
    to be more flexible in the services that they offer.
§    Counties agree to serve on the Managed Care Implementation Team, a group comprised of
    CDHS staff and managed care county representatives.
§    Counties agree to participate in the managed care evaluation.
§    Counties agree to comply with the managed care plan approved by the CDHS.
§    The CDHS agrees to provide technical assistance that may be requested by the counties.
§    The CDHS agrees to monitor the MOUs.




William M. Mercer, Incorporated                   3               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Major Goals of the Managed Care Initiative

The major goals of the initiative include:

§   more efficient and responsive service systems for children, youth, and families;
§   increased flexibility and collaboration across multiple agencies and funding streams to more
    appropriately meet consumer needs and avoid cost shifting between programs;
§   encouragement and authorization for a truly integrated service system that incorporates
    blended funding and administration;
§   focus on quality and outcome-driven services with accountability for an entire array of
    services that families need rather than forcing families to be transferred from agency to
    agency;
§   development of data systems to support these goals and to allow administrators and policy
    makers to better manage and evaluate;
§   authority and incentives for creative solutions at the local level that are not bound by the
    constraints of current agency barriers and categorical funding streams, including authority for
    local policy makers to create new entities incorporating blended funding and administration;
    and
§   successful training efforts directed at county staff, judges, court staff, providers, parents, and
    families and other appropriate entities that are involved in managed care service systems.




William M. Mercer, Incorporated                   4                Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Evaluation of the Managed Care Initiative

As mentioned earlier, Senate Bill 98-165 requires an evaluation of the managed care system
developed by the CDHS and the six managed care counties. William M. Mercer, Incorporated
(Mercer) was selected in April 1999 to conduct the evaluation and has produced a report entitled
Interim Implementation Status Report that summarizes the first year of the evaluation.

The Interim Implementation Status Report provides an overview of the managed care programs
that have been implemented as a result of Senate Bills 97-218 and 98-165. It also provides
findings in the areas of process (how programs are implemented), practice (how services are
delivered), outcomes (what the results are), and costs (how do costs change). Because of the
early stage of the managed care initiative in Colorado, it does not draw relationships between
changes in process and practice and changes in outcomes.

The present report builds on the Interim Implementation Status Report. It includes another year
of data and some additional analyses, particularly in the area of cost. It examines counties as
managed care entities, noting key managed care elements that have and have not been
introduced. It provides a series of recommendations. It is intended to be a tool for use by the
CDHS and the counties as they continue to implement managed care initiatives in the Colorado
child welfare system.

Overall Evaluation Design
The overall evaluation design consists of two parts. The first part consists of a study of the six
managed care pilot counties (Arapahoe, Boulder, El Paso, Jefferson, Mesa, and Pueblo) and is
required by Senate Bill 98-165. It is a comprehensive, multi-year evaluation in which differences
in process, practice, outcomes, and costs for children served by the child welfare system in these
counties are being studied.

This part of the evaluation uses counties as their own controls. This means that, rather than
comparing the performance of the managed care counties to non-managed care counties,
counties are compared to themselves on various measures over time. This makes for a stronger
evaluation as differences among counties in such areas as local economies, population density,
and child populations preclude meaningful comparisons in such a complex area as child welfare.

It should be noted that while differences between the counties are not formally evaluated,
different practices are documented and contrasted so counties can learn what is happening
elsewhere and determine if such practices may be applicable for them. There are also a few
places in the evaluation where four large, somewhat similar counties that have not been part of
the managed care pilot program are used as comparison counties. They are used only to help
gauge the extent to which trends seen in the managed care counties represent broader trends that
are also seen in the comparison counties, or are specific to the managed care counties.




William M. Mercer, Incorporated                 5               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
The second part of the evaluation is a study of the effect that the Title IV-E waiver has on the
child welfare system. As mentioned earlier, Colorado has received a waiver from the federal
Department of Health and Human Services (DHHS) that gives it more flexibility in the use of
federal funds that support out-of-home placements. The DHHS requires a comprehensive, long-
term evaluation of the impact of the waiver.

Work has been underway since late SFY 2000 in Arapahoe County in preparation for the
implementation of the waiver. The evaluation of the demonstration has been designed and
Arapahoe County staff have been oriented and trained. Data have been collected on a baseline
sample and the results are included in this report.

The two parts of the overall evaluation design are complementary and tightly integrated. The
study of the managed care pilot counties provides a large-scale, population-based evaluation of
the impact of managed care approaches. The Title IV-E evaluation provides an in-depth look at
the effects of managed care and the federal waiver on specific youth and their families. Together,
they provide a comprehensive and detailed view of the impact of the critical changes that are
being made throughout Colorado’s child welfare system.

It is important to note that the evaluation is a long-term one. It will answer a host of questions
that have been posed now and that will be posed in the future. Some of these questions may be
answered in a short period of time while others will take the full duration of the evaluation. More
information, for example on outcomes, should emerge as processes have sufficient time to exert
an impact. In addition, as data are collected and enriched over time, the analysis of
interrelationships between processes, practices, outcomes, and costs will become more robust
and useful in determining future policy.

Evaluation Components
The evaluation consists of four major components: process, practice, outcome, and cost
evaluation. The following sections describe those components.

Process Evaluation
This component studies how the initiative is being implemented and will be used to relate
differences in implementation to differences in outcomes. It studies aspects of the
implementation such as the introduction of utilization management, the use of performance
indicators with providers, and the development of processes to improve quality. This component
of the evaluation is conducted annually.

We are using three primary measures to collect data on process. We are using a Managed Care
Implementation Review (MCIR) instrument to assess organizational aspects of the managed care
counties such as financing, relationships with providers, and management information systems.
Information from this instrument is supplemented with interviews to form the basis for
Blueprints, which are diagrams of the flow of funds and families and children through the county
systems. Finally, the MCIR contains questions that are used to assess system process fidelity,
which is the extent to which pilot counties are achieving the anticipated benefits of system
reform. Table 1 provides more detail about these measures.


William M. Mercer, Incorporated                  6               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                                      Table 1
                                Measures for the Evaluation of Process
Managed Care Implementation Reviews are used to determine the quality of a county’s ongoing planning,
staffing structure, funding approaches, and methods of implementing managed care. These reviews study not only
the implementation of managed care but also the processes that contribute to successful programs.
Blueprints are diagrams showing pre-reform and post-reform system organization and funding flows. They are also
used to illustrate system and funding differences, including case rates, fee-for-service, and provider payment
mechanisms. They provide a visual representation of child welfare system processes.
Fidelity of Organizational Implementation Scales are used to assess the extent to which each county implements
its planned reforms. Fidelity of Implementation Scales are helpful for providing feedback on whether the reform
design is being followed.

Practice Evaluation
This component studies the impact of the counties’ managed care initiatives on the services that
staff members provide to children and their families. For example, it explores how
family-friendly services are, whether or not new services are developed, and changes in barriers
to services, such as adoption-related services. The relationship between changes in practice and
outcomes for children and families will be studied using data from this component. Although this
component is scheduled to be conducted every other year of the evaluation, it was conducted in
Arapahoe County this year as part of the Title IV-E demonstration evaluation.

Two primary approaches are used in this component. The first is a Child Welfare Practice
Assessment (CWPA), a survey tool that gathers the perceptions of different aspects of the child
welfare system from a variety of respondents. Within the CWPA are Fidelity of Implementation
Scales that provide a quantitative measure of the degree to which family-oriented practices are
being delivered. Table 2 provides more detail on these two measures.

                                                      Table 2
                                Measures for the Evaluation of Practice
The Child Welfare Practice Assessment consists of three versions. The parent/caregiver version gathers
self-reports of families and other caregivers regarding the quality of their experience with child welfare services.
The staff version gathers information from county staff that assesses their orientation to child welfare practice and
perspectives on the success of managed care and other reforms. Finally, the community/stakeholders version is
administered to judges, education officials, other child-serving agencies, and other members of the community to
assess their views of the child welfare system in the county. All are administered to statistically valid random
samples.
In SFY 2001, the CWPA for families and caregivers was adapted into a fourth version to focus on the specific
residential placement goals of Arapahoe County’s Title IV-E waiver demonstration.
Fidelity of Practice Implementation Scales are embedded in the CWPA and are designed to assess, for example,
the degree to which a family-oriented model of services is being implemented.




William M. Mercer, Incorporated                            7                 Colorado Child Welfare Evaluation
                                                                    Second Interim Implementation Status Report
Outcome Evaluation
This involves the evaluation of the impact of the initiative on the lives of children and their
families. It explores areas such as changes in well-being, permanency of placements, and child
safety. The relationship of outcomes to process, practice, and cost will be studied in later years of
the evaluation to understand their relative contributions. Parts of this component are conducted
annually while other parts are conducted every other year.

The overarching question for the outcome evaluation component of the evaluation is, “Are the
outcomes of services improved through the use of the risk- and performance-based contracts?
That is, are children safer, do children and families achieve permanency at higher rates and more
quickly, and is child and family well-being enhanced?”

There are four major measures used in this component of the evaluation. An Outcome Priority
Survey is used every other year of the evaluation to identify the outcome indicators that are of
the highest priority to Colorado. Benchmarking is conducted annually to compare performance
on these indicators to other states. Criterion-Referenced Standards are conducted every other
year to determine what is an acceptable level of performance. Finally, Outcome Tracking
mechanisms are used each year to track progress on the different outcome indicators. Table 3
provides some detail on these measures.

                                                      Table 3
                              Measures for the Evaluation of Outcomes
The Outcome Priority Survey is used to identify a limited set of outcome indicators upon which the evaluation
should focus. The results yield priorities among all stakeholder groups combined, as well as contrasts between the
outcome priorities for specific stakeholder groups, and for specific counties. Outcome priorities are used to analyze
the relative importance of outcome findings to different stakeholders.
Benchmark analyses involve a review of outcome indicators from other states and from the child welfare literature
to identify average and “best practice” benchmarks. Average and “best practice” benchmarks are then used as a
basis of comparison for the outcome findings in the evaluation.
The purpose of Criterion-Referenced Standards is to identify specific expectations for improvement for each of
the pilot counties on each outcome indicator. The criteria specified indicate outcome improvement thresholds that
represent “unacceptable,” “acceptable,” and “noteworthy” outcome results. Outcome improvement thresholds are
established every other year based on expectations regarding the pace of improvement expected as a result of reform
efforts.
Outcome Tracking mechanisms are established with the CDHS and the counties. They are used to track the
outcome indicators identified by the Outcome Priorities Survey.

The primary design for tracking outcome indicators over the course of the evaluation is the use
of managed care counties as their own controls. That is, key outcome indicators are being
tracked for all managed care counties over the course of the five-year evaluation.

Managed care counties are also being compared to four similar counties that have not
implemented managed care. Examining the trends in these counties alongside the trends for the
managed care counties helps us to be more careful about attributing changes to managed care
that may not actually be due to managed care.



William M. Mercer, Incorporated                           8                 Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Table 4 shows the six managed care counties and the four comparison counties. The managed
care counties are divided into two groups based on when they began implementing managed care
and taking on financial risk for their child welfare programs.

                                                Table 4
                             Counties used in the Evaluation of Outcomes
    Wave 1 Managed Care counties       Wave 2 Managed Care counties
    (began implementation 7/1/97)      (began implementation 7/1/98)     Non-Managed Care counties
           Boulder County                    Arapahoe County                    Adams County
          Jefferson County                    El Paso County                    Denver County
              Mesa County                     Pueblo County                     Larimer County
                                                                                 Weld County

The next component of our design involves the use of benchmarking. We have collected data at
the national level for several of the outcome indicators using sources such as the Child Welfare
League of America’s National Data System, which incorporates data from various external
sources and from its own annual survey of states. We have used these data as a point of
comparison for the findings of this evaluation.

We have also compared Colorado to a number of similar states. The group of comparison states
was derived using the following criteria:

§     Child population (as of 1999) between .5 and 1.5 million;
§     Western states; and
§    States primarily west of the Mississippi River that meet the child population criteria above
    and are county administered/state supervised systems.

These criteria led to the selection of the following states for benchmarking:

§     Arizona,
§     Minnesota,
§     New Mexico,
§     Oregon,
§     Utah,
§     Washington, and
§     Wisconsin.

The final component of our outcomes evaluation involves the use of criterion reference
standards. This component was implemented this year.




William M. Mercer, Incorporated                    9                Colorado Child Welfare Evaluation
                                                           Second Interim Implementation Status Report
Cost Evaluation
This component assesses the relationship between benefits, such as changes in out-of-
home placements, and cost. The overall question for the cost portion of the evaluation is,
“Is the cost efficiency of services improved through the use of risk- and performance-
based contracts?”

The cost analysis does not use any specific instruments for data collection. Instead, it
relies on information from the CDHS’s databases to quantify expenditures over time.




William M. Mercer, Incorporated                  10               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Arapahoe County Findings
Arapahoe County is located just to the south of Denver. It had an estimated child population of
126,148 in 2000. Arapahoe was accepted as a managed care county in SFY 1998 and as a
Title IV-E demonstration county in SFY 2000. Activities to implement the Title IV-E
demonstration began in earnest in SFY 2001.




William M. Mercer, Incorporated                11               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Process Findings
Process findings first examine how Arapahoe County is functioning as a managed care entity.
Blueprints are then presented to show changes in the flow of children and their families, and the
flow of funds through Arapahoe’s child welfare system. Finally, Arapahoe is scored on the
extent of its fidelity to managed care principles using Fidelity of Organizational Implementation
Scales.

Progress in the Introduction of Managed Care Processes
Arapahoe County is functioning as a managed care entity as it is assuming risk by entering into
an MOU with the CDHS. In its original MOU, Arapahoe County was permitted to keep unspent
funds but was required to cover overages that might occur. It is expected to perform at a certain
level on specific outcomes.

In 2001, the CDHS offered counties a choice between keeping any of their own unspent funds or
being eligible for participation in the distribution of unspent funds from other counties at the end
of the year. Arapahoe County chose to give up the ability to retain its own unspent funds in order
to participate in surplus distributions from the state.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                   12               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The main focus of Arapahoe County’s managed care initiative and Title IV-E waiver
demonstration has been on utilization management, the development of an adequate provider
network, and provider incentives. Of these three, the area that is the most developed is utilization
management.

Arapahoe has a clearly identifiable utilization management process that centers on a team called
the Resource Pathways Team. The team meets three times a week to staff all Residential
Treatment Centers (RTC) referrals; all RTC placements must be authorized by this team. Once in
an RTC placement, the team uses dedicated utilization review staff to conduct concurrent
utilization review focused upon the treatment plan within the RTC. The review process is
designed to ensure that the RTC maintains a focus on the county’s placement goals. The team
works to focus care on the goal of discharge back to a community or lower-level out-of-home
placement.

The team has participation from probation, school, mental health, SB 94 (Senate Bill 94, which
provides funds for alternative services for youth in the correctional system), a State Alcohol and
Drug Abuse Division (ADAD) representative, residential providers, and family treatment team
members from the Arapahoe Department of Human Services. Reviews are conducted at 45-day
intervals once a child is placed in a residential treatment center.

The use of a multi-agency team for utilization review represents an important step in Arapahoe
County’s development as a managed care entity. In refining the utilization review process,
Arapahoe might consider changing review intervals from fixed periods of time to periods that are
driven by the expected outcome for individual children. A fixed period might be used as a
maximum.

In regard to network capacity, Arapahoe County has not determined the capacity that it requires.
As a result, it is limited in developing a configuration of services that addresses the needs of
children and families now and in the future. Financial incentives for the development of services
that represent present best practices are also difficult to implement without knowing the overall
capacity that is required.

Although Arapahoe County has not determined its required network capacity, it has moved to
address readily perceived gaps in its network. For example, certain county-administered foster
homes now receive payments higher than the standard foster care rate to serve high-need
children who would normally be placed in an expensive RTC.

Arapahoe County is beginning to explore the use of performance indicators and financial
incentives through an agreement with Colorado Care Management (CCM), a provider network.
At present, there are no performance incentives but Arapahoe is developing a case rate that it
hopes to introduce into the agreement, along with performance expectations. This arrangement is
described in more detail in the section of this report dedicated to Colorado’s Title IV-E
demonstration project. Arapahoe County also passes on the risk for Child Protection Agency
(CPA) treatment costs to its local Mental Health Assessment and Service Agency (MHASA).




William M. Mercer, Incorporated                  13               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
In regard to the other key elements of managed care listed above, Arapahoe County is still in the
development stage. Arapahoe needs to develop a clearly identifiable quality improvement system
so that it may continuously improve its services and those of its providers. It also needs to
improve its information system as the one present during most of this evaluation, which is
operated by the CDHS on behalf of the counties, is not adequate for managing Arapahoe’s
managed care program. The CDHS has just implemented a new system, TRAILS, which has the
potential to yield more useful and timely information.

Process Flow
To illustrate the flow of processes involving people and funds in Arapahoe County, we have
constructed eight Blueprints. These show:
1. the flow of persons through the county in SFY 1998, prior to the initiation of managed care
   reforms;
2. the overall flow of persons through the county in SFY 2001, midway through the third year
   of managed care reforms;
3. additional detail on the flow of persons through the county department;
4. additional detail on the flow of persons through the county’s provider network;
5. the flow of funds through the county in SFY 1998, prior to the initiation of managed care
   reforms;
6. the overall flow of funds through the county in SFY 2001, midway through the third year of
   managed care reforms;
7. additional detail on the flow of funds from the state to the county, including financial
   reporting; and
8. additional detail on the flow of funds through the county’s provider network, including
   risk-sharing arrangements.

We have used six colors in the Blueprints:
1. black is used for the organizations involved in paying for, managing, and providing child
   welfare services;
2. red is used for the unmanaged flow of people through the system and includes county
   populations, populations in need, referral groups, movement through different teams and
   areas within the county child welfare department, and referrals to community providers;
3. green depicts child welfare funds that flow directly to the county child welfare department, as
   well as the state and federal funds that flow through the CDHS to the county child welfare
   department;
4. blue depicts the flow of non-child welfare funds that are critical to providing needed care to
   recipients of child welfare services;
5. purple depicts the managed flow of both persons and funds through the county child welfare
   managed care pilot initiative; and
6. yellow depicts the court system in each county.




William M. Mercer, Incorporated                14               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
In the Blueprints, the flow of people through the system is depicted in a left to right fashion,
beginning with the entire population potentially served by the child welfare system, narrowing to
the group of persons actually served, and following these people through the organizations and
services they would typically encounter as they pass through the system. The flow of funds,
however, goes from the top to the bottom of each chart. The drawing traces the flow of funds
from the payers that fund the system, to the administrative entities that allocate and account for
these funds, to the county child-serving agencies and providers that ultimately spend and manage
them.

A set of symbols is also used. They are illustrated in the following diagram.

Organizational Symbols:
               = Federal, state or county payors
               = State agencies
               = Local agencies or providers
               = Local entities that bears risk

Risk Symbols:
  S   = Shared risk of financial utilization
  F   = Full risk of financial utilization

Out-of-Home Placement Symbol:
  $            = Most expensive out-of-home
                   placements (RTC, CPA)
Financial Reporting Symbol:
   Financial
  Reporting
  by County    = Locus of financial reporting


Flow of People — Pre-Managed Care — Spring 1998
The Blueprint depicted on page 17 (entitled “Arapahoe County — Spring 1998 — Flow of
People”) shows the flow of people through Arapahoe County’s child welfare system in the
Spring of 1998, just prior to the initiation of their managed care pilot later that summer.
Beginning on the far left, the overall county population breaks down into the smaller group of
persons who, at any point in time, are in need of child welfare services. From that group, three
types of referrals are typically seen by the county: youth in conflict, youth with delinquency
problems, and children with reported instances of abuse or neglect.

All calls into the Arapahoe County Department of Social Services (DSS) in 1998 passed through
a “crisis screen,” which screened out calls that clearly did not meet criteria and referred them to
other services in the community. Additionally, callers with basic needs were often referred to the
county’s Temporary Assistance for Needy Families (TANF) program.




William M. Mercer, Incorporated                    15            Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Calls meeting criteria were referred on to the county’s Intake Unit, which opened cases and
conducted investigations. Even prior to the initiation of managed care reforms under SB 97-218,
Arapahoe County had a variety of levels of programs through which youth and families could
receive services in 1998. Cases with lower level needs could be referred to the county’s Family
Empowerment Unit. This specialized intake team worked with families for up to 60 days (in
general) in situations where community-based alternatives were seen as being able to resolve the
situation without need for additional services. All other cases were referred to Ongoing Child
and Family Units for adolescents and younger children, as well as to specialized treatment teams
providing home-based care, sexual abuse, teen parents, family interventions, and adoption.

Many children, youth, and families served by the county required additional services provided by
agencies in the community. Workers on ongoing and specialized teams were able to directly
access county-administered foster homes, as well as outpatient services through Arapahoe and
Aurora Community Mental Health Centers, and outpatient substance abuse services providers.
Children needing inpatient services had to go through the community mental health center,
which managed all indigent and Medicaid-funded mental health benefits. All referrals to RTCs,
CPAs, Residential Child Care Facilities (RCCFs), CHRP Host Homes (for children with
developmental disabilities), and CORE services (e.g., family preservation and day treatment)
passed through the Placement Review and Adolescent Resource Teams to receive multi-agency
staffings and efforts to divert care to the lowest level possible. This served a basic, gate-keeping
oriented managed care function prior to the managed care pilot.

A critical factor for any involuntary case served by the county is the court system. Individual
workers presented recommendations for services to the court and the court implemented them as
it saw fit. While the Blueprint highlights the role of the court at the point where it impacts
decisions regarding referrals to out-of-home placements versus other community services, the
court’s involvement with involuntary cases is much broader, beginning at or prior to intake and
extending throughout each family’s final disposition.




William M. Mercer, Incorporated                 16               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   17            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of People—Post-Managed Care—Spring 2001
In the third year of its implementation of managed care reforms, the flow of persons through the
Arapahoe County child welfare system has changed (see the Blueprint on page 19, entitled
“Arapahoe County—Spring 2001—People Flow”). To review these changes with greater
specificity, two detailed Blueprints were created for this year: a Department Internal Detail
Blueprint and a Provider Network Detail Blueprint.




William M. Mercer, Incorporated                18               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
William M. Mercer, Incorporated   19            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Department Internal Detail. Looking at the Department Internal Detail Blueprint (see the
Blueprint depicted on page 22, entitled “Arapahoe County—Spring 2001—Department Internal
Detail”), it can be seen that the county population still breaks down on the left into a smaller
group of persons in need who, in turn, sort into the three primary groups of youth and families
served by the county: youth in conflict, youth with delinquency concerns, and children with
reports of abuse or neglect. Within the county, the flow through Crisis Screen and Intake is
largely the same as are the referral alternatives within the department.

The primary change in person flow involves the newly created Resource Pathways and
Expedited Permanency Planning (EPP) Pathways Teams. The Resource Pathways Team was
developed to centrally manage RTC referrals and seeks to identify community-based alternatives
through multi-agency planning and a treatment authorization process. All RTC referrals must be
staffed by this team and any RTC referrals may now be diverted to either lower level out-of-
home placements (CPAs or county-administered foster homes), or to community-based
alternatives, such as community mental health center services. While the Resource Pathways
Team works with all RTC referrals, most of the youth it serves are adolescents (as they comprise
the bulk of RTC referrals). The focus of this effort is similar to that of the former Placement
Review and Adolescent Resource Teams but the goals have shifted from an emphasis on the
lowest level of care possible toward achieving the most appropriate, least restrictive placement in
the child’s best interests. Children’s Habilitation/Rehabilitation Program (CHRP) referrals are
also reviewed by this team.

Case workers can still make referrals to lower-level RCCF placements, which are lower-intensity
residential placements within the same facilities that provide higher-cost RTC placements. This
allows a significant number of RTC placements to circumvent the Resource Pathways process,
given that the vast majority of RCCF placements end up converting to an RTC level of care
during the placement. Arapahoe should examine whether or not the utilization of RCCF should
be more closely managed.

The EPP Pathways Team focuses on the treatment of all children, ages birth through six years
old, with an emphasis on identifying permanent homes, either through kinship options, adoption,
or a return home. The primary focus of this team is to achieve permanency within a 12-month
target through expedited treatment planning. This permanency initiative also serves a managed
care function, approving placements and identifying appropriate, least restrictive placements in
each child’s best interests. Kinship placements are also being increasingly pursued.

Through these two teams, RTC placements for all ages and all treatment decisions for children,
ages 1–6, are subject to prior authorization and multi-agency staffings. In addition, high cost
RTC placements are subjected to increased ongoing scrutiny through the utilization review
process. With supervisor authorization, caseworkers can also directly access RCCF, group home,
CPA, foster care, and CORE services. This relaxing of strict gatekeeping for out-of-home
placements and targeting of managed care efforts to promote optimal care choices is a key
feature of the Arapahoe County pilot.




William M. Mercer, Incorporated                 20               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
In early 2000, the following changes in the department were implemented:

§    The department changed its name to the Department of Human Services.
§    The Resource Pathways Team expanded its role to begin serving youth at risk of an RTC
    placement.
§     To increase accountability for managed care initiatives, the county created a Care
    Management Administrator position. All managed care functions report to this position,
    allowing for increased coordination and a single point of responsibility. This position has also
    begun to impact the day-to-day practices associated with the managed care pilot. For example,
    workers make increased use of tools, such as the Needs Based Care form, an assessment
    instrument used to identify the level of care needed by youth placed out of home. Increased
    use of this tool with CPA-referred cases has also been accomplished and other levels of care,
    such as county group home care, are targeted for future standardization of the approach.


Department-level changes in 2000–2001 include:

§    A substance abuse provider representative from Signal, the management service organization
    (MSO) for substance abuse services in Arapahoe County, now attends Resource Pathways as
    needed; and
§    Arapahoe County has hired a specialized foster home recruiter to increase the array of foster
    homes available for youth with special needs.




William M. Mercer, Incorporated                   21              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   22            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Provider Network Detail—People Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 24, entitled “Arapahoe County—Spring 2001—Provider
Network Detail—People Flow”), it can be seen that once in an RTC placement, the county
conducts a concurrent utilization review focused upon the treatment plan within the RTC. The
review process was designed to ensure that the RTC maintains a focus on the county’s placement
goals. Reviews are conducted within 10 days of placement and every 30 days thereafter. Backed
up by the authority to deny payment for ongoing care deemed unnecessary, utilization reviewers
work to focus care on the goal of discharge back to a community or lower level out-of-home
placement.

A primary development in 2000–2001 has been increased collaboration on multiple fronts with
CCM, a network of major providers. While the heart of this collaboration has involved financing
methods described below in the financing section, increased collaboration on cases has also
occurred. These include:

§     Arapahoe County has added a CCM representative to the Resource Pathways Team (the team
    overseeing utilization management for RTCs). Having a provider representative on this team
    has increased coordination, but also allows CCM to be privy to financial decision-making by
    the county and its agency partners.
§     Arapahoe County has worked with CCM to recruit interested and able CCM staff to serve as
    foster parents in the Pilot Homes project described below. This program provides enhanced
    foster care services by recruiting skilled mental health and human service professionals to
    serve as foster parents for children with special needs. Arapahoe County reports that CCM has
    been interested in this for multiple reasons, including enhanced care coordination, better access
    to RTC step-down resources, and increased reimbursement to help retain skilled RTC staff.
    The number of Pilot Homes has increased from 1 to 20 in the last year.
§     CCM has hired their own utilization manager to support utilization management activities for
    their network. However, Arapahoe County continues to provide prior authorization and
    concurrent utilization review for these referrals.

Another change is the increase in coordination and ongoing dialogue between the county and the
court system. While the court continues to exercise ultimate authority over all involuntary cases,
the court is more aware of alternatives to RTC care. In the last year, there has been increased
effort to improve the relationship between Arapahoe County and the court. This has required an
increased amount of attention and effort by the county over the past 12 months and involves the
DHS director, county commissioners, the county attorney’s office, and the chief judge.




William M. Mercer, Incorporated                   23               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
William M. Mercer, Incorporated   24            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of Funds – Pre-Managed Care – Spring 1998
The Blueprint depicted on page 27, (entitled “Arapahoe County—Spring 1998—Funding Flow”)
shows the flow of funds through the Arapahoe County child welfare system prior to the initiation
of the managed care pilot in mid-1998. In July 1997, funds from various federal sources (Title IV
and XX) and the State General Fund were combined by the CDHS and passed on to Arapahoe
County. The bulk of these funds passed through a large block that combined prior line-item
funding for county administration (largely used for staffing), out-of-home placements and other
smaller funding lines [e.g., Special Circumstances Child Care (SCCC) and subsidized adoption
funds]. The other primary funding source for child welfare services was the CORE services
funding stream, which was not integrated into the overall child welfare block in order to preserve
separate accountability for these funds for alternatives to out-of-home care, such as family
preservation and day treatment. Arapahoe County was not an EPP county in 1998 but received
other specialized state funding. These other sources of funds are not depicted.

While some CORE funds were provided entirely by the state, most required a 20 percent match
by the county. The overall child welfare block also required a match derived through an equation
combining the historical 100 percent, 80 percent, and other partially state-funded former separate
funding lines. In addition, SB 97-218 shifted the responsibility for excess costs in all Colorado
counties to the county. Even though Arapahoe County was not a managed care county at this
time, Arapahoe DSS and its county commissioners had the full downside risk of any cost
overages that exceeded the state’s ability to shift unspent funds from other areas of the state.
Because it was not yet a managed care county, Arapahoe County was not able to keep any
unspent funds it might have generated. As a result, Arapahoe County had the incentive to spend
all of its child welfare block funding but not to go over the amount, a partial managed care
incentive that is less clear in its effects than the full risk shift for managed care counties.

Arapahoe County, in turn, contracted with various community agencies for services. In addition
to funding streams flowing through DSS, Arapahoe County’s child welfare system also
depended on services paid for by other agency funding. These include:
§    Medicaid (Title XIX) funding paid through the Colorado Department of Health Care Policy
    and Financing (HCPF) to RTC and CHRP providers for the treatment component of the
    services provided in these facilities—The county DSS pays directly for room and board.
    Treatment costs are paid directly from the State Department of Human Services to HCPF but
    are then charged back to the county at year end against the Child Welfare Block allocation.
    The county has no control over the rates paid to providers since they are determined through a
    cost-based approach within available appropriations.
§     Medicaid (Title XIX) funding for mental health services transferred through the Department
    of HCPF to the Mental Health arm of the DHS Office of Health and Rehabilitation Services—
    In Arapahoe County, funding for all Medicaid inpatient and outpatient mental health services
    has been capitated and, as a result, flows through the entity awarded the MHASA contract for
    that area of the state, Behavioral Healthcare, Incorporated (BHI).




William M. Mercer, Incorporated                  25               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
§    Division of Youth Corrections (DYC) SB 94 funding for preventing detention placements for
    which some youth served by DSS are eligible—Services funded may be similar to CORE
    services.
§     ADAD funding through the DHS Office of Health and Rehabilitation Services—In Arapahoe
    County, these funds are managed by Signal, the managed services organization for substance
    abuse services. Signal pays for additional substance abuse services beyond those paid by DSS
    through its CORE funds.
§     Funds underwriting the treatment component of services provided by CPAs—Arapahoe
    County is able to earn a federal Medicaid (Title XIX) match on the treatment portion of its
    CPA costs through its implementation of a CPA Medicaid transfer that sends these funds from
    Arapahoe County DSS through HCPF to Behavioral Healthcare, Incorporated and, thereby, to
    the CPA providers. Arapahoe County has taken advantage of this in order to receive a roughly
    50 percent federal match on the treatment component of the care provided in CPAs.

As can be seen in the Blueprint, child welfare funds were the least managed of the primary
funding streams for expensive out-of-home placements such as inpatient, RTC, and CPA. In
1998, inpatient and other out-of-home placements through mental health were subject to
managed care controls and RTC placements through DYC were subject to the placement
restrictions of the juvenile justice system.




William M. Mercer, Incorporated                 26               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   27            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of Funds—Post-Managed Care—Spring 2001
The Blueprint depicted on page 29 (entitled “Arapahoe County—Spring 2001—Funding Flow”)
shows the flow of funds under the managed care pilot as they exist in the Spring of 2001. Since
most of the major changes in child welfare funding streams happened in 1997, counties (such as
Arapahoe) that initiated their managed care pilots in the Summer and Fall of 1998 experienced
few changes in funding streams other than (1) the ability of the county to retain unspent funds
from its child welfare funding block, and (2) other changes initiated at the county level.
Furthermore, Arapahoe’s carving-out of CPA treatment cost funded through the Department of
HCPF also pre-dated the initiation of its managed care pilot.

Arapahoe County is now able to retain any unspent funds from the seven combined funding
streams, thus, taking on the upside risk for the services it provides. To do this, Arapahoe County
has to allow non-managed care counties to use unspent funds from other areas of the state to
cover their overages before being eligible for these. The county is still eligible for excess Title
IV-E distributions it helps earn.

In the first two years of operating under a capped allocation, the CDHS returned unspent general
funds to the legislature. During that period of time, as well as during the third year, funding for
the child welfare block was based on “population increase.”

During the third year, SFY 2000, all Colorado counties collectively overspent their budgets by
approximately $20 million. The CDHS requested and received a $14 million supplemental
appropriation in SFY 2001, based on projected caseload growth. Because counties overexpended
so significantly in SFY 2000, the Child Welfare Allocation Committee approved allowing
managed care counties a choice to either participate in the statewide surplus distribution or retain
their own unspent funds.

Arapahoe County has chosen the MOU option where they forego the ability to retain their own
unspent funds and have elected to participate in unspent general funds distributions from the
state. Arapahoe County has chosen to be eligible for the benefits of upside risk, rather than have
the state partially protect it from downside risk. This can be seen in both the base Blueprint and
the State Funding Detail Blueprint for the county.




William M. Mercer, Incorporated                  28               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   29            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
State Funding Detail. Looking at the State Funding Detail Blueprint (see the Blueprint on page
31, entitled “Arapahoe County—Spring 2001—State Funding Detail”), another change can be
seen. The county did establish itself as an EPP county and began to receive these funds at the end
of 1998. EPP funding is an integral part of the funding for one of its managed care initiatives, the
EPP Pathways Team serving children age six and younger. EPP funds are used to manage out-of-
home placements and other costs for this population in an overall effort to move these children
more quickly into permanent living arrangements.




William M. Mercer, Incorporated                 30               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   31            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Provider Network Detail—Funding Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 33, entitled “Arapahoe County—Spring 2001—Provider
Network Funding Detail”), it can be seen that, in order to develop foster care alternatives to
expensive RTC placements, Arapahoe County has begun to designate certain
county-administered foster homes as “Pilot Homes.” These enhanced foster care placements for
children with very difficult behavioral needs include enhanced rates (four to six times higher
than the standard foster care payment) and frequent cost-sharing with CMHCs.

The most notable financial change in 2000–2001 has been the development of a “network fee”
paid to CCM. This new fee pays for CCM to perform some utilization management activities on
cases referred to their network, as well as enhanced central coordination and support of intakes
that results in expedited initial placements. This network fee arrangement has been in place since
the end of March 2000. Negotiations have also been underway with CCM to develop a case rate
arrangement as part of Arapahoe’s Title IV-E waiver demonstration.

Another change has involved more formalization of the contractual process with CMHCs and
other providers. For CORE and CPA treatment expenditures through CMHCs, Arapahoe County
has moved to a fee-for-service reimbursement system from a one-twelfth estimated payment
approach. This has reduced expenditures to only those for services actually delivered, saving the
county significantly. In addition, formal cost sharing agreements have been implemented to
document more clearly child-specific agreements made to split costs for placements involving
both child welfare and mental health needs. Enhanced financial tracking for substance abuse
services through Signal, and close tracking of CHRP referrals, has also been implemented.




William M. Mercer, Incorporated                 32              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
William M. Mercer, Incorporated   33            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. When counties implement managed care in a way that is consistent with
such expectations, we call that “process fidelity.”

Fidelity is scored in each of four domains: funding flexibility, interagency coordination,
integration of informal networks and non-traditional services, and implementation of early
intervention and prevention programs (on a four-point scale, from 1 to 4). Once a score is
calculated, it is placed within one of four fidelity categories—early, emerging, maturing, and
mature stages of fidelity.

Funding Flexibility
Arapahoe was rated in the Early Stage of fidelity on this domain. county officials reported that
they currently are not using financing approaches other than fee-for service and that they are not
using any performance incentives with providers.

Arapahoe County has been working closely over the past several months with CCM to develop
new funding approaches. One approach that has been utilized is the network fee, a preferred
provider rate, which allows Arapahoe County to obtain quicker access to residential treatment
centers for children in need of such placements. This approach is seen as a first step toward a
managed care financing approach. Arapahoe County has also been working with CCM on the
development of a case rate financing approach, in the context of the Title IV-E demonstration
project, but has not yet implemented such an approach.

Interagency Coordination
In this domain, Arapahoe County was given a rating of Maturing Stage fidelity. This was the
domain in which Arapahoe County received its highest fidelity rating. Arapahoe has developed
formal contractual relationships with several other child serving agencies and has coordinated
with several other such agencies through multi-agency service planning processes, and through
regular administrative meetings. Arapahoe has not been involved in integrating data across
child-serving systems and its point of responsibility for coordinating care has not been
articulated as clearly as would be desired in a managed care context.




William M. Mercer, Incorporated                 34               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Integration of Informal Networks and Non-Traditional Services
Arapahoe County’s fidelity score placed it in the Emerging Stage of fidelity in this domain.
Arapahoe County was found to be involved with a few informal networks of support, such as the
Children’s Advocacy and Resource Center and the Foster Parent Association, but not in
extensive or highly integrated ways.

Implementation of Early Intervention and Prevention Programs
Arapahoe County was found to be performing at the Early Stage of fidelity in this domain in that
it did not report that it was providing or developing early intervention and prevention programs
at this time.




William M. Mercer, Incorporated                35              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in Arapahoe County in
the nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in Arapahoe to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.

Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                        Figure 1
           Percentage of Children Who Experienced a Confirmed Recurrence of
                Abuse and/or Neglect within a 12-Month Period after being
                                Abused and/or Neglected

                 10
                  9
                  8
                  7
                  6                       5.2
                  5
                                                         3.2           3.4
                  4         3
                  3
                  2
                  1
                  0
                         SFY 97         SFY 98         SFY 99         SFY 00




William M. Mercer, Incorporated                  36             Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect

                                        Figure 2
                      Confirmed Abuse and Neglect per 1,000 Children
                                in the County Population

                20
                18
                16
                14
                12
                10
                 8
                                                     5            4.2
                 6                    4.2
                          3.2
                 4
                 2
                 0
                        SFY 97      SFY 98         SFY 99        SFY 00




Indicator 3: Referrals and Investigations

                                          Figure 3
                 Referrals of Alleged Abuse and Neglect per 1,000 Children
                                  in the County Population

                70

                60

                50
                          35.7                                    34.9
                40                    33.7          30.8
                30

                20

                10

                  0
                        FFY 97       FFY 98        FFY 99        FFY 00




William M. Mercer, Incorporated               37            Colorado Child Welfare Evaluation
                                                   Second Interim Implementation Status Report
                                          Figure 4
               Investigations of Alleged Abuse and Neglect per 1,000 Children
                                  in the County Population

                100
                 90
                 80
                 70        57.7           55.5
                 60
                                                         43.5           39.9
                 50
                 40
                 30
                 20
                 10
                  0
                         SFY 97         SFY 98          SFY 99         SFY 00



Indicator 4: Discharge to Reunification
It should be noted that this indicator does not directly measure reunification in that is represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.

                                          Figure 5
                 Percentage of Children Placed with Parents at Case Closure

              100
               90
               80
               70        57.7            55.5
               60
                                                        43.5           39.9
               50
               40
               30
               20
               10
                0
                        SFY 97         SFY 98         SFY 99          SFY 00




William M. Mercer, Incorporated                  38               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 5: Children in Out-of-Home Placement

                                           Figure 6
                  Children in Out-of-Home Placement at some Point during the
                    Fiscal Year per 1,000 Children in the County Population

            20
            18
            16
            14
            12
            10                                         8.2
                                                                    7.3
             8        6                     6.1
                                  5.6
             6
             4
             2
             0
                    SFY 96     SFY 97      SFY 98      SFY 99     SFY 00




Indicator 6: Length of Out-of-Home Stay

                                      Figure 7
           Out-of-Home Placement Length of Stay (in Months) for Closed Cases


             12

             10                              8.8
                                                        7.5          7.6
              8        6.7         6.7

              6

              4

              2

              0
                     SFY 96       SFY 97   SFY 98      SFY 99      SFY 00




William M. Mercer, Incorporated                   39            Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 7: Percentage in Positive Placements

                                            Figure 8
                       Percentage of Positive Placements at Case Closure
                        (Parents, Kin, Relatives, Adoption, Guardians)

             100                                                   88.1
                         84          82.8            84.4
                90
                80
                70
                60
                50
                40
                30
                20
                10
                 0
                       SFY 97       SFY 98         SFY 99         SFY 00




Indicator 8: Number of Adoptions

                                           Figure 9
                     Adoptions per 1,000 Children in the County Population

            1


          0.8


          0.6             0.53
                                              0.51

                                                                          0.38
          0.4


          0.2


            0
                        FFY 98               FFY 99                   FFY 00




William M. Mercer, Incorporated               40                 Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Indicator 9: Time to Adoption

                                      Figure 10
        Average Months from Termination of Parental Rights to Finalized Adoption

          20




          15
                                                                 12.3
                                          11.2

          10            8.6




           5




           0
                       FFY 98            FFY 99                 FFY 00




William M. Mercer, Incorporated           41               Colorado Child Welfare Evaluation
                                                  Second Interim Implementation Status Report
Outcome Comparisons

                                         Table 5
        Arapahoe County SFY 2000 Outcome Results Compared to Other Counties, to
          Benchmarks, and to the Criterion Reference Standards Set for SFY 2000
                                        Other           Non-
                                                                                                       Criterion
                                       Managed        Managed
                         Arapahoe                                                                     Reference
                                         Care        Care Large
Outcome Indicator         County                                           Benchmark                  Standard
                                       counties       counties
                                                                      National: 11%
Indicator 1:                                                                                     Acceptable:<10%
Confirmed re-abuse
                            3.4%           3.7%          3.0%         Comparison States: 12%
                                                                                                 Outstanding: <2%
                                                                      (Utah)
Indicator 2:
                                                                      National: 11.8             Acceptable: <10
Confirmed abuse per         4.2            5.0           5.1          Comparison States: 8.8
1,000 child population                                                                           Outstanding: <4
Indicator 3a:
Referrals of suspected                                                National: 72.1             Acceptable:<45
abuse per 1,000 child
                           34.9          46.2          48             Comparison States: 53.4    Outstanding: <40
population
Indicator 3b:
Investigations of                                                     National: 43.5             Acceptable:[not set]
suspected abuse per
                           17.7          26.7          31             Comparison States: 28.3    Outstanding: [not set]
1,000 child population
Indicator 4: Percent                                                  National: 59%*
discharged to              39.9%         46%           57.7%                                     Acceptable: >52%
                                                                      Comparison States: 68%*
reunification
Indicator 5:
Children in out-of-                                                   National: 8.1**            Acceptable: <8
home care per 1,000
                            7.3            7.6         10.6           Comparison States: 5.6**   Outstanding: <4
child population
Indicator 6:
                                                                      National: 10.8             Acceptable: <10
LOS in out-of-home          7.6            8.5           6.2          Comparison States: 12.3
placement (in mos.)                                                                              Outstanding: <7
Indicator 7:
                                                                      National: 83%              Acceptable: >82%
Percent discharged to      88.1%         86%           83.8%          Comparison States: 90%
positive placements                                                                              Outstanding: >90%
Indicator 8:
                                                                      National: .66              Acceptable: >.4
Adoptions per 1,000           .38           .72             .66       Comparison States: .52
child population                                                                                 Outstanding: >.9
Indicator 9:
Time from termination                                                 National: N/A              Acceptable: <18 months
of parental rights to
                           12.3          12.1          20.6           Comparison States: N/A     Outstanding: <9 months
adoption
* Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
** Benchmarks are for point in time (on September 30) versus at any time in fiscal year (Indicator 5).




William M. Mercer, Incorporated                        42                  Colorado Child Welfare Evaluation
                                                                  Second Interim Implementation Status Report
Summary of Findings

Arapahoe’s outcome results compare favorably in all areas except for one, percent discharged to
reunification. Arapahoe County’s SFY 2000 rate of reunification among children whose cases
had been closed and who had ever been in out-of-home placement was just under 40 percent.
This was below the “acceptable” criterion reference standard level and below national and
comparison state benchmarks in addition to being lower than the rate of reunification observed in
other large Colorado counties.

It should be noted, however, that reunification is not a unidirectional variable. Low and high
rates of reunification can be subject to question. As with El Paso County, while the rate of
reunification at case closure was comparatively low, the rate of positive placements for all closed
cases was high for Arapahoe County—nearly reaching the “Outstanding” criterion reference
standard. Arapahoe’s positive placement rate has continued to climb over the past few years and
was higher than the average rates of positive placements for other managed care counties and for
non-managed care counties in SFY 2000. Given the results of Indicators 4 and 7 (placements
with parents and positive placements overall), it is likely that Arapahoe County has increased its
emphasis on expedited permanency through alternatives to reunification.

A notable finding for Arapahoe County is that the number of children in out-of-home placement
at some point in the SFY, per 1,000 children in the county population, markedly decreased from
SFY 1999 to SFY 2000. At the same time, there was only a very slight increase in the average
length of stay in out-of-home placement for children whose cases were closed and who had ever
been in out-of-home placement in SFY 1999 versus children whose cases were closed in SFY
2000.

Arapahoe County continued to show low rates of confirmed abuse and neglect, as well as a low
rate of recurrence of abuse and neglect, relative to other managed care counties and to national
rates of confirmed abuse and neglect. Arapahoe’s rate of confirmed abuse and neglect per 1,000
children in the population rose slightly from SFY 1997 to SFY 2000, despite the fact that the
rates of referrals and investigations of abuse and neglect dropped slightly during that same time
period.

Finally, with regard to adoption, Arapahoe County’s rate of adoption continues to be lower than
the adoption rates for other large counties in Colorado. Its average length of time from
termination of parental rights to adoption is comparable to other managed care counties and
much lower than non-managed care counties’ combined average. Arapahoe’s Federal Fiscal Year
(FFY) 2000 performance on the adoption indicators, however, revealed findings that were not as
positive: from FFY 1998 to FFY 2000, Arapahoe County’s rate of adoption has dropped while
the average length of time to adoption has increased.




William M. Mercer, Incorporated                 43               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as RTCs, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher costs care, as well as trends over several years. The results of these
analyses are shown in this section for Arapahoe County by itself and in a later section for all of
the managed care counties together.

Figures 11 and 12 compare Arapahoe County’s allocation at the beginning of SFY 2000 to its
expenditures during that year. Figure 11 provides a comparison of the two main child welfare
funding streams—CORE services and the child welfare block. Figure 12 provides detail
regarding the funding streams that compose the child welfare block.

Figure 11 shows that Arapahoe County experienced a deficit in SFY 2000 in both CORE
services and the child welfare block. Figure 12 shows that Arapahoe’s use of more expensive
out-of-home placement, RTC and CHRP, was the main cause of the deficit. It appears that
Arapahoe was unable to operate within its allocation in SFY 2000.


                                       Figure 11
             CORE Services and Child Welfare Block Allocation vs. Expenditure*
                              Arapahoe County, SFY 2000



         $30,000,000
         $25,000,000                                                       $22,842,112 $23,781,482
         $20,000,000
                                                         $(346,860)




         $15,000,000
         $10,000,000
                        $3,351,382    $3,698,242
          $5,000,000
                $-
                                                                                                          $(939,370)




         $(5,000,000)                CORE Services                                 Child Welfare Block
        $(10,000,000)



* Before Year End Close Out Process                Allocation              Expenditure       Difference




William M. Mercer, Incorporated                                       44              Colorado Child Welfare Evaluation
                                                                             Second Interim Implementation Status Report
                                                   Figure 12
                                 Child Welfare Block Allocation vs. Expenditure
                                          Arapahoe County, SFY 2000




                                          $23,781,482
                            $22,842,112




                                                        $16,010,003


                                                                       $15,245,207
          $25,000,000




                                                                                                   $5,657,697
                                                                                      $4,259,698
          $20,000,000




                                                                                                                                           $1,761,897


                                                                                                                                                             $1,761,897
                                                                                                                           $1,116,681
          $15,000,000




                                                                                                                $810,515
          $10,000,000
           $5,000,000
                    $0
                         Total CW Block                 Total 80/20                  RTC Medicaid CHRP Medicaid 100% Admin.


                                                                 Allocation                        Expenditure


Figures 13 and 14 provide a picture of how Arapahoe County spent its child welfare block in
SFY 2000. Figure 13 shows the percentage of funds expended by budget item while Figure 14
shows the actual amount expended. Of interest is that Arapahoe and Jefferson spend a
significantly larger portion of their budgets on RTC Medicaid than other managed care counties.


                                            Figure 13
                Percentages of Child Welfare Block Expenditure * by Budget Items
                                   Arapahoe County, SFY 2000
                                                                                                                                        Out of Home Placement
                      100% Admin.                                                                                                       Services *
                      7.4%                                                                                                              31.5%

              CHRP Medicaid
              4.7%
                                                                                                                                        Subsidized Adoptions
                                                                                                                                        8.2%
                                                                                               Other
                                                                                               64.1%                                                    Special Circumstances
                                                                                                                                                        Child Care (SCCC)
             RTC Medicaid                                                                                                                               0.7%
             23.8%
                                                                                                                                                           Case Services
                                                                                                                                                           0.5%

                                                                                                                                           80/20 Admin.
                                                                                                                                           23.2%
  * Includes Medicaid Fund and before Year End Close Out Process




William M. Mercer, Incorporated                                                      45                         Colorado Child Welfare Evaluation
                                                                                                       Second Interim Implementation Status Report
                                                  Figure 14
                              Child Welfare Block Expenditure* by Budget Items
                                         Arapahoe County, SFY 2000


     $8,000,000                                                   $7,485,760


                  $5,657,697                                                                                                                $5,519,441
     $6,000,000


     $4,000,000

                                                  $1,761,897                        $1,956,413
     $2,000,000                     $1,116,681
                                                                                                    $117,042         $166,550
             $0
                  R T C M edicaid      CHRP        100% A dmin.   Out o f H o m e    Subsidized     Case Services        Special            80/20 A dmin.
                                      M edicaid                    P lacement        A do ptio ns                   C ircum s t a n c e s
                                                                    Services                                           Child Care
                                                                                                                        (SCCC)



 * Before Year End Close Out Process



Figures 15–17 illustrate the trends in out-of-home expenditures in Arapahoe County from
SFY 1998 through SFY 2000. In most areas, there are relatively modest changes except for
home- based programs, which shows a steady increase from 6.9 percent to 12.7 percent of the
out-of-home budget. Arapahoe County spends a much larger portion of its budget on home-based
programs than other managed care counties.
                                                       Figure 15
                                           Out-of-Home Placement Expenditure
                                               Arapahoe County, SFY 1998

                                                                            Home Based Program
                                         RCCF             RTC                       6.9%
                                         5.5%            16.8% Relative Foster Care
              Specialized Group                                        4.6%
                    Care
                    8.1%
                                                                                                                    Receiving
                                                                          Other                                      Home
                                                                          17.3%                                       3.2%


                                                                  Transitional                             Shelter Care
                  Family Foster Care                                                                          3.6%
                                                                   Program Independent
                        47.8%
                                                                     2.8%      Living
                                                                               0.7%




William M. Mercer, Incorporated                                         46                     Colorado Child Welfare Evaluation
                                                                                      Second Interim Implementation Status Report
                                            Figure 16
                                Out-of-Home Placement Expenditure
                                    Arapahoe County, SFY 1999

                                           RTC                        Home Based Program
                             RCCF
                                          17.4%                             8.2%
    Specialized Group        4.8%
                                                  Relative Foster Care
          Care
                                    Excludes              3.9%
          7.1%                                                                            Receiving Home
                                      RTC
                                                                                               2.5%
                                    Medicaid


                                                      Other
                                                      16.1%



                                                                                          Shelter Care
                                                        Transitional
       Family Foster Care                                                                    2.6%
                                                         Program          Independent
             50.6%                                         1.6%              Living
                                                                              1.2%




                                            Figure 17
                                Out-of-Home Placement
                                Expenditure County, SFY 2000
                                    Arapahoe

                         RCCF          RTC                            Home Based Program
                         4.4%         15.3%                                 12.7%
                                               Relative Foster Care
     Specialized Group
                                    Excludes           3.6%
           Care
                                      RTC
           5.5%                     Medicaid


                                                     Other                                 Receiving
                                                     20.3%                                  Home
                                                                                             1.8%


                                                   Transitional                      Shelter Care
        Family Foster Care                                                              2.5%
                                                    Program
              50.9%                                                    Independent
                                                      2.4%
                                                                          Living
                                                                           0.9%




William M. Mercer, Incorporated                       47                   Colorado Child Welfare Evaluation
                                                                  Second Interim Implementation Status Report
Figures 18 and 19 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 18 shows that Arapahoe County has
contracted out over half of its family foster care and group homes to CPAs. Figure 19
shows that family foster care homes are the primary contracted service.

                                          Figure 18
               Percentages and Amounts of Expenditure for Family Foster Care and
                       Group Home Services delivered by County vs. CPA
                            Arapahoe County, SFY 1998 to SFY 2000
        100%
         90%
         80%
         70%            $2,666,521                     $2,769,705                       $2,840,184
         60%
         50%
         40%
         30%
         20%            $1,427,463                     $1,503,508                       $1,775,629
         10%
          0%
                         SFY1998                         SFY1999                         SFY2000

    County Family Foster Care Homes and Group Homes          CPA Family Foster Care Homes and Group Homes




                                          Figure 19
              Percentages of Expenditure for Family Foster Care and Group Home
                          Services delivered by County and/or CPA
                           Arapahoe County, SFY 1998 to SFY 2000
       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                        SFY1999                         SFY2000


  CPA Family Foster Care Homes   CPA Group Homes      County Family Foster Care Homes    County Group Homes




William M. Mercer, Incorporated                         48                   Colorado Child Welfare Evaluation
                                                                    Second Interim Implementation Status Report
Boulder County Findings
Boulder County is located in the Denver metropolitan area, 35 miles northwest of Denver. It had
an estimated child population of 67,626 in 2000. Boulder was accepted as a managed care county
in SFY 1998.




William M. Mercer, Incorporated               49              Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
Process Findings
Process findings first examine how Boulder County is functioning as a managed care entity.
Blueprints are then presented to show the flow of children and their families and the flow of
funds through Boulder’s child welfare system. Finally, Boulder is scored on the extent of its
fidelity to managed care principles using Fidelity of Organizational Implementation Scales.

Progress in the Introduction of Managed Care Processes
Boulder County is functioning as a managed care entity as it is assuming risk by entering into an
MOU with the CDHS. In its original MOU, Boulder County was permitted to keep unspent
funds but was required to cover overages that might occur. It is expected to perform at a certain
level on specific outcomes.

In 2001, the CDHS offered counties a choice between keeping any of their own unspent funds or
being eligible for participation in the distribution of unspent funds from other counties at the end
of the year. Boulder County chose to give up the ability to retain its own unspent funds in order
to participate in surplus distributions from the state.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                   50               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The main focus of Boulder County’s managed care initiative has been on utilization management
and the development of its provider network. While gains in both areas have been documented,
the area that is the most developed is utilization management.

Boulder is unique in that it has formed a multi-agency organizational entity that operates with
blended funds from each of these agencies. The goals of this entity, called Integrated Managed
Partnership for Adolescent Community Treatment (IMPACT), are to:

§   reduce out-of-home placements and lengths of stay in out-of-home placements;
§   ensure that levels of care provided are the appropriate levels of care;
§   increase permanency;
§   increase educational involvement;
§   reduce contacts with the criminal justice system; and
§   increase community connectedness.

IMPACT’s key function is to manage out-of-home placements for social services, youth
corrections and mental health. IMPACT employs two teams, with one team focusing on
high-risk youth in the community and the other on children in placements.

The development of IMPACT represents an important step in Boulder County’s development as
a managed care entity. It is unparalleled in its promotion of joint ownership of high-risk children
by the major agencies in the county. This addresses a longstanding problem in child welfare.

In refining the utilization review process, Boulder might consider changing review intervals from
fixed periods of time to periods that are driven by the expected outcome for individual children.
A fixed period might be used as a maximum.

In regard to network capacity, Boulder County has not determined the capacity that it requires.
As a result, it is limited in developing a configuration of services that addresses the needs of
children and families now and in the future. Financial incentives for the development of services
that represent present best practices are difficult to implement without knowing the overall
capacity that is required.




William M. Mercer, Incorporated                 51               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Although Boulder County has not determined its required network capacity, it has moved to
address readily perceived gaps in its network. For example, Boulder has developed a kinship
program and has increased funding for other targeted programs. The funds directed to subsidized
adoptions have doubled, as have the funds for permanency for children with developmental
disabilities. Other new initiatives include a gang initiative, specialized services for girls in the
juvenile justice system (with placements and specialized treatment options), and an independent
living program with a local CPA. Grant funding has also secured enhanced substance abuse
services through the Boulder County Health Department.

Boulder County is not using financial incentives or performance indicators as part of its
operation. Boulder is planning on collecting data on the cost of care and relating it to outcomes.
Boulder intends to determine which providers are yielding the best outcomes per dollar.

In regard to the other key elements of managed care listed above, Boulder County is still in the
development stage. Boulder needs to develop a clearly identifiable quality improvement system
so that it may continuously improve its services and those of its providers. It also needs to
improve its information system as the one present during most of this evaluation, which is
operated by the CDHS on behalf of the counties, is not adequate for managing Boulder’s
managed care program. The CDHS has recently implemented a new system, TRAILS, which has
the potential to yield more useful and timely information.

Process Flow
To illustrate the flow of processes involving people and funds in Boulder County, we have
constructed eight Blueprints. These show:
1. the flow of persons through the county in SFY 1997, prior to the initiation of managed care
   reforms;
2. the overall flow of persons through the county in SFY 2001, midway through the fourth year
   of managed care reforms;
3. additional detail on the flow of persons through the county department;
4. additional detail on the flow of persons through the county’s provider network;
5. the flow of funds through the county in SFY 1997, prior to the initiation of managed care
   reforms;
6. the overall flow of funds through the county in SFY 2001, midway through the fourth year of
   managed care reforms;
7. additional detail on the flow of funds from the state to the county, including financial
   reporting; and
8. additional detail on the flow of funds through the county’s provider network, including
   risk-sharing arrangements.




William M. Mercer, Incorporated                  52               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
We have used six colors in the Blueprints:
1. Black is used for the organizations involved in paying for, managing, and providing child
   welfare services.
2. Red is used for the unmanaged flow of people through the system and includes county
   populations, populations in need, referral groups, movement through different teams and
   areas within the county child welfare department, and referrals to community providers.
3. Green depicts child welfare funds that flow directly to the county child welfare department,
   as well as the state and federal funds that flow through the CDHS to the county child welfare
   department.
4. Blue depicts the flow of non-child welfare funds that are critical to providing needed care to
   recipients of child welfare services.
5. Purple depicts the managed flow of both persons and funds through the county child welfare
   managed care pilot initiatives.
6. Yellow depicts the court system in each county.

In the Blueprints, the flow of people through the system is depicted in a left to right fashion,
beginning with the entire population potentially served by the child welfare system, narrowing to
the group of persons actually served, and following these people through the organizations and
services they would typically encounter as they pass through the system. The flow of funds,
however, goes from the top to the bottom of each chart. The drawing traces the flow of funds
from the payers that fund the system, to the administrative entities that allocate and account for
these funds, to the county child-serving agencies and providers that ultimately spend and manage
them.

A set of symbols is also used. They are illustrated in the following diagram.

Organizational Symbols:
               = Federal, state or county payors
               = State agencies
               = Local agencies or providers
               = Local entities that bears risk

Risk Symbols:
  S   = Shared risk of financial utilization
  F   = Full risk of financial utilization

Out-of-Home Placement Symbol:
  $            = Most expensive out-of-home
                   placements (RTC, CPA)
Financial Reporting Symbol:
   Financial
  Reporting
  by County    = Locus of financial reporting


William M. Mercer, Incorporated                    53            Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Flow of People—Pre-Managed Care—Spring 1997
The Blueprint depicted on page 55 (entitled “Boulder County—Spring 1997—Flow of People”)
shows the flow of people through Boulder County’s child welfare system in the Spring of 1997,
prior to the initiation of their managed care pilot later that year. Beginning on the far left, the
overall county population breaks down into the smaller group of persons who, at any point in
time, were in need of child welfare services.

From that group, three types of referrals were typically seen by the Boulder County Department
of Social Services (DSS): youth in conflict, youth with delinquency problems, and children with
reported instances of abuse or neglect. Referrals passed through one of two intake units—
Adolescent or Child—depending on their age. Youth in conflict, and with delinquency concerns,
were generally adolescents and passed through the Adolescent Intake; abuse and neglect cases
went to either Child or Adolescent Intake depending on their age. The intake units screened out
cases that clearly did not meet criteria and referred them to other services in the community.
Additionally, persons with basic needs were often referred to the county’s TANF program. Child
welfare cases meeting criteria were referred to either ongoing or specialized (e.g., home-based)
treatment teams for services.

Many children, youth, and families served by the county required additional services provided by
agencies in the community. Workers on ongoing and specialized teams were able to directly
access expensive out-of-home placements in RTCs, CPAs, RCCFs and county-administered
foster homes, as well as outpatient mental health services through the Mental Health Center of
Boulder County. Youth and families in need of outpatient substance abuse services accessed
these through the Boulder County Health Department. Detention and commitment beds through
the juvenile justice system were accessed through the Probation Department. Children needing
inpatient services had to go through the CMHC, which managed all indigent and Medicaid-
funded mental health benefits. While some high need or multi-agency involved youth accessed
multi-agency planning from time-to-time, the majority of external referral decisions were not
coordinated.

A critical factor for any involuntary case served by the county is the court system. Individual
workers presented recommendations for services to the court and the court implemented them as
it saw fit. Voluntary cases were not subject to court involvement. While the Blueprints highlight
the role of the court at the point where it impacted decisions regarding referrals to out-of-home
placements versus other community services, the court’s involvement with involuntary cases was
much broader, beginning at or prior to intake and extending throughout each family’s final
disposition.




William M. Mercer, Incorporated                  54               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   55            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of People—Post-Managed Care—Spring 2001
In the fourth year of its implementation of managed care reforms, the flow of persons through
the Boulder County child welfare system has changed significantly. A review of the Base
Blueprint for Spring 2001 (see the Blueprint depicted on page 57 entitled “Boulder County—
Spring 2001—Flow of People”) shows the same basic population served but reorganization
within the Department and its provider network. To review these changes with greater
specificity, two detailed Blueprints were created for this year: a Department Internal Detail
Blueprint and a Provider Network Detail Blueprint.




William M. Mercer, Incorporated                56              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
William M. Mercer, Incorporated   57            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Department Internal Detail. Looking at the Department Internal Detail Blueprint (see the
Blueprint depicted on page 59 entitled “Boulder County—Spring 2001—Department Internal
Detail”), it can be seen that the county population still breaks down on the left into a smaller
group of persons in need, who, in turn, sort into the three primary groups of youth and families
served by the county: youth in conflict, youth with delinquency concerns, and children with
reports of abuse or neglect.

However, within the county, the flow of people differs markedly:
§    Rather than separate intake units for children and adolescents, the county has integrated all
    child welfare intakes through a single point of access.
§     Additionally, intake workers are able to provide prevention and early intervention services,
    following families with lower needs for up to 30 days without opening the case in order to
    connect them with other community services and avoid ongoing child welfare services. Not all
    families take advantage of this voluntary service but those that do are often able to have their
    needs met without additional resources.
§     Families with basic needs are still referred to the county’s TANF program but there is greater
    coordination between the TANF and child welfare programs. Expanded TANF child care
    initiatives, substance abuse services, and women’s services also provide increased support to
    families served by child welfare staff, promoting both enhanced stability for these families and
    better coordination of care.
§    Specialized teams have been expanded using unspent funds, including In-Home Wraparound
    Services and Sex Offender Services.

There is also more internal support of coordination with other child- and family-serving
agencies. DSS intake workers receive cross-systems training. Joint cross-systems training
involving DSS staff providing ongoing care and their counterparts in other county human service
agencies has also helped increased ongoing coordination regarding cases.




William M. Mercer, Incorporated                   58               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
William M. Mercer, Incorporated   59            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Provider Network Detail—People Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 62 (entitled “Boulder County—Spring 2001—Provider
Network Detail—People Flow”), the primary change enacted in Boulder County has been the
creation of the IMPACT program. Administered under the Mental Health Center of Boulder
County (MHCBC), IMPACT is a new, independent program in Boulder County with its own
director and staff. It is overseen by a board composed of the leadership of Boulder DSS, the
Mental Health Center of Boulder, the Colorado Division of Youth Corrections (DYC), the
District Court, Probation, the Boulder County Health Department, Community Corrections,
Community Services, and local school districts. The Board is responsible for program
philosophy, vision, and policy.

The IMPACT Planning Committee is responsible to the Board for implementation of that vision
through programs and program operations. This includes program development, oversight, and
budgeting. The IMPACT Planning Committee is comprised of senior program leaders from each
of the partner organizations.

All CPA, RTC, inpatient and detention referrals, as well as high risk cases, are referred by DSS
to IMPACT for staffing and intensive case management. IMPACT’s Placement Review Team
staffs all cases and authorizes any actual care provided. The Placement Review Team includes
senior managers from all the participating agencies, and the review includes meeting with the
family or their representative. Through this review process, many placement referrals are
diverted to either lower level placements (county-administered foster homes) or community-
based alternatives (such as the MHCBC). Originally working with only adolescents, IMPACT
has been able to use unspent funds to hire additional staff and expand its focus to younger
children.

Once a placement is authorized, IMPACT case managers follow the case, facilitate ongoing
inter-agency planning with the family, and conduct ongoing, concurrent review of the
placement’s efficacy and appropriateness. Backed up by the authority to deny payment for
ongoing care deemed unnecessary, these utilization reviewers work to focus care on the goal of
discharge back to the community or lower level out-of-home placement.

IMPACT’s Community Evaluation Team staffs high-risk cases within the community in a
similar fashion. Here, the multi-agency staffing, authorization process, and ongoing case
management by IMPACT case managers is focused on avoiding out-of-home placement. There
is also a special emphasis on the identification and support of alternative placements with kin. As
with the Placement Review Team, the focus of this team has been expanded since early 1999 to
include children younger than adolescence.

Through IMPACT, Boulder County DSS’s highest need youths receive multi-agency planning
and dedicated intensive case management to coordinate their care across the array of providers,
from inpatient facilities to RTCs to outpatient care. The partners also benefit from each other’s
expertise. The participation of the leadership of all Boulder County child and family serving
agencies on the IMPACT Board, the senior leadership from these agencies on the Placement
Review and Community Evaluation Teams, and staff from each agency in the ongoing care
ensures that agencies work together, resolve conflicts, and avoid duplication of services.


William M. Mercer, Incorporated                 60               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Also, critical to the efforts of IMPACT is the participation of Boulder’s District Court, both
through the participation of the Chief Judge on the IMPACT Board and daily coordination for
high need cases by IMPACT case managers. In addition, cases not served by IMPACT have
greater access to their county caseworkers given the shift of high need cases to IMPACT,
allowing for closer coordination with the Court across the board.

In 2000–2001, the focus has been upon consolidation and refinement of Boulder County DSS
and IMPACT programs. The administrative leadership of DSS reorganized, with responsibility
consolidated under one of the county’s three child welfare administrators for programs most
central to the DSS’s managed efforts: intake, ongoing adolescent services, and DSS
representation to the IMPACT Planning Council.

The IMPACT partnership continues to use unspent funds and grants to create new collaborative
programming in the network of care, including a gang initiative, specialized services for girls in
the juvenile justice system (with placements and specialized treatment options), and an
independent living program with a local CPA. Grant funding has also secured enhanced
substance abuse services through the Boulder County Health Department.




William M. Mercer, Incorporated                 61               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   62            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of Funds—Pre-Managed Care—Spring 1997
The Blueprint depicted on page 65 (entitled “Boulder County—Pre-Managed Care—Spring
1997”) shows the flow of funds through the Boulder County child welfare system prior to the
initiation of the managed care pilot in late 1997. In July 1997, funds from various federal sources
(Title IX and XX) and the state general fund were combined by the CDHS and passed on to
Boulder County. The bulk of these funds flowed through separate funding streams, which had to
be separately tracked and accounted for by the county. Furthermore, unspent funds in one
category could not be used in other categories, creating an incentive to keep spending in line
with historical patterns. There were eight separate streams, including:
§    two streams funding county administration (to fund county staffing, primarily), one
     comprised of 100 percent dollars and one of 80 percent/20 percent county funds;
§    a large stream of funds for out-of-home placements (OOHP);
§    a significant funding stream for CORE services, originally developed to fund alternatives to
     out-of-home care, such as family preservation, outpatient substance abuse, and mental health
     services and day treatment; and
§    four smaller funding lines (e.g., Special Circumstances Child Care, subsidized adoption
     funds, etc.).

Boulder County also received other specialized state funding, including Expedited Permanency
Planning funds, but these other sources of funds are not depicted.

In addition to the budgeted funding streams, the CDHS, and through them the State General
Fund, had the full risk for any cost overages that exceed the state’s ability to shift unspent funds
from other areas of the state. In addition, no county was able to keep any unspent funds it might
generate. As a result, counties had little incentive to limit their spending of child welfare funds
and many spent their entire allocation and more.

Boulder County, in turn, contracted with various community agencies for services. In addition to
funding streams flowing through DSS, Boulder County’s child welfare system also depended on
services paid for by other agency funding. These included:
§     Medicaid (Title XIX) funding paid through the HCPF to RTC providers for the treatment
    component of the services provided in these facilities—The County DSS pays directly for only
    room and board. Treatment costs are paid directly from the State Department of Human
    Services to HCPF but are then charged back to the county at year end against its allocation.
    The county has no control over the rates paid to providers since they are determined through a
    cost-based approach within available appropriations.




William M. Mercer, Incorporated                  63               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
§    Medicaid (Title XIX) funding for mental health services transferred through the HCPF to the
    Mental Health arm of the DHS Office of Health and Rehabilitation Services—In Boulder
    County, funding for all Medicaid inpatient and outpatient mental health services has been
    capitated and, as a result, flows through the entity awarded the MHASA contract for that area
    of the state, Behavioral Healthcare, Incorporated (BHI).
§    DYC SB 94 funding for preventing detention placements for which some youth served by DSS
    are eligible—Services funded may be similar to CORE services.
§    ADAD funding through the DHS Office of Health and Rehabilitation Services—In Boulder
    County, these funds are managed by the Boulder County Health Department, which also
    directly provides substance abuse services.

As can be seen in the Blueprint, child welfare funds were the least managed of the primary
funding streams for expensive out-of-home placements, such as inpatient, RTC, and CPA. In
1997, inpatient and other out-of-home placements through mental health were subject to
managed care controls and RTC placements through DYC were subject to the placement
restrictions of the juvenile justice system.




William M. Mercer, Incorporated                 64               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   65            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of Funds—Post-Managed Care—Spring 2001
The Blueprint depicted on page 67 (entitled “Boulder County—Spring 2001—Funding Flow”)
shows the flow of funds under the managed care pilot as they exist in the Spring of 2001. The
primary state and federal funding sources flowing through the CDHS remain essentially
unchanged; however, the flow of funds from the CDHS has changed dramatically. In addition to
creating the managed care pilots, SB-97 218 collapsed seven of the categorical funding streams
that existed in the Spring of 1997 into a single child welfare block. Only CORE services
remained as a separate primary funding stream with separate accountability and accounting to
prevent funds from being diverted to other uses. This block funding approach was implemented
for all counties, not just managed care counties.

All Colorado counties, including Boulder, are now responsible for any overages from budgeted
expenditures, shifting full downside risk from the state CDHS and General Fund to Boulder DSS
and the county, as designated by the full risk triangle in the DSS box. In addition, as a managed
care pilot, Boulder is able to retain any unspent funds from the seven combined funding streams,
thus, taking on the upside risk for the services it provides. To do this, Boulder has to allow
non-managed care counties to use unspent funds from other areas of the state to cover their
overages before being eligible for these. The county is still eligible for excess Title IV-E
distributions it helps earn.

In the first two years of operating under a capped allocation, the CDHS returned unspent general
funds to the legislature. During that period of time, as well as during the third year, funding for
the child welfare block was based on “population increase.”

During the third year, SFY 2000, all Colorado counties collectively overspent their budgets by
approximately $20 million. The CDHS requested and received a $14 million supplemental
appropriation in SFY 2001, based on projected caseload growth. Because counties overexpended
so significantly in SFY 2000, the Child Welfare Allocation Committee approved allowing
managed care counties a choice to either participate in the statewide surplus distribution or retain
their own unspent funds.

Boulder County has chosen the MOU option where they give up the ability to retain their own
unspent funds and instead are able to participate in unspent funds distributions from the state.
Boulder has chosen to have the state partially protect it from downside risk rather than be eligible
for the benefits of upside risk. This can be seen in both the base Blueprint and the State Funding
Detail Blueprint for the county.




William M. Mercer, Incorporated                  66               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   67            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
State Funding Detail. Looking at the State Funding Detail Blueprint (see the Blueprint depicted
on page 69 (entitled “Boulder County—Spring 2001—State Funding Detail”), another systemic
issue related to the cost overruns can be seen. Boulder DSS now earns a federal Medicaid (Title
XIX) match on the treatment portion of its CPA costs, sending them through a Medicaid CPA
transfer to the state HCPF to the Mental Health Center of Boulder County (MHCBC) and,
through them, to CPA providers. Boulder County takes advantage of this in order to receive a
roughly 50 percent federal match on the treatment component of the care provided in CPAs. Part
of the increased revenue realized through this change has been used to increase the rates paid to
county-administered foster care homes to help the county better compete with CPAs to attract
foster parents.




William M. Mercer, Incorporated                68               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
William M. Mercer, Incorporated   69            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Provider Network Detail—Funding Flow. Looking at the Provider Network detail Blueprint
(see the Blueprint depicted on page 71 (entitled “Boulder County—Spring 2001—Provider
Network Detail—Funding Flow”), it can be seen that the biggest change has been the creation of
IMPACT. This has involved the combining of DSS out-of-home placement funds, DYC funds
for detention and commitment beds, and mental health state and Title XIX (Medicaid) funds for
inpatient care into a single budget to fund IMPACT. All DYC funding for detention and
commitment beds now flows through IMPACT, although projected detention and commitment
beds are pre-paid, and additional beds can be purchased on a per case basis. The integration of
DYC funding and services into this coordinated project allows the DYC system to benefit from
the same management approaches previously available only to child welfare and mental health
services.

In addition, the Boulder County Health Department, Probation Department, and Community
Corrections contribute funding to IMPACT. DYC SB 94 funds are also used to support
IMPACT. In return, these agencies enjoy the leverage of their combined purchasing power with
providers, as well as enhanced coordination. DSS manages all out-of-home placements for the
partners and MHCBC manages all inpatient and community-based services. As a result, MHCBC
still pays directly for inpatient care in coordination with the overall IMPACT program. All
partners also jointly share the risk of any cost overages in out-of-home care and work together to
determine the use of any unspent funds realized, as depicted by the shared risk triangles on each.

IMPACT, in turn, purchases inpatient mental health, detention and commitment, RTC, RCCF
and CPA beds, as well as outpatient mental health, substance abuse, probation, and community
corrections services.

Two additional changes can be seen:
§     Grant funding has been secured to support new programs leveraging unspent funds resulting
    from IMPACT. The use of unspent funds allows the county to match grant money in ways that
    categorical state and federal funds formerly could not. This results in both additional money
    into the system and new program development.
§    Unspent TANF funds are now available to help support Title XX expenses related to
    IMPACT programming.

Beginning in 2000–2001, IMPACT began to manage all DSS provider network service funds. In
addition, county foster home, kinship home, specialty provider, and DSS-funded mental health
service funds are now managed by IMPACT.

In the last year, the IMPACT partnership targeted additional creative funding initiatives, taking
advantage of the spending flexibility of unspent funds resulting from IMPACT. For example,
IMPACT has developed a new joint, independent living program with a local CPA. The program
operates with staff funded by the CPA, DSS staff loaned to the program, and a city-funded
house. Funding comes from diverted out-of-home placement funds. In addition, IMPACT
matching funds continue to leverage new grants, including one for substance abuse services
through the Boulder County Health Department.


William M. Mercer, Incorporated                 70               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   71            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. When counties implement managed care in a way that is consistent with
such expectations, we call that “process fidelity.”

Process fidelity is scored in each of four domains: funding flexibility, interagency coordination,
integration of informal networks and non-traditional services, and implementation of early
intervention and prevention programs (on a four-point scale, from 1 to 4). Once a score is
calculated, it is placed within one of four fidelity categories: early, emerging, maturing, and
mature stages of fidelity.

Funding Flexibility
Boulder was rated in the Mature Stage of fidelity in this domain. Boulder has implemented a
very innovative model in which child serving agencies in the county pool funding to support the
operation of the IMPACT program. IMPACT manages all out-of-home placement cases and
attempts to help children live in the least restrictive alternative to placement as quickly as
possible. Boulder also reported that they are currently using incentives with several providers, as
well as risk-based contracting (sub-capitation and case rate financing) with several providers.
Finally, Boulder County indicated that it makes significant use of flexible funds with families.

Interagency Coordination
In this domain, Boulder County was given a rating of Mature Stage fidelity. Boulder County’s
child serving agencies are highly integrated, particularly through the IMPACT team. The
IMPACT program provides a vehicle for intensive collaboration and coordination between child
serving agencies in Boulder County.

Integration of Informal Networks and Non-Traditional Services
Boulder County’s fidelity score placed it in the highest end of the Emerging Stage of fidelity in
this domain. Boulder County was found to be involved in several different ways with
information networks of support, but not always in extensive or highly integrated ways. Boulder
reported that they have attempted to access informal networks through developing a kinship
program and family meetings designed to access the natural supports of the child and family.
They also reported having developed some informal relationships with churches and other
private child and family-serving agencies.

Implementation of Early Intervention and Prevention Programs
Boulder County was found to be performing at the Mature Stage of fidelity in this domain. The
county was rated as delivering a comprehensive array of prevention and early intervention
programming.




William M. Mercer, Incorporated                 72               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in Boulder County in the
nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in Boulder to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.




William M. Mercer, Incorporated                73               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                        Figure 20
           Percentage of Children Who Experienced a Confirmed Recurrence of
                Abuse and/or Neglect within a 12-Month Period after being
                                Abused and/or Neglected

                                    9.7
          10
           9
           8                                                            6.3
           7          5.7                            5.6
           6
           5
           4
           3
           2
           1
           0
                    SFY 97         SFY 98          SFY 99             SFY 00



Indicator 2: Substantiated Abuse and Neglect

                                      Figure 21
        Confirmed Abuse and Neglect per 1,000 Children in the County Population

           20

           15

                                                     9                 8.6
           10                       7.9
                       6.3

            5

            0
                     SFY 97        SFY 98          SFY 99            SFY 00




William M. Mercer, Incorporated             74            Colorado Child Welfare Evaluation
                                                 Second Interim Implementation Status Report
Indicator 3: Referrals and Investigations

                                            Figure 22
                   Referrals of Alleged Abuse and Neglect per 1,000 Children
                                    in the County Population

          70           60.5
                                        55.9             54.1                55.6
          60
          50
          40
          30
          20
          10
           0
                     FFY 97           FFY 98           FFY 99            FFY 00



                                            Figure 23
                     Investigations of Alleged Abuse and Neglect per 1,000
                              Children in the County Population

            50

            40
                        27.9
            30                          24.8            23.7                 24

            20

            10

               0
                       FFY 97          FFY 98          FFY 99            FFY 00




William M. Mercer, Incorporated                 75            Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
Indicator 4: Discharge to Reunification
It should be noted that this indicator does not directly measure reunification in that it represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.

                                         Figure 24
                 Percentage of Children placed with Parents at Case Closure


           100
            90
            80
            70
            60         51.6                52
                                                             43.9               43.3
            50
            40
            30
            20
            10
             0
                      SFY 97             SFY 98             SFY 99             SFY 00




William M. Mercer, Incorporated                   76              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 5: Children in Out-of-Home Placement

                                          Figure 25
                 Children in Out-of-Home Placement at some Point during the
                   Fiscal Year per 1,000 Children in the County Population

         10
          9
          8
          7                                                   6.1
                                                                             5.7
          6          5
          5                                    4
                                   3.5
          4
          3
          2
          1
          0
                  SFY 96          SFY 97    SFY 98          SFY 99         SFY 00




Indicator 6: Length of Out-of-Home Stay

                                        Figure 26
              Out-of-Home Placement Length of Stay (in Months) for Closed Cases

         20
         18
         16
         14
         12
                                              8.6
         10                         7.8                       6.8
                    6.8
          8                                                                   5.9
          6
          4
          2
          0
                   SFY 96         SFY 97     SFY 98         SFY 99         SFY 00




William M. Mercer, Incorporated              77                Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Indicator 7: Percentage in Positive Placements

                                          Figure 27
                      Percentage of Positive Placements at Case Closure
                       (Parents, Kin, Relatives, Adoption, Guardians)

          100          85.8
                                       78.7                 82.9               81.9
           90
           80
           70
           60
           50
           40
           30
           20
           10
            0
                     SFY 97          SFY 98               SFY 99             SFY 00




Indicator 8: Number of Adoptions

                                         Figure 28
                    Adoptions per 1,000 Children in the County Population


                1

           0.8

           0.6
                                                   0.4                       0.43

           0.4            0.31


           0.2

                0
                         FFY 98               FFY 99                      FFY 00




William M. Mercer, Incorporated               78                  Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 9: Time to Adoption

                                      Figure 29
        Average Months from Termination of Parental Rights to Finalized Adoption

           20


           15                                  13.1

                                                                         9.9
           10              8.3



             5


             0
                        FFY 98             FFY 99                     FFY 00




William M. Mercer, Incorporated           79                   Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Outcome Comparisons

                                        Table 5
Boulder County SFY 2000 Outcome Results Compared to Other Counties, to Benchmarks,
              and to the Criterion Reference Standards Set for SFY 2000
                                        Other           Non-
                                                                                                       Criterion
                                       Managed        Managed
                          Boulder                                                                     Reference
                                         Care        Care Large
Outcome Indicator         County                                           Benchmark                  Standard
                                       counties       counties
                                                                      National: 11%
Indicator 1:                                                                                     Acceptable:<10%
Confirmed re-abuse
                            6.3%           2.9%          3.0%         Comparison States: 12%
                                                                                                 Outstanding: <2%
                                                                      (Utah)
Indicator 2:
                                                                      National: 11.8             Acceptable: <10
Confirmed abuse per         8.6            4.5           5.1          Comparison States: 8.8
1,000 child population                                                                           Outstanding: <4
Indicator 3a:
Referrals of suspected                                                National: 72.1             Acceptable:<45
abuse per 1,000 child
                           55.6          41.8          48             Comparison States: 53.4    Outstanding: <40
population
Indicator 3b:
Investigations of                                                     National: 43.5             Acceptable:[not set]
suspected abuse per
                           24.0          24.6          31             Comparison States: 28.3    Outstanding: [not set]
1,000 child population
Indicator 4: Percent                                                  National: 59%*
discharged to              43.3%         45%           57.7%                                     Acceptable: >52%
                                                                      Comparison States: 68%*
reunification
Indicator 5:
Children in out-of-                                                   National: 8.1**            Acceptable: <8
home care per 1,000
                            5.7            7.8         10.6           Comparison States: 5.6**   Outstanding: <4
child population
Indicator 6:
                                                                      National: 10.8             Acceptable: <10
LOS in out-of-home          5.9            8.6           6.2          Comparison States: 12.3
placement (in mos.)                                                                              Outstanding: <7
Indicator 7:
                                                                      National: 83%              Acceptable: >82%
Percent discharged to      82%           87%           83.8%          Comparison States: 90%
positive placements                                                                              Outstanding: >90%
Indicator 8:
                                                                      National: .66              Acceptable: >.4
Adoptions per 1,000           .43           .67             .66       Comparison States: .52
child population                                                                                 Outstanding: >.9
Indicator 9:
Time from termination                                                 National: N/A              Acceptable: <18 months
of parental rights to
                            9.9          12.3          20.6           Comparison States: N/A     Outstanding: <9 months
adoption
* Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
** Benchmarks are for point in time (on September 30) versus at any time in fiscal year (Indicator 5).




William M. Mercer, Incorporated                        80                  Colorado Child Welfare Evaluation
                                                                  Second Interim Implementation Status Report
Summary of Findings
Boulder County continues to excel in its shorter average lengths of stay in out-of-home
placement relative to other Colorado counties and to national rates of out-of-home placement and
lengths of stay. The County’s short average length of stay in out-of-home placement has declined
each year of the managed care pilot after increasing steadily prior to the start of the pilot. Its
overall rate of out-of-home placement has increased slightly since the initiation of managed care.

Like some of the other managed care counties, Boulder’s reunification rate was somewhat lower
than the non-managed care counties’ reunification rates. That may be due, in large part, to the
fact that Boulder has low rates for out-of-home placement. This may be because Boulder does
not utilize out-of-home placements as much as other counties in Colorado and nationally,
children placed out-of-the home may be the most difficult to reunify.

Boulder’s rates of referrals and investigations have dropped slightly from FFY 1997 to
FFY 2000, while its rate of confirmed abuse and neglect has risen during that same time period.
Boulder’s rates of confirmed abuse and neglect and recurrence of abuse and neglect are low
compared to national and comparison state benchmarks. Boulder’s rates also are well within the
criterion reference standards’ acceptable ranges; however, they are higher compared to the other
large counties in Colorado.

When Boulder County’s rates of abuse and neglect and its rates of recurrence of abuse and
neglect are removed from the managed care county totals, the managed care county rates are
lower than the non-managed care counties’ rates. Boulder County has a higher rate of referrals
than other managed care counties and a slightly lower rate of investigations. It is not yet clear
why Boulder’s rate of confirmed abuse and neglect is higher than for other managed care
counties. It is interesting that Boulder’s rate of out-of-home placement is relatively low, despite
the relatively higher rate of confirmed abuse and neglect compared to other counties.

Finally, with regard to the adoption indicators, Boulder continues to evidence a lower rate of
adoption but had a shorter average time to adoption in FFY 2000 compared to other large
counties in Colorado. Boulder has steadily increased its rate of adoption from FFY 1998 to
FFY 2000.




William M. Mercer, Incorporated                  81               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as residential treatment centers, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher cost care, as well as trends over several years. The results of these
analyses are shown in this section for Boulder County by itself and in a later section for all of the
managed care counties together.

Figures 30 and 31 compare Boulder County’s allocation at the beginning of SFY 2000 to its
expenditures for the year. Figure 30 provides a comparison of the two main child welfare
funding streams, CORE services and the child welfare block. Figure 31 provides detail regarding
the funding streams that compose the child welfare block.

                                    Figure 30
          CORE Services and Child Welfare Block Allocation vs. Expenditure*
                            Boulder County, SFY 2000




           $18,000,000




                                                                                                  $(1,985,578)
           $14,000,000                                                             $11,900,347
                                                                      $9,914,769
                                                      $(115,752)




           $10,000,000
            $6,000,000      $2,343,761 $2,459,513
            $2,000,000
           $(2,000,000)
                                      CORE Services                       Child Welfare Block


* Before Year End Close Out Process                 Allocation           Expenditure         Difference




William M. Mercer, Incorporated                     82                      Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Figure 31 shows that Boulder County experienced a substantial deficit in SFY 2000, primarily in
the child welfare block. Figure 31 shows that Boulder’s deficit resulted from the overuse of
RTC, CHRP, and its 80/20 budget. Boulder was the only county that overspent in all of these
areas. It appears that Boulder County was unable to operate within its allocation in SFY 2000.

                                               Figure 31
                         Child Welfare Block Allocation vs. Expenditure before
                                Close Out, Boulder County, SFY 2000
      $30,000,000

      $25,000,000
                                        $11,900,347
                           $9,914,769




      $20,000,000


                                                                    $8,256,342
                                                       $6,793,790
      $15,000,000




                                                                                                $1,592,854
                                                                                   $1,451,557




                                                                                                                                         $1,364,118

                                                                                                                                                      $1,364,118
      $10,000,000




                                                                                                                          $687,033
                                                                                                               $305,304
       $5,000,000

                $0
                      Total CW Block                  Total 80/20                RTC Medicaid CHRP Medicaid                           100% Admin.
                                                       Allocation                    Expenditure


Figures 32 and 33 provide a picture of how Boulder County spent its child welfare block in
FY 2000. Figure 32 shows the percentage of funds expended by budget item, while Figure 33
shows the actual amount expended. Of interest is that Boulder County spends only a small
portion of its budget on RTC but spends a greater portion on CHRP than other managed care
counties.

                                           Figure 32
                Percentages of Child Welfare Block Expenditure* by Budget Items
                                   Boulder County, SFY 2000


                                                                                                                          Out of Home Placement
                                                                                                                          Services
                100% Admin.                                                                                               17.3%     Subsidized
                11.5%                                                                                                               Adoptions
                                                                                                                                    6.7%
                                                                                                                                Case Services
         CHRP Medicaid                                                                                                          0.9%
                                                                                   Other
         5.8%
                                                                                   69.4%                                        Special Circumstances
                                                                                                                                Child Care (SCCC)
                                                                                                                                1%
               RTC Medicaid                                                                                                    80/20 Admin.
               13.4%                                                                                                           43.1%

         * Includer Medicaid Fund and before Year End Close Out Process




William M. Mercer, Incorporated                                                  83                             Colorado Child Welfare Evaluation
                                                                                                       Second Interim Implementation Status Report
                                                Figure 33
                            Child Welfare Block Expenditure* by Budget Items
                                        Boulder County, SFY 2000
   $6,000,000
                                                                                                                                            $5,133,436


   $4,000,000


                                                                 $2,064,001
   $2,000,000   $1,592,854
                                               $1,364,118
                                  $687,033                                          $792,256
                                                                                                    $107,162         $159,487
           $0
                R T C M edicaid     CHRP        100% A d m in.   Out o f H o m e     Subsidized     Case Services        Special             80/20 A dmin.
                                   M edicaid                      P lacement         A do ptio ns                   C ircum s t a n c e s
* Before Year End Close Out Process                                Services                                            Child Care
                                                                                                                         (SCCC)




Figures 34–36 illustrate the trends in out-of-home expenditures in Boulder County from
SFY 1998 through SFY 2000. Boulder and Jefferson are unique in the proportion of expenditures
for out-of-home care that is devoted to receiving homes and shelter care. Boulder is unique in
that it spends a smaller proportion of its budget for family foster care than other managed care
counties. Boulder and Jefferson spend a far smaller percentage of their budgets on family foster
care than other managed care counties.

                                                     Figure 34
                                        Out-of-Home Placement Expenditure
                                             Boulder County, SFY 1998
                  RCCF
                  27.6%                                  RTC                   Home Based Program
                                                        10.9%                                                                               Receiving
                                                                                     0.2%
                                                                                                                                             Home
                                               Excludes            Relative Foster                                                            3.0%
                                                 RTC                    Care
                                               Medicaid
                                                                        4.2%


Specialized
Group Care
  10.7%
                                                                  Other
                                                                  7.4%                                                                      Shelter Care
                                                                                   Transitional                                                1.0%
                                                                                    Program                   Independent
         Family Foster Care                                                           2.3%                       Living
               39.3%                                                                                              0.9%




William M. Mercer, Incorporated                                          84                     Colorado Child Welfare Evaluation
                                                                                       Second Interim Implementation Status Report
                                            Figure 35
                               Out-of-Home Placement Expenditure
                                    Boulder County, SFY 1999
                                          RTC                     Home Based Program
                     RCCF                 9.0%                          2.1%
                     24.3%                          Relative Foster
                                     Excludes
                                                                                        Receiving
                                                         Care                            Home
                                       RTC
                                     Medicaid            5.4%                             5.1%


       Specialized                                     Other
       Group Care                                      14.3%
         10.1%
                                                                                          Shelter Care
                                                                                             2.4%
                                                        Transitional
                                                         Program          Independent
                Family Foster Care                         2.4%              Living
                      37.0%                                                   2.1%




                                           Figure 36
                              Out-of-Home Placement Expenditure
                                   Boulder County, SFY 2000

                     RCCF              RTC                        Transitional     Home Based
                     10.0%            15.3%                         Program         Program
                                                                     3.0%             1.4%
                                                Relative Foster Care
                               Excludes
                                                        6.2%
     Specialized                 RTC
     Group Care                Medicaid                                                      Receiving
        9.9%                                                                                  Home
                                                                                               2.9%

                                                      Other
                                                      12.2%




                                                                Independent         Shelter Care
       Family Foster Care                                          Living              2.6%
             46.3%                                                  2.4%




William M. Mercer, Incorporated                        85                Colorado Child Welfare Evaluation
                                                                Second Interim Implementation Status Report
Figures 37 and 38 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 37 shows that Boulder County
directly operates most of its family foster care and group homes. Figure 38 shows that the
direct operation of group homes has increased over time, while there has been increased
contracting of family foster care to CPAs.
                                         Figure 37
              Percentages and Amounts of Expenditure for Family Foster Care and
                      Group Home Services delivered by County vs. CPA
                            Boulder County, SFY 1998 to SFY 2000
      100%
       90%
                       $320,163                       $278,176                          $478,887
       80%
       70%
       60%
       50%
       40%
                       $619,088                       $617,089                          $828,634
       30%
       20%
       10%
        0%
                        SFY1998                       SFY1999                           SFY2000

    County Family Foster Care Homes and Group Homes         CPA Family Foster Care Homes and Group Homes




                                          Figure 38
              Percentages of Expenditure for Family Foster Care and Group Home
                          Services delivered by County and/or CPA
                            Boulder County, SFY 1998 to SFY 2000
       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                      SFY1999                           SFY2000


  CPA Family Foster Care Homes     CPA Group Homes    County Family Foster Care Homes      County Group Homes




William M. Mercer, Incorporated                        86                 Colorado Child Welfare Evaluation
                                                                 Second Interim Implementation Status Report
El Paso County Findings
El Paso County is located in the central part of Colorado, 60 miles south of Denver. It had an
estimated child population of 138,382 in 2000. El Paso County was accepted as a managed care
county in SFY 1999.




William M. Mercer, Incorporated               87               Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Process Findings
Process findings first examine how El Paso County is functioning as a managed care entity.
Blueprints are then presented to show the flow of children and their families, and the flow of
funds through El Paso’s child welfare system. Finally, El Paso is scored on the extent of its
fidelity to managed care principles using Fidelity of Organizational Implementation Scales.

Progress in the Introduction of Managed Care Processes
El Paso County is functioning as a managed care entity as it is assuming risk by entering into an
MOU with the CDHS. In its original MOU, El Paso County was permitted to keep unspent funds
but was required to cover overages that might occur. It is expected to perform at a certain level
on specific outcomes.

In 2001, the CDHS offered counties a choice between keeping any of their own unspent funds or
being eligible for participation in the distribution of unspent funds from other counties at the end
of the year. El Paso County chose to give up the ability to retain its own unspent funds in order
to participate in surplus distributions from the state.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                   88               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
El Paso County is unique in that it has not developed a formal utilization management process as
part of its managed care initiative. Instead, it has focused on the development of its provider
network, the introduction of performance indicators, and the use of financial incentives.

In developing its network, El Paso County, like all other managed care counties, has not
determined the capacity that it requires. As a result, it is limited in developing a configuration of
services that addresses the needs of children and families now and in the future. Financial
incentives for the development of services that represent present best practices are difficult to
implement without knowing the overall capacity that is required.

Although El Paso County has not determined its required network capacity, it and Arapahoe are
the only two counties that contract with formal provider networks. El Paso uses two vendors to
manage all CORE services. United Health’s EPICS Network provides services under a risk
corridor arrangement to approximately 80 percent of cases, replacing the hundreds of separate
providers who formerly provided such services. Savio House Home-Based Intervention serves
the 20 percent of cases with the highest level of need through a case rate arrangement, primarily
through intensive community alternatives such as family preservation and day treatment services.

El Paso has also filled some obvious gaps in its network. For example, a partnership among
El Paso County, youth corrections, probation, and the mental health center has created a team
approach to provide multi-systemic therapy to high-risk youth and their families. Of special note
is that El Paso has received a grant for integrating child welfare and domestic violence
interventions, a key area of need.

El Paso County has introduced financial incentives coupled with performance indicators, most
notably in the area of adoptions. El Paso has increased the adoption subsidy to the level of the
foster care subsidy and then used bonuses for adoptions to increase adoptions to the point where
the county is now the leader among both managed care and non-managed care counties in this
evaluation. Of interest is that this has had a negative impact on El Paso’s finances as adoption
subsidies last much longer than foster care subsidies and costs are compounded with each new
year of subsidies incurred.

El Paso County also has risk sharing arrangements with its CPA Project (the MHASA is fully at
risk for treatment service costs and the county is at risk for room and board costs), EPICS
Network (through a risk corridor that splits unspent funds 40/60 and cost overages 50/50), and
Savio House (through a case rate). Another example of the use of performance indicators is
Project REDIRECT, which focuses on the 100 most resource-intensive children in the system. It
has developed outcome criteria and has collected data for several years on these children.

In general, providers of service are required to obtain measurable outcomes and collect and
report statistical, fiscal, or performance data. If a contractor fails to attain outcomes, it can
constitute a failure to perform.




William M. Mercer, Incorporated                   89               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
In regard to the other key elements of managed care listed above, El Paso County is still in the
development stage. As mentioned earlier, it does not have a distinct utilization management
program. Instead, it relies on individual caseworkers and their supervisors to manage utilization.
The impact of such an approach on utilization, compared to other approaches, might be a focus
of study as El Paso continues to develop its managed care program.

In other areas, El Paso leadership has articulated a vision for quality improvement and included a
focus on using data to improve performance over time. El Paso is the only county with a written
quality improvement plan.

El Paso County needs to improve its information system. While El Paso tracks its own costs for
its purchased services, it take specialized, labor-intensive queries to use those data. As a result,
El Paso County has relied on the CDHS reporting to track provider expenditures and manage
utilization.

El Paso encountered a major problem in SFY 2000 when the CDHS reported a large deficit that
had gone unreported the quarter before; this delay made it difficult for El Paso to manage the
deficit. The importance of timely and accurate financial reporting has been a major lesson cited
by El Paso County in the last year. In response to the need for more reliable and timely financial
reporting information, El Paso County has used vacancies to create a lead financial position in
the county. The CDHS has recently implemented a new system, TRAILS, which has the
potential to yield more useful and timely information.

Process Flow
To illustrate the flow of processes involving people and funds in El Paso County, we have
constructed eight Blueprints. These show:
1. the flow of persons through the county in SFY 1997, prior to the initiation of managed care
   reforms;
2. the overall flow of persons through the county in SFY 2001, midway through the fourth year
   of managed care reforms;
3. additional detail on the flow of persons through the county department;
4. additional detail on the flow of persons through the county’s provider network;
5. the flow of funds through the county in SFY 1997, prior to the initiation of managed care
   reforms;
6. the overall flow of funds through the county in SFY 2001, midway through the fourth year of
   managed care reforms;
7. additional detail on the flow of funds from the state to the county, including financial
   reporting; and
8. additional detail on the flow of funds through the county’s provider network, including
   risk-sharing arrangements;

We have used six colors in the Blueprints:

William M. Mercer, Incorporated                  90               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
1. black is used for the organizations involved in paying for, managing, and providing child
   welfare services;
2. red is used for the unmanaged flow of people through the system and includes county
   populations, populations in need, referral groups, movement through different teams and
   areas within the county child welfare department, and referrals to community providers;
3. green depicts child welfare funds that flow directly to the county child welfare department, as
   well as the state and federal funds that flow through the CDHS to the county child welfare
   department;
4. blue depicts the flow of non-child welfare funds that are critical to providing needed care to
   recipients of child welfare services;
5. purple depicts the managed flow of both persons and funds through the county child welfare
   managed care pilot initiative; and
6. yellow depicts the court system in each county.

In the Blueprints, the flow of people through the system is depicted in a left to right fashion,
beginning with the entire population potentially served by the child welfare system, narrowing to
the group of persons actually served, and following these people through the organizations and
services they would typically encounter as they pass through the system. The flow of funds,
however, goes from the top to the bottom of each chart. The drawing traces the flow of funds
from the payers that fund the system, to the administrative entities that allocate and account for
these funds, to the county child-serving agencies and providers that ultimately spend and manage
them.

A set of symbols is also used. They are illustrated in the following diagram.

Organizational Symbols:
               = Federal, state or county payors
               = State agencies
               = Local agencies or providers
               = Local entities that bears risk

Risk Symbols:
  S   = Shared risk of financial utilization
  F   = Full risk of financial utilization

Out-of-Home Placement Symbol:
  $            = Most expensive out-of-home
                   placements (RTC, CPA)
Financial Reporting Symbol:
   Financial
  Reporting
  by County    = Locus of financial reporting




William M. Mercer, Incorporated                    91            Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Flow of People—Pre-Managed Care—Spring 1997
The Blueprint depicted on page 94 (entitled “El Paso County—Spring 1997—People Flow”)
shows the flow of people through El Paso County’s child welfare system in the Spring of 1997.
Although El Paso County did not receive designation as a managed care pilot county
until 1998, the county did apply to become a managed care county during the first round in 1997
and initiated its managed care programs despite not being designated as such. Given this, it was
determined that a comparison to 1997 would best fulfill the goal of describing the pre-managed
care system there.

Beginning on the far left of this diagram, the overall county population is broken down into the
smaller group of persons who, at any point in time, were in need of services through the El Paso
DHS. From that group, two sets of referrals emerged. The first set involved two groups typically
seen by the Child Welfare Division of DHS: youth in conflict (including youth with delinquency
problems) and children with reported instances of abuse or neglect. The second set, families in
poverty, comprised the main source of referrals for the county’s TANF program. These two sets
of referrals passed through their own dedicated intake units, one for child welfare and one for
TANF. The intake units screened out cases that clearly did not meet criteria and referred them to
other services in the community. Cases meeting criteria on the child welfare side were referred to
an array of Child Welfare Ongoing Units, including child protection teams, EPP teams, youth
and family services teams, and various initiatives with community partners, including Project
REDIRECT, Metcalf Hall Community Center, and on-site mental health services.

Many children, youth, and families served by the county required additional services provided by
agencies in the community. Workers on ongoing teams referred children to out-of-home
placements in RTCs, RCCFs, CPAs and county administered foster homes through the El Paso
County Placement Team. Workers could directly access outpatient substance abuse services,
specialty mental health providers, and mental health services through the Pike’s Peak
Community Mental Health Center. Children needing inpatient services first had to go through the
CMHC, which managed all indigent and Medicaid-funded mental health benefits. While some
high need or multi-agency involved youth accessed multi-agency planning from time-to-time, the
majority of external referral decisions were not centrally coordinated.

A critical factor for any involuntary case served by the county is the court system. Individual
workers presented recommendations for services to the court and the court implemented them as
it saw fit. DHS leadership conducted ongoing dialogue with the courts to facilitate better
coordination. The El Paso County approach focused on providing the most appropriate, least
intrusive service available. Voluntary cases, including both child welfare and those served
through TANF, were not subject to court involvement. While the Blueprint highlights the role of
the court at the point where it impacted decisions regarding referrals to out-of-home placements
versus other community services, the court’s involvement with involuntary cases was much
broader, beginning at or prior to intake and extending throughout each family’s final disposition.




William M. Mercer, Incorporated                 92               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Another important factor in El Paso County was the management approach of the DHS
leadership. Child welfare and TANF programs were independently managed but management
focused on outcomes tied to clearly specified guiding principles, such as an emphasis on
permanency. Management was also very hands-on.




William M. Mercer, Incorporated             93              Colorado Child Welfare Evaluation
                                                   Second Interim Implementation Status Report
William M. Mercer, Incorporated   94            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Flow of People—Post-Managed Care—Spring 2001
In the fourth year of its implementation of managed care reforms, and the third year following
El Paso County’s designation as a managed care pilot county, the flow of persons through the
El Paso County child welfare system has significantly changed. A review of the Base Blueprint
for Spring 2001 (see the Blueprint depicted on page 96 entitled “El Paso County—Spring
2001—People Flow”) shows the same basic population served but reorganization within El Paso
County DHS and its provider network. To review these changes with greater specificity, two
detail Blueprints were created for this year: a Department Internal Detail Blueprint and a
Provider Network Detail Blueprint.




William M. Mercer, Incorporated               95              Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
William M. Mercer, Incorporated   96            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Department Internal Detail. Looking at the Department Internal Detail Blueprint (see the
Blueprint depicted on page 99 entitled “El Paso County—Spring 2001—Department Internal
Detail”), it can be seen that the county population still breaks down on the left into a smaller
group of persons in need. These, in turn, sort into the three primary groups of persons served by
the county: youth in conflict (including youth with delinquency concerns), children with reports
of abuse or neglect, and families in poverty.

Within the county, the flow of people differs markedly from 1997. To better coordinate the
functions of its separate intake units for child welfare and TANF cases, the county department
uses training and active support so that either unit can facilitate entry into the entire system.
TANF intake workers assess referrals for both basic and child welfare needs, as do child welfare
intake workers, referring people to the most appropriate, least intrusive service. Additionally,
intake workers receive training and supervision to help shift their focus away from a strict
investigation stance to focus more on an assessment of the families’ strengths and needs. Intake
workers are also able to provide prevention and early intervention services, working with
families with lower level needs without opening the case in order to connect them with other
community resources and avoid ongoing child welfare services. Not all families take advantage
of this voluntary service, but those that do are often able to have their needs met without
additional resources.

Families with basic needs are still referred to the county’s TANF program, but there is now
greater coordination between the TANF and child welfare programs under a unified leadership
team. All El Paso County Human Services agency programs (Agency) and services are now
guided by an overall vision to eliminate poverty and family violence. TANF and self-sufficiency
programs serve as the Agency’s principal prevention programs, assisting families to move out of
poverty, and work to prevent the need for more involved child welfare services. Similarly, child
welfare services serve an anti-poverty mission in addition to their primary child safety focus.

All department self-sufficiency programs are viewed as preventative in this fashion, including
TANF, state and county diversion, food stamps, Medicaid, Low-Income Childcare, Teen
Services, TANF Kinship Care, and the Center on Fathering. In addition to department staff
providing services through these programs, on-site community partners play a significant role in
these activities, including employment services through Goodwill Industries of Colorado
Springs, employment and training through One-stop Career Centers and Welfare to Work, child
support through Maximus, non-custodial parent employment and support through The Parent
Opportunity Program, child care resources through Child Care Connections, domestic violence
services through the Center for the Prevention of Domestic Violence, EPSDT services through
the El Paso County Health Department, and mentoring through Faith Partners.

This closer integration and re-focusing TANF services to include an explicit prevention and early
intervention role in support of the child welfare system forms a large part of El Paso County’s
efforts. At the root of this change is a change in county DHS leadership that took the existing
outcome oriented management style and focused it upon better program integration and managed
care principles across the board.




William M. Mercer, Incorporated                 97              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Child welfare partner initiatives expanded as well. The Metcalf Hall Community Center
continued to evolve and was renamed the Lorraine Education and Community Center. Other new
or expanded initiatives included Teen Self-sufficiency, the Placement Team, a Multi-systemic
Therapy team, and the CPA Monitoring Unit.

In 2000–2001, the focus in El Paso County (as in the other managed care counties) has been on
program consolidation and refinement. As has been the case with other counties, staff turnover
has increased related to the increasingly competitive job market for human service employees.




William M. Mercer, Incorporated               98               Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
William M. Mercer, Incorporated   99            Colorado Child Welfare Evaluation
                                       Second Interim Implementation Status Report
Provider Network detail—People Flow. Looking at the Provider Network Detail Blueprint (see
the Blueprint depicted on page 102 entitled “El Paso County—Spring 2001—Provider Network
Detail—People Flow”), it can be seen that all out-of-home referrals go through Placement
Services, which pre-dates the managed care pilot. Placement Services provides centralized
capacity tracking and information on program specialties, and also reviews referrals to see if
alternative out-of-home placements or community-based services would be appropriate.
Placement Services functions in an advisory and consultative role to caseworkers. Placement
Services does not prior authorize care. Differences with caseworkers are worked out
collaboratively through the supervisory structure.

Beginning in 2000–2001, El Paso’s contract with its county attorneys was broadened to support a
24-hour a day, 7-days a week review capacity for all out-of-home referrals by the county
attorneys. This serves a consensus-oriented, gatekeeping role for all out-of-home placements,
ensuring that all out-of-home referrals meet legal sufficiency standards. Caseworkers submit
cases for review to the county attorneys within two days following the determination of a
placement being needed.

The primary managed care project in El Paso County is the CPA Project. The CPA Project is a
partnership between DHS, key CPA providers who share a common vision for the project, and
Pike’s Peak-Options Colorado Health Networks, the entity that manages the Medicaid capitated
mental health contract for El Paso County. The primary goal of the project is to engage the key
CPA providers in the community, train them, and provide them with incentives to better support
child welfare goals, including the conduct of specific case management duties. This program has
the goal to become the primary certified foster care resource for the county and no new county
foster placements have been pursued since it was started (current placements continue to be
supported, however, to prevent disruption to youth already under care).

In addition to the CPA Project, multi-agency cases can receive multi-agency treatment planning
through various multi-agency and community review forums. These forums include both child-
specific and system coordination level multi-agency coordination. The forums include: the
Placement Team, the Multi-Agency Review Team, administrative reviews, Community Child
Protection Team, early release meetings, and permanency reviews. DHS, mental health, DYC,
substance abuse services, and probation representatives participate on these teams.

Through specific initiatives related to permanency, there has been an increase in the number of
families from the community coming forward to adopt children through DHS. Support of kin
placements also has increased through the Kinship Foster Care program and other supports.
Kinship homes in the community conform with the same licensing requirements as foster homes.




William M. Mercer, Incorporated               100              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
In Spring 2000, El Paso County continued to evolve its system by restructuring its CORE
services program. It let an RFP in the Fall of 1999 and selected two vendors to co-manage all
CORE services. The CORE Services System of Care was now co-managed as follows:

§     United Health’s EPICS Network of the 17 top providers in the area provides the full array of
    CORE services to approximately 80 percent of cases. This single, coordinated network
    replaced the hundreds of separate providers who formerly provided such services.
§     Savio House HBI serves the 20 percent of cases with the highest level of needs, primarily
    through intensive community alternatives such as family preservation and day treatment
    services. The Intake Unit, Ongoing Teams, and CPAs directly refer to this coordinated
    network.




William M. Mercer, Incorporated                 101               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   102            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Pre-Managed Care—Spring 1997

The Blueprint depicted on page 104 (entitled “El Paso County—Spring 1997—Funding Flow”)
shows the flow of funds through the El Paso County child welfare system prior to the initiation
of their managed care reforms in late 1997. In the Spring of 1997, funds from various federal
sources (Title IV-E and XX) and the State General Funds were combined by the CDHS and
passed on to El Paso County. The bulk of these funds flowed through separate funding streams
which had to be separately tracked and accounted for by the county. Furthermore, unspent funds
in one category could not be used in other categories, creating an incentive to keep spending in
line with historical patterns. There were eight separate state funding streams, including:
§    two streams funding county administration (to fund county staffing, primarily), one
    comprised of 100 percent state dollars and one of 80 percent state/20 percent county funds;
§    a large stream of funds for out-of-home placements (OOHP);
§    a significant funding stream for CORE services, originally developed to fund alternatives to
    out-of-home care such as family preservation, outpatient substance abuse and mental health
    services, and day treatment; and
§    four smaller funding lines (e.g., Special Circumstances Child Care, Subsidized Adoption,
    CHRP, and Case Services).

El Paso County also received other specialized state funding, including Expedited Permanency
Planning funds, but these other sources of funds are not depicted.

In addition to the budgeted funding streams, the Colorado DHS, and through them, the State
General Fund, has the full risk for any cost overages that exceeded the state’s ability to shift
unspent funds from other areas of the state. In addition, no county was able to keep any unspent
funds. As a result, counties had little incentive to limit their spending of child welfare funds and
many spent their entire allocation and more.

El Paso County, in turn, contracted with various community agencies for services. In addition to
funding streams flowing through DHS, El Paso County’s child welfare system also depended on
services paid for by other agency funding. These included:
§    Medicaid (Title XIX) funding paid through the Colorado Department of HCPF to RTC
    providers for the treatment component of the services provided in these facilities—The county
    DHS directly pays for only room and board. Treatment costs are paid directly from the State
    Department of Human Services to HCPF but are then charged back to the county at year end
    against its allocation. The county has no control over the rates paid to providers since they are
    determined through a cost-based approach within available appropriations.




William M. Mercer, Incorporated                  103               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
§     Medicaid (Title XIX) funding for mental health services transferred through the Department
    of HCPF to the Mental Health arm of the DHS Office of Health and Rehabilitation Services—
    In El Paso County, funding for all Medicaid inpatient and outpatient mental health services has
    been capitated and, as a result, flows through the entity awarded the MHASA contract for that
    area of the state, Pike’s Peak-Options Colorado Health Networks.
§    DYC SB 94 funding for preventing detention placements for which some youth served by DSS
    are eligible—Services funded may be similar to CORE services.
§     ADAD funding through the DHS Office of Health and Rehabilitation Services—In El Paso
    County, these funds are managed by Signal, the managed services organization for substance
    abuse services. Signal pays for additional substance abuse services beyond those paid by DSS
    through its CORE funds.

As can be seen in the Blueprint, child welfare funds were the least managed of the primary
funding streams for expensive out-of-home placements, such as inpatient, RTC, and CPA. In
1997, inpatient and other out-of-home placements through mental health were subject to
managed care controls, and RTC placements through DYC were subject to the placement
restrictions of the juvenile justice system.




William M. Mercer, Incorporated                 104               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   105            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Post-Managed Care—Spring 2001
The Blueprint depicted on page 107 (entitled “El Paso County—Spring 2001—Funding Flow”)
shows the flow of funds under the managed care pilot as they exist in the Spring of 2001. The
primary state and federal funding sources flowing through the CDHS remains essentially
unchanged with one exception. Now state and federal Title IV-E funds supporting TANF
services are included in the description of child welfare funding sources. All Colorado counties
receive these funds, but El Paso County has directly aligned its Child Welfare and TANF
programs so they are depicted here as another source of child welfare funding support.

The flow of funds from the CDHS has changed dramatically. In addition to creating the managed
care pilots, SB 97-218 collapsed seven of the categorical funding streams that existed in the
Spring of 1997 into a single child welfare block. Only CORE services remained as a separate
primary funding stream, with separate accountability and accounting to prevent funds from being
diverted to other uses. This block funding approach was implemented for all counties, not just
managed care counties.

All Colorado counties, including El Paso, are now responsible for any overages from budgeted
expenditures, shifting full downside risk from the state CDHS and General Fund to El Paso DHS
and the county, as designated by the risk triangle in the DHS box. In addition, as a managed care
pilot, El Paso is able to retain any unspent funds from the seven combined funding streams, thus
taking on the upside risk for the services it provides. To do this, El Paso has to allow
non-managed care counties to use unspent funds from other areas of the state to cover their
overages before being eligible for these. The county is still eligible for excess Title IV-E
distributions it helps earn.

In the first two years of operating under a capped allocation, the CDHS returned unspent general
funds to the legislature. During that period of time, as well as during the third year, funding for
the child welfare block was based on “population increase.”

During the third year, SFY 2000, all Colorado counties collectively overspent their budgets by
approximately $20 million. The CDHS requested and received a $14 million supplemental
appropriation in SFY 2001, based on projected caseload growth. Because counties overexpended
so significantly in SFY 2000, the Child Welfare Allocation Committee approved allowing
managed care counties a choice to either participate in the statewide surplus distribution or retain
their own unspent funds.

El Paso County has chosen the MOU option where they give up the ability to retain unspent
funds and, instead, are able to participate in surplus distributions from the state. This can be seen
in both the Base Blueprint and the State Funding Detail Blueprint for the county (also see the
Blueprint depicted on page 110 entitled “El Paso County—Spring 2001—State Funding Detail”).




William M. Mercer, Incorporated                 106               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
El Paso County had a challenging year in terms of funding in SFY 2000, experiencing the largest
deficit among the managed care counties. Several of the factors accounting for this deficit are
related to El Paso County’s pilot. Despite decreasing numbers of youths in foster care and RTC
placements, spending in other areas has increased. Subsidized adoption spending, CORE services
spending, and increased attorney costs (largely to fund the 24/7 placement review process
described earlier) account for the bulk of the increase in costs. On the surface, El Paso County’s
move toward more community-based programming seems to be costing it more.




William M. Mercer, Incorporated                107              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
William M. Mercer, Incorporated   108            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
State Funding Detail. Looking at the State Funding Detail Blueprint (see the Blueprint depicted
on page 110 entitled “El Paso county—Spring 2001—State Funding Detail”), another systemic
issue related to the cost overruns can be seen. This is the areas of financial reporting. All
counties track their internal costs for staff and overhead. However, unlike many other counties,
El Paso tracks its own costs for its purchased services, although it takes specialized,
labor-intensive queries to use those data. The primary difficulty is that the financial reporting by
the CDHS and reporting by El Paso County use different systems that are not readily
reconcilable. In addition, because of the difficulty aggregating its service data, El Paso County
has relied on the CDHS reporting to track provider expenditures and manage utilization.

The transition in the CDHS electronic provider payment and financial reporting system in
SFY 2000 kept El Paso County from having accurate financial information from the CDHS until
after the close of the fiscal year. At the second to last month of the fiscal year, CDHS reports led
El Paso County to believe that it was within one percent of its budget. However, the final
close-out reports showed it to be over-expended by over $3 million.

In response to the need for more reliable and timely financial reporting information, El Paso
County has used vacancies to create a lead financial position in the county. This new position
will work to help the county develop a financial reporting system that is reconcilable with the
CDHS system and more timely in its availability than the CDHS reports. The importance of
timely and accurate financial reporting has been a major lesson cited by El Paso County in the
last year. Fortunately, strong support from the county leadership and a strong financial reserve
has allowed El Paso time to make these changes without major crisis.




William M. Mercer, Incorporated                 109               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   110            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Provider Network Detail—Funding Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 112 entitled “El Paso County—Spring 2001—Provider
Network Detail—Funding Flow”), it can be seen that, through the CPA Project, El Paso DHS
also now earns a federal Medicaid (Title XIX) match on the treatment portion of its CPA costs.
This was accomplished by implementing a CPA Medicaid transfer that sent these funds through
the State Department of HCPF to the Pike’s Peak-Options CHN and, thereby, to the CPA
providers. Part of the increased revenue realized through this change is used to increase the rates
paid to CPAs to cover the additional costs of requiring them to perform case work for the youth
they serve, as well as special therapeutic supports (e.g., respite, specialized services) and
increased administrative and training costs.

Financial risk for the CPA Project is shared between El Paso County DHS and Pike’s
Peak-Options CHN. DHS is fully responsible for the room and board rates for all CPA
placements. Pike’s Peak-Options CHN is responsible for all treatment costs incurred.

Subsidized adoption payments have also been increased to match the rate paid for CPA certified
foster care. In addition, the CPA Project offers an incentive to CPA providers whose foster
homes end up adopting their placed youth. This incentive helps defray the additional recruitment
costs that CPA providers incur when a foster parent adopts their foster child. Incentives are
higher for more timely adoptions.

The county’s CORE Services System of Care simplified the funding stream for these services.
The two provider groups co-managing this system also share in the financial risk for these
services. EPICS has entered into a risk corridor arrangement in which unspent funds are split
60/40, with more going to EPICS as an additional incentive. Downside risk is split 50/50,
creating an equal shared risk situation for cost overruns. Savio House HBI receives a fixed per
case rate and provides all needed care within that rate. Savio House HBI is responsible for any
overage and benefits from any unspent funds.




William M. Mercer, Incorporated                 111              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   112            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. When counties implement managed care in a way that is consistent with
such expectations, we call that “process fidelity.”

Process fidelity is scored in each of four domains: funding flexibility, interagency coordination,
integration of informal networks and non-traditional services, and implementation of early
intervention and prevention programs (on a four-point scale, from one to 4). Once a score is
calculated, it is placed within one of four fidelity categories – early, emerging, maturing, and
mature stages of fidelity.

Funding Flexibility
El Paso was rated in the Mature Stage of fidelity in this domain. They are using various types of
financing approaches other than fee-for-service and they have incentives built into some of their
contracts with providers. El Paso has an arrangement with CPAs in which they transfer some
administrative responsibility and financial risk to the CPAs. They also use a risk-corridor with a
capped allocation and case rate financing for CORE services.

Interagency Coordination
In this domain, El Paso County was given a rating of Mature Stage fidelity. El Paso has
developed formal, contractual relationships with numerous child-serving agencies and they have
developed service planning processes and administrative coordination processes with multiple
child-serving agencies. El Paso has forged extensive and creative internal relationships between
TANF and child welfare programs to leverage dollars for prevention and other services.

Integration of Informal Networks and Non-Traditional Services
El Paso County’s fidelity score placed it in the Maturing Stage of fidelity in this domain. El Paso
has forged relationships with many organizations outside of the formal child-serving system and
has engaged these organizations in many different ways. An area in which El Paso County
officials did not report as much integration of informal networks was funding—they did not
report much community involvement in funding for programs.

Implementation of Early Intervention and Prevention Programs
El Paso County was found to be performing at the Mature Stage of fidelity in this domain. The
County reported delivering a prevention program in conjunction with TANF and that they are
beginning to implement an early intervention/prevention program through a nurse visitation
program, based on the evidence-based approach developed by David Olds.




William M. Mercer, Incorporated                 113              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in El Paso County in the
nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in El Paso to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.

Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                        Figure 39
           Percentage of Children Who Experienced a Confirmed Recurrence of
                Abuse and/or Neglect within a 12-Month Period after being
                                Abused and/or Neglected

    10
     9
     8
     7
     6
     5                                                4.2
     4                             2.8                                    2.7
     3          2.2
     2
     1
     0
              SFY 97              SFY 98             SFY 99             SFY 00




William M. Mercer, Incorporated               114               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect

                                      Figure 40
        Confirmed Abuse and Neglect per 1,000 Children in the County Population


       20
       18
       16
       14
       12
       10
        8
        6                          4.3             4.7              4.5
        4         2.8
        2
        0
                SFY 97            SFY 98          SFY 99           SFY 00

Indicator 3: Referrals and Investigations

                                          Figure 41
                 Referrals of Alleged Abuse and Neglect per 1,000 Children
                                  in the County Population

      70
                                                                    59
      60                           54.4           52.7
                 50.6
      50
      40
      30
      20
      10
       0
                FFY 97            FFY 98          FFY 99          FFY 00




William M. Mercer, Incorporated             115               Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
                                              Figure 42
                       Investigations of Alleged Abuse and Neglect per 1,000
                                Children in the County Population

              50

              40                                                          35
                                                         29.3
              30           27            25.9

              20

              10

                0
                         FFY 97        FFY 98           FFY 99          FFY 00

Indicator 4: Discharge to Reunification
It should be noted that this indicator does not directly measure reunification in that is represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.

                                            Figure 43
                    Percentage of Children placed with Parents at Case Closure

            100
             90
             80
             70                          63.4
                                                         57.5
             60           53.2
             50                                                          37.1
             40
             30
             20
             10
              0
                        SFY 97         SFY 98           SFY 99          SFY 00



William M. Mercer, Incorporated                  116              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 5: Children in Out-of-Home Placement

                                         Figure 44
                Children in Out-of-Home Placement at some Point during the
                  Fiscal Year per 1,000 Children in the County Population

         20
         18
         16
         14
         12
         10        7.7                                    7.7
                                            6.7                          7.2
          8                        6.3
          6
          4
          2
          0
                 SFY 96           SFY 97   SFY 98       SFY 99         SFY 00

Indicator 6: Length of Out-of-Home Stay

                                       Figure 45
             Out-of-Home Placement Length of Stay (in Months) for Closed Cases

        12
                                            9.3           9.4             9.6
        10

         8                         7.5

         6        4.9

         4

         2

         0
                 SFY 96           SFY 97   SFY 98        SFY 99         SFY 00




William M. Mercer, Incorporated            117               Colorado Child Welfare Evaluation
                                                    Second Interim Implementation Status Report
Indicator 7: Percentage in Positive Placements

                                           Figure 46
                       Percentage of Positive Placements at Case Closure
                        (Parents, Kin, Relatives, Adoption, Guardians)

        100                                            86.4             88.5
         90          79.7             83.7
         80
         70
         60
         50
         40
         30
         20
         10
          0
                  SFY 97             SFY 98           SFY 99            SFY 00

Indicator 8: Number of Adoptions

                                          Figure 47
                     Adoptions per 1,000 Children in the County Population

               1.8
               1.6                             1.45
               1.4                                               1.22
                             1.17
               1.2
                 1
               0.8
               0.6
               0.4
               0.2
                 0
                            FFY 98            FFY 99            FFY 00




William M. Mercer, Incorporated                118               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Indicator 9: Time to Adoption

                                      Figure 48
        Average Months from Termination of Parental Rights to Finalized Adoption

          20


          15
                         11                11
                                                                 9.2
          10


           5


           0
                      FFY 98              FFY 99                FFY 00




William M. Mercer, Incorporated           119             Colorado Child Welfare Evaluation
                                                 Second Interim Implementation Status Report
           Outcome Comparisons

                                            Table 7
 El Paso County SFY 2000 Outcome Results Compared to Other Counties, to Benchmarks, and to
                      the Criterion Reference Standards Set for SFY 2000
                                                          Non-
                                              Other
                                                        Managed
                              El Paso        Managed                                             Criterion Reference
  Outcome Indicator                                       Care            Benchmark
                              County           Care                                                   Standard
                                                         Large
                                             counties
                                                        counties
                                                                    National: 11%
Indicator 1: Confirmed                                                                         Acceptable: <10%
re-abuse
                              2.7%            4.0%       3.0%       Comparison States: 12%
                                                                                               Outstanding: <2%
                                                                    (Utah)
Indicator 2: Confirmed                                                                         Acceptable: <10
                                                                    National: 11.8
abuse per 1,000 child          4.5            5.2         5.1       Comparison States: 8.8     Outstanding: <4
population
Indicator 3a: Referrals of                                                                     Acceptable: <45
                                                                    National: 72.1
suspected abuse per 1,000       59            37.9        48        Comparison States: 53.4    Outstanding: <40
child population
Indicator 3b:                                                                                  Acceptable: [not set]
Investigations of suspected                                         National: 43.5             Outstanding: [not set]
abuse per 1,000 child
                                35            20.8        31        Comparison States: 28.3
population
Indicator 4: Percent                                                National: 59%*
discharged to reunification
                              37.1%           48%       57.7%       Comparison States: 68%*
                                                                                               Acceptable: >52%

Indicator 5:
Children in out-of-home                                             National: 8.1**            Acceptable: <8
care per 1,000 child
                               7.2            7.6        10.6       Comparison States: 5.6**   Outstanding: <4
population
Indicator 6:                                                                                   Acceptable: <10
                                                                    National: 10.8
LOS in out-of-home             9.6            7.8         6.2       Comparison States: 12.3    Outstanding: <7
placement (in mos.)
Indicator 7:                                                                                   Acceptable: >82%
                                                                    National: 83%
Percent discharged to         88.5%           86%       83.8%       Comparison States: 90%     Outstanding: >90%
positive placements
Indicator 8: Adoptions                                              National: .66              Acceptable: >.4
per 1,000 child population
                               1.22           .43         .66       Comparison States: .52     Outstanding: >.9
Indicator 9:                                                                                   Acceptable: <18 months
                                                                    National: N/A
Time from termination of       9.2            15.1       20.6       Comparison States: N/A     Outstanding: <9 months
parental rights to adoption
* Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
**Benchmarks are for point in time (on September 30) versus at any time in fiscal year (Indicator 5).




           William M. Mercer, Incorporated                120               Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Summary of Findings
El Paso County’s rates of abuse and neglect were low compared to national and comparison state
benchmarks. They were low even in comparison to other large counties in Colorado. In
particular, El Paso County’s rate of recurring abuse and neglect was quite low (2.7 percent). At
the same time, El Paso’s rates of referrals and investigations of abuse and neglect are high
compared to other large counties in Colorado and, in particular, compared to other managed care
counties. It is not clear whether El Paso’s investigations of abuse and neglect apply standards for
confirming abuse and neglect that are more strict than other counties, or if El Paso is more
vigilant with regard to reporting and investigating suspected abuse.

El Paso has evidenced an increase from SFY 1997 to SFY 2000 in reports, investigations, and
confirmations of abuse and neglect per 1,000 children in the population. It is not yet clear
whether there is an actual increase in abuse and neglect in the community or whether El Paso
County has become more vigilant with respect to reporting, investigating, and confirming abuse
and neglect in the county.

El Paso County had one of the lowest rates of children living with parents at case closure
(27.1 percent). At the same time, the percentage of positive placements at case closure (88.5
percent) was one of the highest among large counties in Colorado. El Paso has placed great
emphasis on permanency, particularly on incentives for adoption. It is possible that El Paso’s
relatively low rate of placement with parents is related to its outstanding performance in the area
of adoption. Indeed, the trends in positive placements over time show that rates of placements
with parents at case closure have steadily decreased, while rates of positive placements overall
have steadily increased in El Paso County. Given El Paso County’s outlier status in both high
rates of positive discharges and low rates of children returning to their parents, this pattern will
continue to be examined as the evaluation progresses.

El Paso County’s rate of out-of-home placement was somewhat low relative to other counties but
its average length of stay in out-of-home placement was relatively high compared to other large
counties in Colorado. As with Jefferson County, it may be that the lower the rate of placement,
the more difficult the cases tend to be and, generally, the longer the lengths of stay in out-of-
home placement. An examination of trends in out-of-home placements in El Paso County,
however, indicates that rates of out-of-home placement have remained relatively stable over
time, while average lengths of stay increased precipitously prior to the implementation of
managed care, and then leveled off once managed care was implemented.

Finally, with regard to adoptions, El Paso County continues to stand out in both its rate of
adoptions, which is much higher than any of the other counties, and in its short length of time
between termination of parental rights and adoption.




William M. Mercer, Incorporated                 121               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as residential treatment centers, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher cost care, as well as trends over several years. The results of these
analyses are shown in this section for El Paso County by itself and in a later section for all of the
managed care counties together.

Figures 49 and 50 compare El Paso County’s allocation at the beginning of SFY 2000 to its
expenditures at the end of the year. Figure 49 provides a comparison of the two main child
welfare funding streams, CORE services and the child welfare block. Figure 50 provides detail
regarding the funding streams that compose the child welfare block.

Figure 49 shows that El Paso County experienced a large deficit in SFY 2000 in both CORE
services and the child welfare block. Figure 50 shows that El Paso’s deficit did not result from
the overuse of RTC and CHRP, the most costly services in its array, but from an overuse of less
costly services contained in its 80/20 allocation. It appears that El Paso County was unable to
operate within its allocation in SFY 2000.

                                             Figure 49
                               CORE Services and Child Welfare Block
                       Allocation vs. Expenditure*, El Paso County, SFY 2000


        $40,000,000
        $35,000,000                                                   $30,542,273 $31,538,697
        $30,000,000
        $25,000,000
        $20,000,000
        $15,000,000
        $10,000,000    $4,342,543 $6,052,659
         $5,000,000
               $-
                                                                                                     $(996,423)
                                                 $(1,710,116)




        $(5,000,000)             CORE Services                                 Child Welfare Block
       $(10,000,000)



          * Before Year End Close Out Process
                                                                      Allocation      Expenditure                 Difference




William M. Mercer, Incorporated                                 122               Colorado Child Welfare Evaluation
                                                                         Second Interim Implementation Status Report
                                               Figure 50
                                 Child Welfare Block Allocation versus
                        Expenditure before Close Out, El Paso County, SFY 2000




                                         $31,538,697
                           $30,542,273




                                                                         $25,100,108
        $40,000,000




                                                           $20,643,141
        $35,000,000
        $30,000,000
        $25,000,000




                                                                                         $6,332,717
        $20,000,000




                                                                                                         $3,206,825




                                                                                                                                                   $2,328,240

                                                                                                                                                                $2,328,240
                                                                                                                      $1,238,175
        $15,000,000




                                                                                                                                   $903,525
        $10,000,000
         $5,000,000
                 $0
                       Total CW Block                     Total 80/20                  RTC Medicaid CHRP Medicaid                               100% Admin.
                                                       Allocation                          Expenditure


Figures 51 and 52 provide a picture of how El Paso County spent its child welfare block in
FY 2000. Figure 51 shows the percentage of funds expended by budget item, while Figure 52
shows the actual amount expended. Of interest is that El Paso spends a smaller percentage of its
budget on RTC than other managed care counties, while taking the lead in the proportion of
expenditures for subsidized adoption, reflecting its emphasis on adoption.

                                          Figure 51
              Percentages of Child Welfare Block Expenditure* by Budget Items,
                                  El Paso County, SFY 2000



                                                                                                                                              Out of Home Placement
                                                                                                                                              Services
                      100% Admin.                                                                                                             35%
                      7%
                CHRP Medicaid                                                                                                                 Subsidized Adoptions
                                                                                                      Other
                3%                                                                                                                            17%
                                                                                                      80%
                   RTC Medicaid                                                                                                               Case Services
                   10%                                                                                                                        1%
                                                                                                                                              Special Circumstances
                                                                                                                                              Child Care (SCCC)
                                                                                                                                              2%
                                                                                                                                              80/20 Admin.
                                                                                                                                              25%
              * Includes Medicaid Fund and before Year End Close Out Process




William M. Mercer, Incorporated                                                        123                             Colorado Child Welfare Evaluation
                                                                                                              Second Interim Implementation Status Report
                                             Figure 52
                         Child Welfare Block Expenditure* by Budget Items,
                                     El Paso County, SFY 2000


    $12,000,000                                                $11,120,315

    $10,000,000
                                                                                                                            $7,821,118
     $8,000,000

     $6,000,000                                                               $5,300,249

     $4,000,000      $3,206,825
                                                $2,328,240
     $2,000,000                     $903,525
                                                                                            $358,974        $499,451
                $0
                     RTC Medicaid     CHRP       100% Admin.    Out of Home    Subsidized   Case Services        Special    80/20 Admin.
                                     Medicaid                    Placement     Adoptions                    Circumstances
                                                                  Services                                    Child Care
    * Before Year End Close Out Process                                                                         (SCCC)




Figures 53–55 illustrate the trends in out-of-home expenditures in El Paso County from SFY
1998 through SFY 2000. El Paso and Pueblo share the distinction of having a much larger
proportion of their out-of-home expenditures used for family foster care than other managed care
counties. Of note, is that El Paso has decreased the proportion of its out-of-home expenditures
devoted to RTC while increasing other services.

                                           Figure 53
                      Percentages of Out-of-Home Placement Expenditure,
                                  El Paso County, SFY 2000
                                                                              Specialized Group
                                                                                 Care 7.8%                    RCCF
                                                                                                              3.1%

                                                                                                        RTC
                                                                                                        8.5%

                                                                                                   Relative Foster Care
                                                                                                           3.0%

                                                                                                  All Other *
                                      Family Foster Care                                             0.6%
                                            77.1%

                                                                               * Includes independent living, transitional
                                                                               program, home-based program, receiving
                                                                               home, and shelter care.




William M. Mercer, Incorporated                                  124                   Colorado Child Welfare Evaluation
                                                                              Second Interim Implementation Status Report
                                         Figure 54
                     Percentages of Out-of-Home Placement Expenditure,
                                 El Paso County, SFY 1999
                                   Specialized Group Care
                                            10.3%                        RCCF
                                                                         3.5%

                                                                           RTC
                                                                           5.3%
                                                Excludes RTC
                                                    Medicaid                Relative Foster
                                                                             Care 4.4%

                                                                     All Other *
                       Family Foster Care                               0.1%
                             76.3%
                                                                           * Includes independent living, transitional program, home-
                                                                           based program, receiving home, and shelter care.




                                     Figure 55
                 Percentages of Out-of-Home Placement Expenditure,
                             El Paso County, SFY 2000
                                                    Specialized Group Care
                                                             9.9%
                                                                                RCCF
                                                                                4.0%



                                                                              RTC
                                                      Excludes RTC           5.6%
                                                        Medicaid                    Relative Foster Care
                                                                                            5.4%



                                                                                   All Other *
                               Family Foster Care                                     0.2%
                                     75.0%
                                                                           * Includes independent living, transitional
                                                                           program, home-based program, receiving
                                                                           home, and shelter care.




William M. Mercer, Incorporated                          125                    Colorado Child Welfare Evaluation
                                                                       Second Interim Implementation Status Report
Figures 56 and 57 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 56 shows that El Paso County has
contracted out most of its family foster care and group homes to CPAs. Figure 57 shows
that family foster care homes are the primary contracted service.
                                         Figure 56
              Percentages and Amounts of Expenditure for Family Foster Care and
                      Group Home Services delivered by County vs. CPA,
                            El Paso County, SFY 1998 to SFY 2000
       100%
        90%
        80%
        70%
        60%                                           $8,062,493
                       $8,682,333                                                   $8,746,622
        50%
        40%
        30%
        20%
        10%            $1,096,982                     $1,232,612                    $1,127,393
         0%
                         SFY1998                       SFY1999                        SFY2000

   County Family Foster Care Homes and Group Homes        CPA Family Foster Care Homes and Group Homes




                                          Figure 57
              Percentages of Expenditure for Family Foster Care and Group Home
                          Services delivered by County and/or CPA,
                            El Paso County, SFY 1998 to SFY 2000

       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                       SFY1999                        SFY2000

   CPA Family Foster Care Homes     CPA Group Homes        County Family Foster Care Homes       County Group Homes




William M. Mercer, Incorporated                        126                  Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Jefferson County Findings

Jefferson County is located adjacent to and west of Denver. It had an estimated child population
of 127,888 in 2000. Jefferson County was accepted as a managed care county in SFY 1998.




William M. Mercer, Incorporated               127               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Process Findings
Process findings first examine how Jefferson County is functioning as a managed care entity.
Blueprints are then presented to show the flow of children and their families and the flow of
funds through Jefferson’s child welfare system. Finally, Jefferson is scored on the extent of its
fidelity to managed care principles using Fidelity of Organizational Implementation Scales.

Progress in the Introduction of Managed Care Processes
Jefferson County is functioning as a managed care entity as it is assuming risk by entering into
an MOU with the CDHS. In its original MOU, Jefferson County was permitted to keep unspent
funds but was required to cover overages that might occur. It is expected to perform at a certain
level on specific outcomes.

In 2001, the CDHS offered counties a choice between keeping any of their own unspent funds or
being eligible for participation in the distribution of unspent funds from other counties at the end
of the year. Jefferson County chose to give up the ability to retain unspent funds in order to
participate in surplus distributions from the state.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                  128               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The main focus of Jefferson County’s managed care initiative has been on utilization
management and the development of its provider network. Of these two, the one that is the most
developed is utilization management.

The central feature of Jefferson County’s managed care initiative has been the collaborative
project between Jefferson County Department of Social Services (JCDSS) and Jefferson Center
for Mental Health (JCMH). JCDSS and JCMH first entered into a managed care collaborative in
1997 through the development of the Service Utilization Review Team (SURT) and Child
Placement Agency (CPA) Medicaid Transfer project.

The SURT, which is facilitated by JCDSS, meets weekly to review referrals for day treatment,
non-preferred CPA providers, and RTC residential treatment center placements. Panel members
are able to commit fiscal resources from their respective agencies for the children and families
reviewed by the team and include a JCDSS rotating child welfare program administrator, the
JCDSS delinquency intake supervisor, the clinical supervisor of the Jefferson County Health
Department Substance Abuse program, a rotating JCMH outpatient manager, the JCMH
supervisor of in-home delinquency services (Turnabout Program), and a Jefferson county public
schools representative. Representatives from SB 94 and Probation occasionally sit on the panel.

The SURT does not review all referrals for out-of-home placement. Caseworkers, given
supervisor approval, have direct access to kin placements, internal JCDSS administered foster
and group homes, and JCDSS Preferred Provider Child Placement Agency placements. CPAs
may become preferred providers if they sign an annual letter of agreement with JCMH indicating
their willingness to negotiate rates, based on the Needs Based Care Assessment, for each
child/youth placed into their facility.

A particular strength of Jefferson County’s utilization management process is that out-of-home
placements are reviewed on a monthly basis. This is the most ambitious time frame for review
among all the managed care counties.

In regard to network capacity, Jefferson County has not determined the capacity that it requires.
As a result, it is limited in developing a configuration of services that addresses the needs of
children and families now and in the future. Financial incentives for the development of services
that represent present best practices are also difficult to implement without knowing the overall
capacity that is required.

Jefferson County has moved to address gaps in its provider network by developing various
contracts for multi-systemic therapy in-home delinquency services, bilingual home-based
therapeutic services, and emergency crisis placements. Jefferson County contracts for three
guaranteed shelter beds for adolescents and awarded contracts for emergency crisis CPA
placements.




William M. Mercer, Incorporated                129              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
In the area of financial incentives, Jefferson County has not entered into relationships with
providers that involve financial incentives. Jefferson has not yet developed a formal quality
improvement program and has not introduced performance indicators for providers.

Like other counties, Jefferson County does not have a management information system that
allows it to adequately manage utilization or its system of care. It relies on state information
systems that do not provide information that is timely or sufficiently useful for administering a
managed care program. The CDHS has recently implemented a new system, TRAILS, which has
the potential to yield more useful and timely information.

Process Flow
To illustrate the flow of processes involving people and funds in Jefferson County, we have
constructed eight Blueprints. These show:
1. the flow of persons through the county in SFY 1997, prior to the initiation of managed care
   reforms;
2. the overall flow of persons through the county in SFY 2001, midway through the fourth year
   of managed care reforms;
3. additional detail on the flow of persons through the county department;
4. additional detail on the flow of persons through the county’s provider network;
5. the flow of funds through the county in SFY 1997, prior to the initiation of managed care
   reforms;
6. the overall flow of funds through the county in SFY 2001, midway through the fourth year of
   managed care reforms; and
7. additional detail on the flow of funds from the state to the county, including financial
   reporting.
8. Additional detail on the flow of funds through the county’s provider network, including risk-
   sharing arrangements.

We have used six colors in the Blueprints:
1. black is used for the organizations involved in paying for, managing, and providing child
   welfare services;
2. red is used for the unmanaged flow of people through the system and includes county
   populations, populations in need, referral groups, movement through different teams and
   areas within the county child welfare department, and referrals to community providers;
3. green depicts child welfare funds that flow directly to the county child welfare department, as
   well as the state and federal funds that flow through the CDHS to the county child welfare
   department;
4. blue depicts the flow of non-child welfare funds that are critical to providing needed care to
   recipients of child welfare services;



William M. Mercer, Incorporated                 130              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
5. Purple depicts the managed flow of both persons and funds through the county child welfare
   managed care pilot initiative; and
6. Yellow depicts the court system in each county.

In the Blueprints, the flow of people through the system is depicted in a left to right fashion,
beginning with the entire population potentially served by the child welfare system, narrowing to
the group of persons actually served, and following these people through the organizations and
services they would typically encounter as they pass through the system. The flow of funds,
however, goes from the top to the bottom of each chart. The drawing traces the flow of funds
from the payers that fund the system, to the administrative entities that allocate and account for
these funds, to the county child-serving agencies and providers that ultimately spend and manage
them.

A set of symbols is also used. They are illustrated in the following diagram.

Organizational Symbols:
               = Federal, state or county payors
               = State agencies
               = Local agencies or providers
               = Local entities that bears risk

Risk Symbols:
  S   = Shared risk of financial utilization
  F   = Full risk of financial utilization

Out-of-Home Placement Symbol:
  $            = Most expensive out-of-home
                   placements (RTC, CPA)
Financial Reporting Symbol:
   Financial
  Reporting
  by County    = Locus of financial reporting




William M. Mercer, Incorporated                    131            Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Flow of People—Pre-Managed Care—Spring 1997
The Blueprint depicted on page 134 (entitled “Jefferson County—Spring 1997—People Flow”)
shows the flow of people through Jefferson County’s child welfare system in the Spring of 1997,
just prior to the initiation of their managed care pilot later that year. Beginning on the far left of
this diagram, the overall county population breaks down into the smaller group of persons who,
at any point in time, are in need of child welfare services.

From that group, three sub-groups of referrals emerge for the DSS. The county maintained a
centralized informational phone number to give callers information for contacting the correct
JCDSS department, be it child welfare, TANF, or other program. Children with reported
instances of abuse or neglect accessed services directly through the department’s Intake Unit.
Families with basic needs were referred to the county’s TANF program.

At Intake, workers were able to provide prevention services, following families with lower needs
for up to 30 days in order to connect them with other community services and avoid ongoing
child welfare services. Not all families took advantage of this voluntary service but those who
did were often able to have their needs met without additional resources.

For youth in conflict and youth with delinquency concerns, a partnership between JCDSS, the
JCMH, SB 94, the District Attorney’s Office, the school district, the sheriff’s department and
local law enforcement had developed a separate multi-agency Juvenile Assessment Center
(JAC). Co-located with the JAC, the Family Adolescent Crisis Team (FACT) was able to follow
youth with lower level needs and their families for up to 30 days, often helping them stay in the
community without additional services. If ongoing services were needed, the JAC and/or FACT
were able to refer cases to JCDSS. Delinquent youth could also access JCDSS services through
placement evaluators referred by the court or probation.

The Intake units screened out cases that clearly did not meet criteria and referred them to other
services in the community. Cases meeting criteria were referred to an array of ongoing and
specialized teams, including ongoing teams for adolescents, children, and expedited permanency
planning, as well as specialized services (such as intensive family treatment and home-based
intensive intervention services). Resource Development staff supported the work of the ongoing
teams through the development of specialized resources, including foster parent recruitment and
adoption supports.

Many children, youth, and families served by the county required additional services provided by
agencies in the community. Workers on ongoing and specialized teams were able to directly
access expensive out-of-home placements in RTCs, CPAs, and county administered foster
homes, as well as outpatient substance abuse services, specialty mental health providers, and
mental health services through JCMH. Children needing inpatient services first had to go through
JCMH, which managed all indigent and Medicaid-funded mental health benefits. While some
high need or multi-agency involved youth accessed multi-agency planning from time-to-time, the
majority of external referral decisions were not formally coordinated.

A critical factor for any involuntary case served by the county was the court system. Individual
workers presented recommendations for services to the court and the court implemented them as

William M. Mercer, Incorporated                   132              Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
it saw fit. Voluntary cases were not subject to court involvement. While the Blueprints highlight
the role of the court at the point where it impacted decisions regarding referrals to out-of-home
placements versus other community services, the court’s involvement with involuntary cases was
much broader, beginning at or prior to Intake and extending throughout each family’s final
disposition.




William M. Mercer, Incorporated                133              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
William M. Mercer, Incorporated   134            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of People—Post-Managed Care—Spring 2001

In the fourth year of its implementation of managed care reforms, the flow of persons through
the Jefferson County child welfare system has significantly changed. A review of the Base
Blueprint for Spring 2001 (see the Blueprint depicted on page 136 entitled “Jefferson County—
Spring 2001—People Flow”) shows the same basic population served, but reorganization within
JCDSS and its provider network. To review these changes with greater specificity, two detailed
Blueprints were created for this year: a Department Internal Detail Blueprint and a Provider
Network Detail Blueprint.




William M. Mercer, Incorporated              135                Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
William M. Mercer, Incorporated   136            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Department Internal Detail. Looking at the Department Internal Detail Blueprint (see the
Blueprint depicted on page 138 entitled “Jefferson County—Spring 2001—Department Internal
Detail”), it can be seen that the county population still breaks down on the left into a smaller
group of persons in need, who, in turn, sort into the three primary groups of persons served by
the county: youth in conflict, youth with delinquency concerns, and children with reports of
abuse or neglect. Youth in conflict and with delinquency concerns still present through the JAC
and children with abuse and neglect directly through the county Intake unit. In the last year, the
JAC was strengthened by a formal intergovernmental agreement for its funding.

The FACT and Intake teams still offer preventative services. Child welfare and TANF programs
are separate overall but there are some joint programs. For example, TANF supports preventive
kin case managers who work alongside child welfare kin caseworkers but with different
subgroups of children. Cases meeting criteria are referred to ongoing and specialized teams
within JCDSS. In Spring 2000, Jefferson County combined its adolescent and child ongoing
services into an integrated unit.

In 2000–2001, the focus in Jefferson County (as in the other managed care counties) has been on
program consolidation and refinement. As with other counties, staff turnover has increased
related to the increasingly competitive job market for human service employees. A new
partnership with the JCMH can also be seen, where JCMH now operates a satellite mental health
office at JCDSS that provides both Intake and liaison services for JCMH.

Jefferson County has developed a new support for caseworkers making referrals to out-of-home
placements it refers to as the “Placement Desk.” This position gathers information daily on
available beds at RTCs, CPAs, and county-administered foster homes. It is in the process of
developing a capacity to handle rate negotiations with county-administered foster care providers.




William M. Mercer, Incorporated                 137              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   138            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Provider Network Detail—People Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 141 entitled “Jefferson County—Spring 2001—Provider
Network Detail—People Flow”), the primary managed care reform implemented by Jefferson
County can be seen. The SURT is collaboration between JCDSS and JCMH. That initial
collaboration expanded in early 2001 to include participation by the school district. The SURT
meets weekly with the county foster care unit to staff cases believed to need CPA levels of care.

The SURT meets weekly to review referrals for day treatment, non-preferred CPA providers, and
RTC placements. The JCDSS Managed Care Administrator facilitates the SURT. Team members
are able to commit fiscal resources from their respective agencies for the children and families
reviewed by the team, which includes a JCDSS rotating child welfare program administrator, the
JCDSS delinquency intake supervisor, the clinical supervisor of the Jefferson County Health
Department Substance Abuse program, a rotating JDMH outpatient manager, the JDMH
supervisor of in-home delinquency services (Turnabout Program), and a Jefferson County Public
Schools representative. Representatives from Senate Bill 94 and Probation occasionally sit on the
panel. The Child’s family and the youth, probation officer, therapist, and JCDSS caseworker are
invited to individual reviews.

Initially, the SURT focused its activities on CPA referrals. Late in 1999, JCDSS and JCMH
identified key CPAs willing to negotiate rates and provide quality services and formalized
collaborative relationships with them as preferred CPA providers. Caseworkers, given supervisor
approval, may directly access foster and group home placement with CPAs on the JCDSS
Preferred Provider List without obtaining prior approval from the SURT. This has freed up
SURT resources to begin to focus on RTC placements, and those remaining CPA placements not
on the preferred provider list. In 2000–2001, the Managed Care Administrator who manages the
SURT began to manage a prior authorization process for all new CHRP and day treatment
placements, two populations that have increased significantly in recent years.

Another change in 2000-2001 involved Jefferson County’s substance abuse providers for CORE
services and referrals to substance abuse services funded under the state’s ADAD. Signal, the
managed care entity for ADAD-funded substance abuse services in Jefferson County, requires
JCDSS to contract CORE funding for substance abuse services through Signal in order for
JCDSS referrals to gain access to Signal’s ADAD-funded services. This limits substance abuse
referrals under both CORE and ADAD funding to Signal’s provider network, which includes
only one local provider (the Jefferson County Health Department) and no local youth resources
(two adolescent providers are in Denver).

The relationship between JCDSS and the local court has evolved in SFY 2001. Similar to
procedures already used in Arapahoe and El Paso counties, Jefferson County has instituted an
Administrative Review process, which avoids additional court proceedings for cases where the
child remains in out-of-home placement following the Permanency Planning Hearing, and all
parties concur with the Family Service Plan. Instead of going to court for annual reviews, the
state-conducted foster care review is used as a planning forum. The foster care review
recommendations are submitted to the court to be made an order of the court; however, any party
disagreeing with the finding may request an appearance before the court. This is designed to save
staff, court, and family time, and promote a more collaborative decision-making process.


William M. Mercer, Incorporated                139              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
JCDSS also gave the First Judicial District a grant award to cover the cost of a mediator in
Dependency and Neglect (D&N) cases. In addition, the First Judicial District hired a D&N court
facilitator to conduct case conferences and promote a positive working relationship between the
court and social services. Overall, these are reported to have strengthened Jefferson County’s
relationship to its courts. However, the JCDSS/court relationship continues to have a major
impact on the managed care initiative and JCDSS operations as a whole. Some magistrates
reportedly continue to intervene directly in JCDSS operations and relationships with JCDSS
vendors.

Other features of the provider network include:
§    Jefferson County offers both certified foster care homes that pay kin regular foster care rates
    and certified kinship homes that have lower credentialing requirements and, as a result, pay kin
    providers less for a child’s cost of care.
§    Specialty provider capacity on the Provider Network Detail Blueprint refers to approximately
    30 CORE service contracts for day treatment, outpatient sex offender services, home-based
    services, and life skills providers.
§     Placements through DYC take place in a separate system of care with little interaction at the
    case-planning level with JCDSS. After reviewing a case, the SURT may make a
    recommendation to the court for commitment. Although invited to SURT Review Team
    meetings, probation does not sit as a regular panel member of SURT when such
    recommendations are made. During the past six months, probation and DYC have been
    conducting their own case reviews of youth at risk of commitment that include an invitation to
    the JCDSS caseworker.

Other provider network developments in 2000–2001 include:
§    CPA network capacity has been enhanced through an award to Maple Star CPA to develop
    up to 12 receiving/crisis home slots (there are currently two receiving homes). These are
    available for children and adolescents.
§    Multisystemic therapy contracts are in place with one provider and are under development
    with another.
§    Specialized case management, home-based care, and other services for Spanish-speaking
    populations were put in place in the last six months.




William M. Mercer, Incorporated                  140              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   141            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Pre-Managed Care—Spring 1997
The Blueprint depicted on page 144 (entitled “Jefferson County—Spring 1997—Funding Flow”)
shows the flow of funds through the Jefferson County child welfare system prior to the initiation
of their managed care reforms in late 1997. In July 1997, funds from various federal sources
(Titles IX and XX) and the state general fund were combined by the CDHS and passed on to
Jefferson County. The bulk of these uncapped funds flowed through separate funding streams
which had to be separately tracked and accounted for by the county. Furthermore, unspent funds
in one category could not be used in other categories, creating an incentive to keep spending in
line with historical patterns. There were eight separate state funding streams, including:
§    Two streams funding county administration (to fund county staffing, primarily), one
    comprised of 100 percent state dollars and one of 80 percent state/20 percent county funds;
§    A large stream of funds for out-of-home placements (OOHP);
§    A significant funding stream for CORE services, originally developed to fund alternatives to
    out-of-home care such as family preservation, outpatient substance abuse, and mental health
    services and day treatment; and
§    Four smaller funding lines (e.g., Special Circumstances Child Care, Subsidized Adoption,
    CHRP, and Case Services).

Jefferson County also received other specialized state funding, including EPP funds. It should be
noted that reductions in the level of such specialized funds over time have been seen as
impacting the availability of services.

In addition to the budgeted funding streams, the CDHS, and through them the State General
Fund, had the full risk for any cost overages that exceeded the CDHS’s ability to shift unspent
funds from other areas of the state. In addition, no county was able to keep any unspent funds.
As a result, counties had little incentive to limit their spending of child welfare funds and many
spent their entire allocation and more.

Jefferson County, in turn, contracted with various community agencies for services. In addition
to funding streams flowing through DSS, Jefferson County’s child welfare system also depended
on services paid for by other agency funding. These included:
§     Medicaid (Title XIX) funding paid through the HCPF to RTC providers for the treatment
    component of the services provided in these facilities—The county DHS directly pays for only
    room and board. Treatment costs are paid directly from the State Department of Human
    Services to HCPF but are then charged back to the county at year end against its allocation.
    The county has no control over the rates paid to providers since they are determined through a
    cost-based approach within available appropriations;




William M. Mercer, Incorporated                 142               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
§     Medicaid (Title XIX) funding for mental health services transferred through the Department
    of HCPF to the Mental Health arm of the DHS Office of Health and Rehabilitation Services—
    In Jefferson County, funding for all Medicaid inpatient and outpatient mental health services
    has been capitated and, as a result, flows through the entity awarded the MHASA contract for
    that area of the state, Pike’s Peak-Options Colorado Health Networks;
§     ADAD funding through the DHS Office of Health and Rehabilitation Services—In Jefferson
    County, these funds are managed by Signal, the managed services organization for substance
    abuse services. Signal pays for additional substance abuse services beyond those paid by DSS
    through its CORE funds; and
§     JAC—JCMH, the District Attorney, the school district, SB 94, the sheriff’s department, and
    local law enforcement co-fund the JAC with JCDSS.

As can be seen in the Blueprint, child welfare funds were the least managed of the primary
funding streams for expensive out-of-home placements, such as inpatient, RTC, and CPA. In
1997, inpatient and other out-of-home placements through mental health were subject to
managed care controls, and RTC placements through DYC were subject to the placement
restrictions of the juvenile justice system.




William M. Mercer, Incorporated                 143              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   144            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Post-Managed Care—Spring 2001
The Blueprint depicted on page 146 (entitled “Jefferson County—Spring 2001—Funding Flow”)
shows the flow of funds under the managed care pilot in the Spring of 2001. The primary federal
and state funding sources flowing to the CDHS remains essentially unchanged; however, the
flow of funds from the CDHS has changed dramatically. In addition to creating the managed care
pilots, SB-218 collapsed seven of the categorical funding streams that existed in the Spring of
1997 into a single child welfare block. Only CORE services remained as a separate primary
funding stream with separate accountability and accounting to prevent funds from being diverted
to other uses. This block funding approach was implemented for all counties, not just managed
care counties.

All Colorado counties, including Jefferson, are now responsible for any overages from budgeted
expenditures, shifting full downside risk from the state CDHS and General Fund to JCDSS and
the county, as designated by the full risk triangle in the JCDSS box. In addition, as a managed
care pilot, Jefferson is now able to retain any unspent funds from the seven combined funding
streams, thus taking on the upside risk for the services it provides. To do this Jefferson has to
allow non-managed care counties to use unspent funds from other areas of the state to cover their
overages before being eligible for these. The county is still eligible for excess Title IV-E
distributions it helps earn.

In the first two years of operating under a capped allocation, the CDHS returned unspent general
funds to the legislature. During that period of time, as well as during the third year, funding for
the child welfare block was based on “population increase.”

During the third year, SFY 2000, all Colorado counties collectively overspent their budgets by
approximately $20 million. The CDHS requested and received a $14 million supplemental
appropriation in SFY 2001 based on projected caseload growth. Because counties overexpended
so significantly in SFY 2000, the Child Welfare Allocation Committee approved allowing
managed care counties a choice to either participate in the statewide surplus distribution or retain
their own unspent funds.

Jefferson County has chosen the MOU option where they give up the ability to retain their own
unspent funds and, instead, are able to participate in surplus distributions from the state. This can
be seen in both the Base Blueprint and the State Funding Detail Blueprint for the county (also
see the Blueprint depicted on page 148 entitled “Jefferson County—Spring 2001—State Funding
Detail”).




William M. Mercer, Incorporated                 145               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   146            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
State Funding Detail. Looking at the State Funding Detail Blueprint (see the Blueprint depicted
on page 148 entitled “Jefferson County—Spring 2001—State Funding Detail”), another systemic
issue related to the cost overruns can be seen. This is the area of financial reporting. All counties
track their internal costs for staff and overhead. Unlike many other counties, Jefferson tracks its
own costs for its purchased services, although it take specialized, labor-intensive queries to use
those data. Because of the difficulty aggregating its service data, Jefferson County has relied on
CDHS reporting to track provider expenditures and manage utilization.

Jefferson County cites these reports as a key reason for choosing the MOU option to protect it
from downside risk. Both data accuracy and timeliness are seen as issues limiting Jefferson
County’s risk management capacity, either locally or from state-produced reports. Jefferson
County created a new financial budget director position in the last year to begin to address these
data needs. Jefferson County currently relies on the state for their primary data on utilization of
purchased services. They track costs, but their data is not easily reconcilable with the state
formats currently used.

JCDSS now earns a federal Medicaid (Title XIX) match on the treatment portion of its CPA
costs through its implementation of a CPA Medicaid transfer that sends these funds through the
state Department of HCPF to JCMH and, thereby, to CPA providers. Jefferson County takes
advantage of this to receive a roughly 50 percent federal match on the treatment component of
the care provided in CPAs.

The risk sharing arrangement for the CPA Medicaid Transfer Project between JCDSS and JCMH
is outlined in the annual MOU signed by both agencies. The MOU clearly indicates that JCMH
incurs all fiscal risk in the program and unspent funds are mutually split between the agencies;
however, the 2000–2001 treatment costs exceeded the contracted amount primarily due to an
increase in the number of children placed in CPAs and the two agencies negotiated a 50/50 split
of the overage.




William M. Mercer, Incorporated                 147               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   148            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Provider Network Detail—Funding Flow. Looking at the Provider Network detail Blueprint
(see the Blueprint depicted on page 151 entitled “Jefferson County—Spring 2001—Provider
Network Detail Funding”), it can be seen that, for CPA services, JCDSS, and JCMH have
identified preferred providers. Preferred providers sign a letter of agreement with JCMH as the
administrator of treatment funds for CPA care. The agreement includes providers’ willingness to
negotiate their CPA rates and to use the standardized needs-based level of care worksheet (the
Needs-Based Care Assessment). This Assessment is focused on the individual needs of each
child and is used to negotiate with each CPA an individualized rate for room, board, and other
special needs. Part of the increased revenue realized through the CPA contract with JCMH was
used to increase the rates paid to county-administered foster care homes to help the county better
compete with CPAs to attract foster parents.

In 2000, in keeping with its expanded focus to include RTC care, the SURT had begun to use the
Colorado Client Assessment Record (CCAR) to compute level of care scores to help set the RTC
treatment rate paid by HCPF. In 2001, RTC costs have increased, largely through new categories
of costs in addition to the basic room and board and Medicaid treatment rates. The two primary
networks of RTC providers (CCM and RocNet) are negotiating network fees to fund enhanced
access and coordination for Jefferson County referrals to their network providers.

There has been upward pressure on CPA room and board and specialty provider costs funded
under CORE in 2001. This is apparently a result of increasing provider rates attributed to rising
labor costs. Some rates have increased while others are still under negotiation. JCDSS is working
with other child welfare agencies in the Northern Consortium to develop a uniform approach to
CPA rates that would include incentives tied to performance goals.

JCDSS reports that CORE and ADAD funding together did not cover all the needs for JCDSS
clients during the most recent year. Signal reallocated unspent ADAD funding to cover some
services for Jefferson County clients and transferred some mothers into a specialized federal
program; however, JCDSS also directly covered these under-allocated substance abuse treatment
needs with child welfare block funds.




William M. Mercer, Incorporated                149               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   150            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. When counties implement managed care in a way that is consistent with
such expectations, we call that “process fidelity.”

Process fidelity is scores in each of four domains: funding flexibility, inter-agency coordination,
integration of informal networks and non-traditional services, and implementation of early
intervention and prevention programs, on a scale from one to four. Once a score is calculated, it
is placed within one of four fidelity categories: early, emerging, maturing, and mature states of
fidelity.

Funding Flexibility
Jefferson County was rated in the Emerging Stage of fidelity in this domain. The county has a
contract with JCMH that requires JCMH to provide “appropriate therapeutic services to all
children in Child Placement Agencies,” but implementation of the contract has not resulted in
JCMH bearing the entire financial risk. In addition, Jefferson County provides access to flexible
funds for families. The county does not currently utilize managed care financing techniques with
other providers, nor does it use incentives with providers or other agencies.

Interagency Coordination
In this domain, Jefferson County was given a rating of Mature Stage fidelity. Jefferson has
developed formal, contractual relationships with several child-serving agencies and they have
developed service planning processes and administrative coordination processes with multiple
child-serving agencies. In addition, Jefferson County appears to have the capacity to integrate
data with many other child serving agencies in the county.

Integration of Informal Networks and Non-Traditional Services
Jefferson County’s fidelity score placed it in the Emerging Stage of fidelity in this domain.
Jefferson County reported that they have developed relationships with various informal networks
of support in the community, including community shelters, churches, and action centers, but it
is not clear how extensive the collaboration is with these informal networks of support. They
utilize paraprofessionals on a limited basis. Informal networks of support were not reported to be
involved in service planning nor are there organizations outside of the formal child-serving
system that are involved with the county in providing funding for services.




William M. Mercer, Incorporated                151               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Implementation of Early Intervention and Prevention Programs
Jefferson County was found to be performing at the Emerging Stage of fidelity in this domain.
The county currently is delivering limited prevention programming and did not report any new
developments in the area of prevention and early intervention.




William M. Mercer, Incorporated              152               Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in Jefferson County in
the nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in Jefferson to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.

Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                          Figure 58
             Percentage of Children Who Experienced a Confirmed Recurrence of
                  Abuse and/or Neglect within a 12-Month Period after being
                                  Abused and/or Neglected

       10
         9
         8
                                         6.5
         7
         6
         5
                    3.3                                       3.4                  3.1
         4
         3
         2
         1
         0
                   SFY 97              SFY 98               SFY 99               SFY 00




William M. Mercer, Incorporated                153              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect

                                      Figure 59
        Confirmed Abuse and Neglect per 1,000 Children in the County Population




       20

       15

       10
                                                         5.7
                    4.3               4.7                                     4.6
        5

        0
                  SFY 97             SFY 98             SFY 99               SFY 00


Indicator 3: Referrals and Investigations

                                          Figure 60
                 Referrals of Alleged Abuse and Neglect per 1,000 Children
                                  in the County Population




        70
        60
        50
        40          31.4
                                     28.9              26.7                26.5
        30
        20
        10
         0
                   FFY 97           FFY 98            FFY 99              FFY 00




William M. Mercer, Incorporated              154            Colorado Child Welfare Evaluation
                                                   Second Interim Implementation Status Report
                                            Figure 61
                     Investigations of Alleged Abuse and Neglect per 1,000
                              Children in the County Population



       50

       40

       30
                    18.6                                       16.4                  16.8
       20                                 16.2

       10

        0
                  FFY 97                FFY 98                FFY 99                FFY 00


Indicator 4: Discharge to Reunification
It should be noted that this indicator does not directly measure reunification in that is represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.

                                         Figure 62
                 Percentage of Children placed with Parents at Case Closure

       100
        90
        80
        70           57.1                 54.8                  50.9
        60                                                                           47.5
        50
        40
        30
        20
        10
         0
                   SFY 97                SFY 98                SFY 99               SFY 00




William M. Mercer, Incorporated                  155              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 5: Children in Out-of-Home Placement

                                         Figure 63
                Children in Out-of-Home Placement at some Point during the
                  Fiscal Year per 1,000 Children in the County Population

      20
      18
      16
      14
      12
      10
       8
                                                                   5                 5
       6         3.6                             3.7
                                   3.2
         4
         2
         0
               SFY 96             SFY 97     SFY 98              SFY 99           SFY 00


Indicator 6: Length of Out-of-Home Stay

                                       Figure 64
             Out-of-Home Placement Length of Stay (in Months) for Closed Cases




    14                                       12
    12
    10          8.4               8.9                                               9

     8                                                          6.6

     6
     4
     2
     0
              SFY 96          SFY 97        SFY 98             SFY 99           SFY 00




William M. Mercer, Incorporated            156                  Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 7: Percentage in Positive Placements

                                          Figure 65
                      Percentage of Positive Placements at Case Closure
                       (Parents, Kin, Relatives, Adoption, Guardians)




        100                                                87.2                88.7
                  85.6               86.2
        80

        60

        40

        20

          0
                 SFY 97             SFY 98                SFY 99               SFY 00

Indicator 8: Number of Adoptions

                                        Figure 66
                   Adoptions per 1,000 Children in the County Population



  1.8
  1.6
  1.4
  1.2
   1
  0.8
                    0.5
  0.6
                                                                              0.32
  0.4                                          0.28
  0.2
   0
                  FFY 98                      FFY 99                         FFY 00




William M. Mercer, Incorporated              157                Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 9: Time to Adoption

                                      Figure 67
        Average Months from Termination of Parental Rights to Finalized Adoption



   20
                                                                         16.3
                   14.7
   15
                                           11.8

   10


    5


    0
                  FFY 98                  FFY 99                        FFY 00




William M. Mercer, Incorporated          158                Colorado Child Welfare Evaluation
                                                   Second Interim Implementation Status Report
Outcome Comparisons

                                        Table 8
       Jefferson County SFY 2000 Outcome Results Compared to Other Counties, to
         Benchmarks, and to the Criterion Reference Standards Set for SFY 2000
                                      Other        Non-
                                                                                               Criterion
                         Jefferson   Managed     Managed
Outcome Indicator                                                     Benchmark                Reference
                          County       Care     Care Large
                                                                                               Standard
                                     counties    counties
                                                                National: 11%
Indicator 1:                                                                               Acceptable: <10%
Confirmed re-abuse
                         3.1%         3.9%        3.0%          Comparison States: 12%
                                                                                           Outstanding: <2%
                                                                (Utah)
Indicator 2:
                                                                National: 11.8             Acceptable: <10
Confirmed abuse per        4.6         5.2            5.1       Comparison States: 8.8
1,000 child population                                                                     Outstanding: <4
Indicator 3a:
Referrals of suspected                                          National: 72.1             Acceptable: <45
abuse per 1,000 child
                          26.5        49.1            48        Comparison States: 53.4    Outstanding: <40
population
Indicator 3b:
Investigations of                                               National: 43.5             Acceptable: [not set]
suspected abuse per
                          16.8        27.0            31        Comparison States: 28.3    Outstanding: [not set]
1,000 child population
Indicator 4: Percent
                                                                National: 59%*             Acceptable: >52%
discharged to            47.5%        45%         57.7%         Comparison States: 68%*
reunification
Indicator 5:
Children in out-of-                                             National: 8.1**            Acceptable: <8
home care per 1,000
                            5          8.3         10.6         Comparison States: 5.6**   Outstanding: <4
child population
Indicator 6:
                                                                National: 10.8             Acceptable: <10
LOS in out-of-home         9.0         8.2            6.2       Comparison States: 12.3
placement (in mos.)                                                                        Outstanding: <7
Indicator 7:
                                                                National: 83%              Acceptable: >82%
Percent discharged to    88.7%        86%         83.8%         Comparison States: 90%
positive placements                                                                        Outstanding: >90%
Indicator 8:
                                                                National: .66              Acceptable: >.4
Adoptions per 1,000        .32         .75            .66       Comparison States: .52
child population                                                                           Outstanding: >.9
Indicator 9:                                                                               Acceptable:
Time from termination                                           National: N/A              <18 months
of parental rights to
                          16.3        11.5         20.6         Comparison States: N/A     Outstanding:
adoption                                                                                   <9 months
 * Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
** Benchmarks are for point in time (on September 30) versus at any time in fiscal year (Indicator 5).




William M. Mercer, Incorporated                 159                  Colorado Child Welfare Evaluation
                                                            Second Interim Implementation Status Report
Summary of Findings
Jefferson County had one of the lowest rates of out-of-home placement (5 per 1,000 children in
the population) compared to other large counties in Colorado. At the same time, Jefferson
County’s average length of stay in out-of-home placement was one of the longest and increased
in the last year. As was mentioned above, in the discussion of Jefferson County’s outcome
results, it may be that, in general, lower rates of out-of-home placement lead to the most difficult
cases being placed out-of-the-home. This might increase the average length of stay. This is a
hypothesis that we will explore further in the next year or the evaluation. It should be noted that
Jefferson County’s average length of stay was lower than both the national and comparison state
benchmarks.

Like several of the other managed care counties, Jefferson County has evidenced a steady
decrease in the percentage of children placed with parents at case closure (among children who
had ever been in out-of-home placement). At the same time, the overall percentage of children in
positive placements at case closure has steadily increased from SFY 1997 to SFY 2000. Again,
as more emphasis is placed on expedited permanency in Colorado, rates of reunification may
decrease, even though overall rates of positive placement increase. We will explore this
hypothesis further in the next year of the evaluation.

The rate of recurrence of abuse and neglect in Jefferson County during SFY 2000 was low in
comparison to other managed care counties, nearly the same as the rate for all other
(non-managed care) large counties combined, and much lower than national and comparison
state benchmarks. The overall number of children who were victims of abuse or neglect per
1,000 children in the population was slightly lower than in other large counties in Colorado and
much lower than national and comparison state benchmarks. It is not clear whether Jefferson
County’s relatively low rate of abuse and neglect per 1,000 children in the population is due to
the county’s extremely low rates of referrals and investigations of abuse or whether there is a low
base rate of abuse and neglect in the population relative to other counties.

With regard to adoption-related outcomes, Jefferson County continues to have a low rate of
adoptions and a relatively long time from termination of parental rights to adoptions relative to
other managed care counties. The county’s rates of adoption have dropped somewhat from
FFY 1998 to FFY 2000 and its average number of months from termination of parental rights to
adoption has increased slightly during that same time.




William M. Mercer, Incorporated                160                Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as residential treatment centers, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher cost care as well as trends over several years. The results of these
analyses are shown in this section for Jefferson County by itself and in a later section for all of
the managed care counties together.

Figures 68 and 69 compare Jefferson County’s allocation at the beginning of SFY 2000 to its
expenditures at the end of the year. Figure 68 provides a comparison of the two main child
welfare funding streams, CORE services and the child welfare block. Figure 69 provides detail
regarding the funding streams that compose the child welfare block.

Figure 68 shows that Jefferson County experienced a deficit in SFY 2000 in the child welfare
block. Figure 69 shows that Jefferson’s use of more expensive out-of-home placements, RTC
and CHRP, was the main cause of the deficit. It appears that Jefferson was unable to operate
within its allocation in SFY 2000.
                                          Figure 68
               CORE Services and Child Welfare Block Allocation vs. Expenditure*
                                 Jefferson County, SFY 2000

        $25,000,000
                                                                                       $19,268,858
        $20,000,000                                                      $16,928,544
        $15,000,000

        $10,000,000
                                                         $76,793




         $5,000,000     $3,152,244     $3,075,451

                $-
                                     CORE Services                               Child Welfare Block
                                                                                                           $(2,340,314)




         $(5,000,000)

        $(10,000,000)



 * Before Year End Close Out Process                Allocation            Expenditure         Difference




William M. Mercer, Incorporated                                    161               Colorado Child Welfare Evaluation
                                                                            Second Interim Implementation Status Report
                                                            Figure 69
                                         Child Welfare Block Allocation vs. Expenditure*
                                                   Jefferson County, SFY 2000




                                          $19,268,858
                           $16,928,544




                                                         $11,623,067

                                                                       $10,847,169
       $25,000,000
       $20,000,000




                                                                                                    $5,832,875
                                                                                       $3,104,634
       $15,000,000




                                                                                                                                              $1,566,715

                                                                                                                                                           $1,566,715
                                                                                                                               $1,022,100
                                                                                                                    $634,128
       $10,000,000
        $5,000,000
                 $0
                       Total CW Block                   Total 80/20                  RTC Medicaid                CHRP Medicaid              100% Admin.


* Before Year End Close Out Process                                                                 Allocation                   Expenditure



Figures 70 and 71 provide a picture of how Jefferson County spent its child welfare block in
SFY 2000. Figure 70 shows the percentage of funds expended by budget item, while Figure 71
shows the actual amount expended. Of interest is that Jefferson County spent a greater
proportion of its budget on RTC in SFY 2000 than any other managed care county.

                                          Figure 70
               Percentages of Child Welfare Block Expenditure* by Budget Items
                                 Jefferson County, SFY 2000



                       100% Admin.                                                                                             Out of Home Placement
                       8.1%                                                                                                    Services
               CHRP Medicaid                                                                                                   21.5%
               5.3%                                                                                                            Subsidized Adoptions
                                                                                                                               6.8%
                                                                                     Other                                     Case Services
                                                                                     56.3%                                     0.7%

               RTC Medicaid                                                                                                    Special Circumstances
               30.3%                                                                                                           Child Care (SCCC)
                                                                                                                               1.3%

                                                                                                                               80/20 Admin.
                                                                                                                               26.0%



            * Includes Medicaid Fund and before Year End Close Out Process




William M. Mercer, Incorporated                                                      162                            Colorado Child Welfare Evaluation
                                                                                                           Second Interim Implementation Status Report
                                                Figure 71
                            Child Welfare Block Expenditure* by Budget Items
                                       Jefferson County, SFY 2000
  $8,000,000

                $5,832,875
  $6,000,000
                                                                                                                                               $5,016,991
                                                                     $4,147,612
  $4,000,000


  $2,000,000                                         $1,566,715                        $1,304,423
                                 $1,022,100
                                                                                                       $127,101         $251,043
          $0
               R T C M edicaid   C H R P M edicaid    100% A dmin.   Out o f H o m e    Subsidized     Case Services        Special            80/20 A dmin.
                                                                      P lacement        A do ptio ns                   C ircum s t a n c e s
                                                                       Services                                           Child Care
                                                                                                                            (SCCC)

  * Before Year End Close Out Process




Figures 72–74 illustrate the trends in out-of-home expenditures in Jefferson County from
SFY 1998 through SFY 2000. In most areas there are relatively modest changes except for RTC,
which shows an increase from 16.4 percent to 24.1 percent of the out-of-home budget. Jefferson
and Arapahoe spend a large proportion of their out-of-home placement budgets on shelter care
and receiving homes than other managed care counties.

                                                          Figure 72
                                             Out-of-Home Placement Expenditure
                                                 Jefferson County, SFY 1998


                                                                                   Home Based Program
                                                                                         2.4%
                                             RCCF
                                                                  RTC
                         Specialized         8.4%
                                                                 16.4%
                         Group Care                                                                           Receiving Home
                                                         Excludes    Relative Foster
                           16.4%                           RTC                                                     4.0%
                                                         Medicaid
                                                                      Care 1.0%
                                                                       Other
                                                                       13.1%

                                                                                                                Shelter Care
                                                                                                                   1.7%
                                                                             Transitional
                            Family Foster Care                                Program              Independent
                                  44.7%                                         4.6%                  Living
                                                                                                       0.4%




William M. Mercer, Incorporated                                           163                      Colorado Child Welfare Evaluation
                                                                                          Second Interim Implementation Status Report
                                              Figure 73
                                 Out-of-Home Placement Expenditure
                                     Jefferson County, SFY 1999



                         RCCF               RTC
                         11.8%             16.9%
              Specialized                                                    Receiving Home
              Group Care              Excludes   Relative Foster
                                        RTC                                       4.6%
                 9.6%                 Medicaid     Care 2.5%
                                                  Other
                                                  11.7%
                                                                            Shelter Care
                                                                               0.6%
                                                                          Independent
              Family Foster Care                             Transitional    Living
                    47.4%                                     Program         0.3%
                                          Home Based Program
                                                                3.0%
                                                3.2%




                                            Figure 74
                               Out-of-Home Placement Expenditure
                                   Jefferson County, SFY 2000



                                                   Home Based Program
                                           RTC           0.8%
                       RCCF
                                          24.1%
                        7.6%                                                  Receiving Home
              Specialized           Excludes                                       3.8%
              Group Care              RTC         Relative Foster
                 7.9%               Medicaid        Care 2.0%
                                                   Other
                                                  10.6%
                                                                                  Shelter Care
                                                                                     0.8%

                                                        Transitional          Independent
               Family Foster Care                                                Living
                                                         Program
                     47.7%                                                        0.1%
                                                           5.1%




William M. Mercer, Incorporated                      164                    Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Figures 75 and 76 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 75 shows that Jefferson County
directly operates most of its family foster care and group homes. Figure 76 shows that
Jefferson is the primary operator of both foster care and group homes.
                                         Figure 75
              Percentages and Amounts of Expenditure for Family Foster Care and
                      Group Home Services delivered by County vs. CPA
                           Jefferson County SFY 1998 to SFY 2000
       100%
        90%
                                                      $769,815                          $947,932
        80%             $1,255,960
        70%
        60%
        50%
        40%
                                                     $1,729,344                     $1,894,266
        30%             $1,387,333
        20%
        10%
         0%
                         SFY1998                      SFY1999                           SFY2000


   County Family Foster Care Homes and Group Homes         CPA Family Foster Care Homes and Group Homes




                                        Figure 76
            Percentages of Expenditure for Family Foster Care and Group Home
                        Services delivered by County and/or CPA
                         Jefferson County, SFY 1998 to SFY 2000

       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                     SFY1999                       SFY2000

  CPA Family Foster Care Homes     CPA Group Homes    County Family Foster Care Homes       County Group Homes




William M. Mercer, Incorporated                      165                   Colorado Child Welfare Evaluation
                                                                  Second Interim Implementation Status Report
Mesa County Findings

Mesa County is located in the western part of Colorado, 250 miles west of Denver. It had an
estimated child population of 29,138 in 2000. Mesa was accepted as a managed care county in
SFY 1998.

Mesa County is unique in this study as it decided not to continue as a managed care county in
SFY 2001. As a result, the process findings do not include Blueprints or scoring on the Fidelity
of Organizational Implementation Scales as data could not be collected to support these. Other
process findings are based on data collected in early 2000, rather than in 2001 like the other
counties.




William M. Mercer, Incorporated               166               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Process Findings
Process findings examine how Mesa County is functioning as a managed care entity. As
mentioned above, Blueprints and Fidelity of Organizational Implementation Scales are not
included due to the unavailability of data.

Progress in the Introduction of Managed Care Processes
Mesa County, when it was a managed care county, functioned as a managed care entity as it
assumed risk by entering into an MOU with the CDHS. Mesa was permitted to keep unspent
funds, but was required to cover overages that might occur. It was expected to perform at a
certain level on specific outcomes.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                 167                Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The main focus of Mesa County’s managed care initiative was on utilization review, the
expansion of its network, the development of performance indicators, and improvements in its
management information system. Of these three, the one that is the most developed is utilization
management.

Mesa County has a clearly identifiable utilization management program called the CORE Team.
The CORE Team staffs all new cases and authorizes all new treatment requests, including
requests for higher level care for families already in care. It coordinates all referrals to
out-of-home placements and performs ongoing utilization review. Residential treatment centers
and CPA placements that are across the mountains on the Front Range (over 200 miles away)
have concurrent utilization reviews performed by reviewers in that area to ensure treatment
efficacy and a focus upon Mesa County’s goals.

In regard to network capacity, Mesa County has not determined the capacity that it requires. As a
result, it is limited in developing a configuration of services that addresses the needs of children
and families now and in the future. Financial incentives for the development of services that
represent present best practices are difficult to implement without knowing the overall capacity
that is required.

Although Mesa County has not determined its required network capacity, it has moved to
address readily perceived gaps in its network. For example, it has started its own residential
treatment program to reduce the need to place children outside the county. It has also developed
a number of early intervention and prevention programs, including higher quality day care
standards, collaborative efforts to increase newborn birth weight, a professional nurse home
visiting program for all children born in Mesa County, dental access for Medicaid-eligible
children, and community education to promote healthy families. Another example is the use of
experienced parents for mentoring parents in the child welfare system.

Among the managed care counties, Mesa has the most highly developed client information
system. The system captures data from the point of intake to case closure. The system tracks
multiple outcomes related to services, such as overall effectiveness of individual services,
outcomes by provider and family by service, time between service authorizations and service
beginning, and attendance by caseworker, family, and providers present at reviews.
Approximately seven years of detailed data are available in Mesa County’s local database.

In regard to the other key elements of managed care listed above, Mesa County is still in the
development stage. Mesa needs to develop a clearly identifiable quality improvement system so
that it may continuously improve its services and those of its providers. It also might consider
linking performance indicators to financial incentives. Finally, although it has an advanced client
information system, the adequacy of its financial information system is not known.




William M. Mercer, Incorporated                168               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in Mesa County in the
nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in Mesa to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.

Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                        Figure 77
           Percentage of Children Who Experienced a Confirmed Recurrence of
                Abuse and/or Neglect within a 12-Month Period after being
                                Abused and/or Neglected

           10
                                       8.6
            9
            8
            7
            6
                                                        4.8
            5         4.3

            4                                                            2.9
            3
            2
            1
            0
                    SFY 97           SFY 98           SFY 99            SFY 00




William M. Mercer, Incorporated               169              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect

                                      Figure 78
        Confirmed Abuse and Neglect per 1,000 Children in the County Population

   20
   18
   16
   14
   12
   10
                 7                                     7.1
    8                              6.2                                   5.9
    6
    4
    2
    0
               SFY 97             SFY 98              SFY 99           SFY 00



Indicator 3: Referrals and Investigations

                                           Figure 79
                  Referrals of Alleged Abuse and Neglect per 1,000 Children
                                   in the County Population

          76                               75
          74
          72
                         70.3
          70
                                                                                68.1
          68
          66
                                                               63.9
          64
          62
          60
          58
                        FFY 97           FFY 98              FFY 99            FFY 00




William M. Mercer, Incorporated                 170              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                            Figure 80
                     Investigations of Alleged Abuse and Neglect per 1,000
                              Children in the County Population

       50
                   44.4
                                         42

       40                                                   35.7                 35


       30


       20


       10


        0
                  FFY 97              FFY 98               FFY 99              FFY 00



Indicator 4: Discharge to Reunification

It should be noted that this indicator does not directly measure reunification in that is represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.




William M. Mercer, Incorporated                 171               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
                                        Figure 81
                Percentage of Children Placed with Parents at Case Closure

       100
         90
         80
         70         61.5
                                                          55.6               52
         60                              53.6
         50
         40
         30
         20
         10
          0
                  SFY 97                SFY 98            SFY 99           SFY 00



Indicator 5: Children in Out-of-Home Placement


                                        Figure 82
               Children in Out-of-Home Placement at some Point during the
                 Fiscal Year per 1,000 Children in the County Population

        20
        18
        16
        14                                                       11.9         11.7
                 11.3
        12                                       10.5
                                  9.2
        10
         8
         6
         4
         2
         0
                SFY 96        SFY 97             SFY 98          SFY 99      SFY 00




William M. Mercer, Incorporated                  172               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Indicator 6: Length of Out-of-Home Stay

                                       Figure 83
             Out-of-Home Placement Length of Stay (in Months) for Closed Cases

        12
                                                  9.8              9.6
        10
                                                                                  7.9
         8
                                  6.6
                  5.8
         6

         4

         2

         0
                SFY 96         SFY 97            SFY 98         SFY 99        SFY 00



Indicator 7: Percentage in Positive Placements

                                            Figure 84
                        Percentage of Positive Placements at Case Closure
                         (Parents, Kin, Relatives, Adoption, Guardians)

       100
        90          79
                                         76
        80                                                 68                68
        70
        60
        50
        40
        30
        20
        10
         0
                  SFY 97                SFY 98            SFY 99            SFY 00




William M. Mercer, Incorporated                  173                Colorado Child Welfare Evaluation
                                                           Second Interim Implementation Status Report
Indicator 8: Number of Adoptions

                                        Figure 85
                   Adoptions per 1,000 Children in the County Population

         1.8

         1.6

         1.4

         1.2
                                                                       0.93
           1
                                             0.76
         0.8

         0.6

         0.4            0.28

         0.2

           0
                       FFY 98               FFY 99                    FFY 00

Indicator 9: Time to Adoption

                                      Figure 86
        Average Months from Termination of Parental Rights to Finalized Adoption

         30                                  27.2

         25                                                           22.5


         20
                        15
         15


         10


          5


          0
                      FFY 98                FFY 99                   FFY 00




William M. Mercer, Incorporated            174                Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
 Outcome Comparisons


                                           Table 9
     Mesa County SFY 2000 Outcome Results Compared to Other Counties, to Benchmarks,
                 and to the Criterion Reference Standards Set for SFY 2000
                                                       Non-
                                      Other                                                         Criterion
                           Mesa                      Managed
Outcome Indicator                  Managed Care                            Benchmark                Reference
                          County                    Care Large
                                     counties                                                       Standard
                                                     counties
                                                                     National: 11%              Acceptable: <10%
Indicator 1:
Confirmed re-abuse
                          2.9%        3.7%              3.0%         Comparison States: 12%     Outstanding: <2%
                                                                     (Utah)
Indicator 2:
                                                                     National: 11.8             Acceptable: <10
Confirmed abuse per        5.9         5.0               5.1         Comparison States: 8.8
1,000 child population                                                                          Outstanding: <4
Indicator 3a: Referrals
                                                                     National: 72.1             Acceptable: <45
of suspected abuse per    68.1         42.0              48          Comparison States: 53.4
1,000 child population                                                                          Outstanding: <40
Indicator 3b:
Investigations of                                                    National: 43.5             Acceptable: [not set]
suspected abuse per
                          35.0         23.9              31          Comparison States: 28.3    Outstanding: [not set]
1,000 child population
Indicator 4: Percent
                                                                     National: 59%*             Acceptable: >52%
discharged to             52%         44%               57.7%        Comparison States: 68%*
reunification
Indicator 5:
Children in out-of-                                                  National: 8.1**            Acceptable: <8
home care per 1,000
                          11.7         7.3              10.6         Comparison States: 5.6**   Outstanding: <4
child population
Indicator 6:
                                                                     National: 10.8             Acceptable: <10
LOS in out-of-home         7.9         8.4               6.2         Comparison States: 12.3
placement (in mos.)                                                                             Outstanding: <7
Indicator 7:
                                                                     National: 83%              Acceptable: >82%
Percent discharged to     68%         88%               83.8%        Comparison States: 90%
positive placements                                                                             Outstanding: >90%
Indicator 8: Adoptions
                                                                     National: .66              Acceptable: >.4
per 1,000 child            .93         .63               .66         Comparison States: .52
population                                                                                      Outstanding: >.9

Indicator 9:
                                                                                                Acceptable: <18
Time from termination                                                National: N/A
of parental rights to
                          22.5         11.2             20.6         Comparison States: N/A
                                                                                                 months
                                                                                                Outstanding: <9
adoption                                                                                         months
 * Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
** Benchmarks are for point in time (on 9/30) versus at any time in fiscal year (Indicator 5)




 William M. Mercer, Incorporated                  175                   Colorado Child Welfare Evaluation
                                                               Second Interim Implementation Status Report
Summary of Findings
Mesa County’s rate of positive placements was much lower than the rate for other large counties
in Colorado, and much lower than it had been prior to the managed care pilot. The rate was so
low that we will look into possible reasons, including data anomalies, that might explain the low
rate. At the same time, Mesa’s rate of placement with parents at case closure was high relative to
other managed care counties (although it was below the rate for non-managed care, large
counties combined). Like several of the other managed care counties, Mesa’s rate of placement
with parents has decreased over time.

Mesa’s rate of out-of-home placement per 1,000 children in the county population was much
higher than the rate for other managed care counties and somewhat higher than for non-managed
care counties. The average length of stay, however, was slightly lower than for other managed
care counties. This pattern is similar to that for Pueblo, which is the county most like Mesa in
terms of distance from the relatively more resource-rich major metropolitan area of Denver.

In examining the trends for the out-of-home placement indicators, the number of children per
1,000 in the population who were placed out of the home at some point during the fiscal year,
has increased slightly from before managed care (SFY 1996 to SFY 2000). The trend in average
length of stay in out-of-home placement, which appeared to be rising prior to managed care
taking effect, appears to have begun to reverse.

Mesa County had one of the lowest rates of recurrence of abuse and neglect of the large counties
in Colorado during SFY 2000. The overall rate of abuse and neglect per 1,000 children in the
population was slightly higher than the rate of other large counties in Colorado and much lower
than national and comparison state benchmarks. Mesa’s rates on both indicators of abuse and
neglect declined from SFY 1999 to SFY 2000.

Mesa County has an extraordinarily high rate of referrals and investigations in comparison to
other large counties in Colorado. Mesa has a relatively low rate of confirmed cases in
relationship to the number of cases investigated. It is not clear whether Mesa has more stringent
criteria for confirming abuse and neglect than other counties or if it has broader criteria for
referring and investigating alleged abuse and neglect. With regard to trends, Mesa’s rate of
referrals has dropped slightly from SFY 1997 to SFY 2000 as has its rates of investigations and
confirmed cases.

Mesa’s rate of referrals of abuse and neglect per 1,000 children in the population was much
higher than the “acceptable” range of referral rates. This finding may signify a limitation of the
criterion reference standard for this outcome indicator. It is possible that Mesa does not have a
significantly higher rate of actual abuse and neglect in the population but has a more aggressive
approach to reporting potential abuse and neglect. Further investigation of this indicator is
needed in year three of the evaluation in order to examine this finding.

Mesa has dramatically increased its rate of adoption from FFY 1998 to FFY 2000 from .28
adoptions per 1,000 children in the population to .93, an increase of 232 percent. Mesa’s rate of
adoption is higher than all other large counties in Colorado, except for El Paso.


William M. Mercer, Incorporated                176               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as residential treatment centers, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher cost care as well as trends over several years. The results of these
analyses are shown in this section for Mesa County by itself and in a later section for all of the
managed care counties together.

Figures 87 and 88 compare Mesa County’s allocation at the beginning of SFY 2000 to its
expenditures at the end of the year. Figure 87 provides a comparison of the two main child
welfare funding streams, CORE services and the child welfare block. Figure 88 provides detail
regarding the funding streams that compose the child welfare block.

Figure 87 shows that Mesa County had unspent funds in SFY 2000. Figure 88 shows a slight
deficit in CORE services that was more than offset by unspent funds in the child welfare block.
Mesa County was the only managed care county that spent less than its allocation at the end of
the year in SFY 2000.

                                       Figure 87
             CORE Services and Child Welfare Block Allocation vs. Expenditure*,
                                 Mesa County, SFY 2000

      $15,000,000


      $10,000,000                                                   $8,331,396   $7,909,465
                                                 $(103,299)




                                                                                                   $421,931




       $5,000,000
                      $848,010     $951,309
             $-
                                 CORE Services                               Child Welfare Block
      $(5,000,000)


     $(10,000,000)


          * Before Year End Close Out Process                   Allocation       Expenditure                  Difference




William M. Mercer, Incorporated                               177                Colorado Child Welfare Evaluation
                                                                        Second Interim Implementation Status Report
                                                            Figure 88
                                         Child Welfare Block Allocation vs. Expenditure*,
                                                     Mesa County, SFY 2000




                            $8,331,396
       $14,000,000




                                           $7,909,465
       $12,000,000




                                                                      $5,984,611
                                                         $5,666,771
       $10,000,000
        $8,000,000




                                                                                         $1,700,512

                                                                                                      $1,001,899
        $6,000,000




                                                                                                                                                 $627,331

                                                                                                                                                            $627,331
                                                                                                                      $336,782

                                                                                                                                 $295,623
        $4,000,000
        $2,000,000
                $0
                       Total CW Block                   Total 80/20                RTC Medicaid                    CHRP Medicaid            100% Admin.



   * Excludes Medicaid Fund and before Year end Close-Out Process                                                        Allocation                         Expenditure



Figures 89 and 90 provide a picture of how Mesa County spent its child welfare block in
SFY 2000. Figure 89 shows the percentage of funds expended by budget item while Figure 90
shows the actual amount expended. Mesa County was among the lower tier of managed care
counties in the proportion of expenditures for RTC. Of particular interest is a comparison of
Figure 89 with the same figure (108) for Pueblo County as the proportions in each item are very
similar.
                                        Figure 89
            Percentages of Child Welfare Block Expenditure* by Budget Items,
                                 Mesa County, SFY 2000


                                                                                                                                 Out of Home Placement
                100% Admin.                                                                                                          Services 32%
                    8%
                                                                                                                                 Subsidized Adoptions
             CHRP Medicaid                                                                                                                8%
                 4%                                                                Other
                                                                                                                                 Case Services
                                                                                   75%
                                                                                                                                      2%
               RTC Medicaid
                                                                                                                                 Special Circumstances Child
                   13%
                                                                                                                                         Care (SCCC)
                                                                                                                                              1%

                                                                                                                                 80/20 Admin.
                                                                                                                                     32%


                     * IncludesMedicaid Fund and before Year End Close Out Process




William M. Mercer, Incorporated                                                    178                              Colorado Child Welfare Evaluation
                                                                                                           Second Interim Implementation Status Report
                                                 Figure 90
                             Child Welfare Block Expenditure* by Budget Items,
                                           Mesa County, SFY 2000


    $4,000,000


    $3,000,000                                                          $2,580,143                                                                  $2,536,982

    $2,000,000

                 $1,001,899
    $1,000,000                                         $627,331                           $644,862
                                    $295,623                                                                 $143,176         $79,448
            $0
                 R T C M edicaid   C H R P M edicaid   100% A d m in.   Out o f H o m e   Subsidized        Case Services        Special            80/20 A dmin.
                                                                         P lacement       A do ptio ns                      C ircum s t a n c e s
                                                                          Services                                             Child Care
                                                                                                                                 (SCCC)
   * Before Year End Close Out Process



Figures 91–93 illustrate the trends in out-of-home expenditures in Mesa County from SFY 1998
through SFY 2000. Of note is the decrease in the use of family foster care over time and an
increase in the use of specialized group care.

                                              Figure 91
                         Percentages of Out-of-Home Placement Expenditure,
                                      Mesa County, SFY 1998


                    Specialized Group
                      Care 11.5%                                                                                  RCCF
                                                                                                                  9.0%
                                                                                                 RTC
                                                                           Excludes RTC          6.9%
                                                                             Medicaid                        Relative Foster Care
                                                                                                                     2.5%
                                                                                                         All Other *
                                                                                                            3.4%
                                          Family Foster Care
                                                66.7%
                                                                                                   * Include independent living, transitional
                                                                                                   program, home-based program, receiving
                                                                                                   home, and shelter care.




William M. Mercer, Incorporated                                           179                      Colorado Child Welfare Evaluation
                                                                                          Second Interim Implementation Status Report
                                         Figure 92
                    Percentages of Out-of-Home Placement Expenditure,
                                 Mesa County, SFY 1999


                                                                       RCCF
          Specialized Group Care                                       5.8%
                   19.0%
                                                                      RTC
                                                    Excludes RTC      9.9%
                                                      Medicaid
                                                                             Relative Foster Care
                                                                                     2.6%
                                                                              All Other *
                                                                                 2.5%
                                   Family Foster Care
                                         60.2%                         * Include independent living, transitional
                                                                       program, home-based program, receiving home,
                                                                       and shelter care.




                                         Figure 93
                    Percentages of Out-of-Home Placement Expenditure,
                                  Mesa County, SFY 2000
           Specialized Group                            RCCF
             Care 20.0%                                 7.7%

                                                                      RTC
                                                                      7.5%
                                               Excludes R T C
                                                 M edicaid
                                                                   Relative Foster Care
                                                                           3.2%
                          Family Foster Care                       All Other *
                                59.0%                                 2.6%

                                                        * Include independent living, transitional
                                                        program, home-based program, receiving
                                                        home, and shelter care.




William M. Mercer, Incorporated                          180                   Colorado Child Welfare Evaluation
                                                                      Second Interim Implementation Status Report
Figures 94 and 95 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 56 shows that Mesa County has
contracted out most of its family foster care and group homes to CPAs. Figure 95 shows
that family foster care homes are the primary contracted service.

                                         Figure 94
              Percentages and Amounts of Expenditure for Family Foster Care and
                      Group Home Services delivered by County vs. CPA,
                             Mesa County, SFY 1998 to SFY 2000
      100%
       90%
       80%
       70%                                            $1,334,119                       $1,407,716
                       $1,536,276
       60%
       50%
       40%
       30%
       20%                                             $782,894                         $952,329
                        $615,408
       10%
        0%
                        SFY1998                        SFY1999                          SFY2000

     County Family Foster Care Homes and Group Homes            CPA Family Foster Care Homes and Group Homes




                                          Figure 95
              Percentages of Expenditure for Family Foster Care and Group Home
                          Services delivered by County and/or CPA,
                             Mesa County, SFY 1998 to SFY 2000
       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                       SFY1999                          SFY2000


   CPA Family Foster Care Homes     CPA Group Homes         County Family Foster Care Homes    County Group Homes




William M. Mercer, Incorporated                       181                   Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Pueblo County Findings
Pueblo County is located in the south central part of Colorado, 100 miles south of Denver. It had
an estimated child population of 34,810 in 2000. Pueblo was accepted as a managed care county
in SFY 1999.




William M. Mercer, Incorporated                182              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Process Findings
Process findings first examine how Pueblo County is functioning as a managed care entity. Flow
charts (Blueprints) are then presented to show the flow of children and their families and funds
through the county’s child welfare system. Finally, the county is scored on the extent of its
fidelity to managed care principles using Fidelity of Organizational Implementation Scales.

Progress in the Introduction of Managed Care Processes
Pueblo County is functioning as a managed care entity as it is assuming risk by entering into an
MOU with the CDHS. In its original MOU, Pueblo County was permitted to keep unspent funds
but was required to cover overages that might occur. It is expected to perform at a certain level
on specific outcomes.

In 2001, the CDHS offered counties a choice between keeping any of their own unspent funds or
being eligible for participation in the distribution of unspent funds from other counties at the end
of the year. Pueblo County was the only managed care county to select the option to keep its own
unspent funds.

There are key processes that a managed care entity should have in place to effectively bear risk.
These include:

§     Utilization management. This is a formal process for ensuring that the right service is offered
    in the right amount at the right time. Utilization management is an especially useful process
    for controlling the cost of the most expensive services, such as residential treatment. In
    managed care programs, it is usually one of the first processes introduced when a program is
    implemented.
§     Quality improvement. This is a formal process in which specific areas for improvement are
    identified, data are collected, solutions are developed and implemented, and data are collected
    again to determine if solutions are working and to refine them as needed.
§    Adequate provider network. An adequate network contains services that are accessible, are
    appropriate to the needs of children and their families, are of demonstrated effectiveness, and
    are available in adequate quantity.
§    Performance indicators. Performance indicators encompass measures of process and
    outcomes. They provide a basis for determining if the service system is effective and for the
    use of financial incentives.
§    Provider financial incentives. The methods by which providers are paid encourage
    performance that supports the mission of the county child welfare system. Financial incentives
    may range from bonuses to the use of case rates.
§    Appropriate management information system. An information system is available that
    provides the appropriate and timely reports required by the county to properly manage the
    system of care.




William M. Mercer, Incorporated                  183               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The main focus of Pueblo County’s managed care initiative has been on the use of utilization
management and network development. Of these two, utilization management is the most
developed.

The county has implemented the Utilization Review Team, which manages out-of-home
placement referrals and identifies community-based alternatives through multi-agency planning
and a treatment authorization process. All out-of-home referrals are staffed by this team and any
placement has to be authorized in order to be paid. All discharges from out-of-home placement
also must be approved by the Utilization Review Team.

In regard to network capacity, Pueblo County has not determined the capacity that it requires. As
a result, it is limited in developing a configuration of services that addresses the needs of
children and families now and in the future. Financial incentives for the development of services
that represent present best practices are also difficult to implement without knowing the overall
capacity that is required.

Although Pueblo County has not determined its required network capacity, it has moved to
address readily perceived gaps in its network. For example, Pueblo has developed day treatment
programming, a sexual abuse program, and specialized foster care placements. Pueblo is
presently working to develop more county foster homes to reduce the cost of foster care
placements.

In the area of financial incentives, Pueblo County has not entered into relationships with
providers that involve financial incentives other than passing on the risk for CPA treatment costs
to its local MHASA. Pueblo has not yet developed a formal quality improvement program and
has not introduced performance indicators for providers.

Like other counties, Pueblo County has relied on state information systems that most counties do
not view as providing sufficiently detailed or timely information to manage utilization or
otherwise fulfill the data requirements for administering a managed care program. The CDHS
has recently implemented a new system, TRAILS, which has the potential to yield more useful
and timely information.




William M. Mercer, Incorporated                184               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Process Flow
To illustrate the flow of processes involving people and funds in Pueblo County, we have
constructed eight Blueprints. These show:
1. the flow of persons through the county in SFY 1998, prior to the initiation of managed care
   reforms;
2. the overall flow of persons through the county in SFY 2001, midway through the third year
   of managed care reforms;
3. additional detail on the flow of persons through the county department;
4. additional detail on the flow of persons through the county’s provider network;
5. the flow of funds through the county in SFY 1998, prior to the initiation of managed care
   reforms;
6. the overall flow of funds through the county in SFY 2001, midway through the third year of
   managed care reforms;
7. additional detail on the flow of funds from the state to the county, including financial
   reporting; and
8. additional detail on the flow of funds through the county’s provider network, including
   risk-sharing arrangements.

We have used six colors in the Blueprints:
1. black is used for the organizations involved in paying for, managing, and providing child
   welfare services;
2. red is used for the unmanaged flow of people through the system and includes county
   populations, populations in need, referral groups, movement through different teams and
   areas within the county Child Welfare Department, and referrals to community providers;
3. green depicts child welfare funds that flow directly to the county child welfare department, as
   well as the state and federal funds that flow through the CDHS to the county Child Welfare
   Department;
4. blue depicts the flow of non-child welfare funds that are critical to providing needed care to
   recipients of child welfare services;
5. purple depicts the managed flow of both persons and funds through the county child welfare
   managed care pilot initiative; and
6. yellow depicts the court system in each county.




William M. Mercer, Incorporated                 185              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
In the Blueprints, the flow of people through the system is depicted in a left to right fashion,
beginning with the entire population potentially served by the child welfare system, narrowing to
the group of persons actually served, and following these people through the organizations and
services they would typically encounter as they pass through the system. The flow of funds,
however, goes from the top to the bottom of each chart. The drawing traces the flow of funds
from the payers that fund the system, to the administrative entities that allocate and account for
these funds, to the county child-serving agencies and providers that ultimately spend and manage
them.

A set of symbols is also used. They are illustrated in the following diagram.

Organizational Symbols:
               = Federal, state or county payors
               = State agencies
               = Local agencies or providers
               = Local entities that bears risk

Risk Symbols:
  S   = Shared risk of financial utilization
  F   = Full risk of financial utilization

Out-of-Home Placement Symbol:
  $            = Most expensive out-of-home
                   placements (RTC, CPA)
Financial Reporting Symbol:
   Financial
  Reporting
  by County    = Locus of financial reporting




William M. Mercer, Incorporated                    186            Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Flow of People—Pre-Managed Care—Spring 1998
The Blueprint depicted on page 189 (entitled “Pueblo County–Spring 1998–People Flow”)
shows the flow of people through Pueblo County’s child welfare system in the Spring of 1998,
just prior to the initiation of their managed care pilot later that summer. Beginning on the far left,
the overall county population breaks down into the smaller group of persons who, at any point in
time, were in need of child welfare services. From that group, two types of referrals were
typically seen by the county: youth in conflict (including youth with delinquency problems) and
children with reported instances of abuse or neglect.

All calls into the Pueblo County DSS in 1998 passed through a minimal screening process in
which clerical staff briefly tracked calls and passed them on to a supervisor for assignment or
referral back to the community. All calls meeting criteria were referred directly to the Intake
Unit. Additionally, callers with basic needs were often referred to the county’s TANF program.
The Intake Unit also investigated all allegations of abuse regarding children under care at the
Colorado Mental Health Institute at Pueblo, regardless of the county in which the child resided.

Even prior to the initiation of managed care reforms, Pueblo County had a variety of levels of
programs through which youth and families could receive services in 1998. Ongoing teams
included an Adolescent Unit, Child Protection Services, and an Incest Treatment Team.
Specialized teams included two teams providing intensive services separately for adolescents and
children, an Adoption Unit, a Permanency Planning Unit, and SB 94 services through the PRIDE
program (which provided alternatives to detention).

Many children, youth, and families served by the county required additional services provided by
agencies in the community. Workers on ongoing and specialized teams were able to directly
access expensive out-of-home placements in RTCs, RCCFs, CPAs, and county-administered
foster homes, as well as outpatient services through Spanish Peaks CMHC, outpatient substance
abuse services, family preservation services through the Pueblo Youth Services Bureau, a
visiting nurse program through La Familia Fuerte, and services through other specialty outpatient
providers. Children needing inpatient services had to go first through the CMHC, which
managed all indigent and Medicaid-funded mental health benefits. While some high need or
multi-agency involved youth accessed multi-agency planning from time-to-time, the majority of
external referral decisions were not centrally coordinated.

A critical factor for any involuntary case served by the county is the court system. Individual
workers presented recommendations for services to the court and the court implemented them as
it saw fit. While the Blueprint highlights the role of the court at the point where it impacted
decisions regarding referrals to out-of-home placements versus other community services, the
court’s involvement with involuntary cases was much broader, beginning at or prior to intake,
and extending throughout each family’s final disposition.




William M. Mercer, Incorporated                  187               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
William M. Mercer, Incorporated   188            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of People—Post-Managed Care—Spring 2001
In the third year of its implementation of managed care reforms, the flow of persons through the
Pueblo County child welfare system appears changed. A review of the Base Blueprint for Spring
2001 (see the Blueprint depicted on page 191 entitled “Pueblo County—Spring 2001—People
Flow”) shows the same basic population served, but reorganization within Pueblo DSS and its
provider network. To review these changes with greater specificity, two detailed Blueprints were
created for this year: a Department Internal Detail Blueprint and a Provider Network Detail
Blueprint.




William M. Mercer, Incorporated               189              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
William M. Mercer, Incorporated   190            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Department Internal Detail. Looking at the Department Internal Detail Blueprint (see the
Blueprint depicted on page 193 entitled “Pueblo County—Spring 2001—Department Internal
Detail”), it can be seen that the county population still breaks down on the left into a smaller
group of persons in need, who, in turn, sort into the two primary groups of youth and families
served by the county: youth in conflict (including youth with delinquency concerns) and children
with reports of abuse or neglect. Within the county, however, the flow is reorganized. The
screening process has been augmented and re-tooled into the Screening and Prevention Unit. Key
changes include restaffing with senior caseworkers, adding the option to provide time-limited
prevention services (that often result in families being able to have their concerns resolved
without ongoing services), and providing training to support a more assessment-oriented
approach in general. Families needing additional services are referred to the Intake Supervisor
for referrals to TANF, Intake, Ongoing, and Specialized Teams, largely the same as before. The
county also participates in joint projects with other child- and family-serving agencies. For
example, the local substance abuse provider, Crossroads, places substance abuse clinicians on
site at DSS. Overall, system level multi-agency planning with other agencies is conducted
informally through existing relationships and ad hoc meetings. This is seen as the best fit by
DSS, given Pueblo’s size and culture.




William M. Mercer, Incorporated               191              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
William M. Mercer, Incorporated   192            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Provider Network Detail—People Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 196 entitled “Pueblo County—Spring 2001—Provider
Network Detail—People Flow”), the primary change in person flow can be seen in the
establishment of the Utilization Review Team (URT). The URT was developed to centrally
manage out-of-home placement referrals and identify community-based alternatives through
multi-agency planning and a treatment authorization process. All out-of-home referrals are
staffed by the URT, and any placement has to be authorized in order to be paid. Many referrals
are diverted to either lower-level out-of-home placements (e.g., kinship homes or county-
administered foster homes) or to community-based alternatives, such as community mental
health center services.

Once in an out-of-home placement, the URT relies on supervisors to conduct concurrent review
of placement appropriateness and refer needed cases to the URT. The review process has a goal
of trying to ensure that placements maintain a focus on the county’s placement goals.

In 2000–2001, the URT began reviewing all cases, in addition to its initial role reviewing all out-
of-home placement referrals. Ensuring that cases are not inappropriately discharged or closed
adds an important quality monitoring role to the URT. Also, additional representation on the
Utilization Review Team by a drug and alcohol provider, the school district, and probation has
been pursued.

Pueblo County reports that RTC placements increased this past year and that the URT has
increased its scrutiny and planning for these cases. In addition, increased coordination with
magistrates to educate them regarding placement options and increased supervision have been
implemented to lower placement levels.

Another major development for Pueblo County in 2000–2001 has been the development of a new
set of services called Supportive Ongoing Services (SOS). This service has been developed for
families with high rates of repeat openings within the system and is provided under contract by
La Familia Fuerte. Under the new program, a senior caseworker tracks approximately 50–60
closed cases. These families agree to receive follow-up monitoring and assistance services
through La Familia Fuerte after their formal case with the county is closed. The goal of the
program is to provide tertiary prevention services to help these families from becoming
re-involved with Pueblo DSS.




William M. Mercer, Incorporated                193               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Other changes in 2000–2001 include:
§    Foster care recruitment resources have been increased by assigning two new foster care
    workers and an assistant to recruit county foster homes. These positions were redirected from
    ongoing services with the goal of reducing dependence on CPAs.
§    Like other counties, Pueblo County has experienced higher rates of staff turnover in the past
    year. County staff see this as primarily related to the competition for experienced and able
    human service workers in Colorado.
§     Staff caseload ratios for the PRIDE project in collaboration with SB 94 have been increased
    from the 1:10 level, although the service still offers an array of enhanced services.




William M. Mercer, Incorporated                 194               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   195            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Pre-Managed Care—Spring 1998
The Blueprint depicted on page 199 (entitled “Pueblo County—Spring 1998—Flow of Funds”)
shows the flow of funds through the Pueblo County child welfare system prior to the initiation of
the managed care pilot in mid-1998. In early 1998, funds from various federal sources (Title IV
and XX) and the state general fund were combined by the CDHS and passed on to Pueblo
County. The bulk of these funds passed through a large block that combined prior line-item
funding for county administration (largely used for staffing), out-of-home placements, and other
smaller funding lines (e.g., Special Circumstances Child Care, subsidized adoption funds). The
other primary funding source for child welfare services was the CORE services funding stream,
which was not integrated into the overall child welfare block in order to preserve separate
accountability for these funds for alternatives to out-of-home care, such as family preservation,
outpatient substance abuse and mental health services, and day treatment. Pueblo County
received other specialized state funding, including EPP funds, but these other sources of funding
are not depicted.

While some CORE funds were provided entirely by the state, most required a 20 percent match
by the county. The overall child welfare block also required a match derived through an equation
combining the historical 100 percent, 80 percent, and other partially state-funded. In addition, SB
97-218 had shifted the responsibility for excess costs in all Colorado counties to the county.
Even though Pueblo County was not a managed care county at this time, Pueblo DSS and its
county Commissioners had the full downside risk for any cost overages that exceeded the state’s
ability to shift unspent funds from other areas of the state, or provide excess Title IV-E
distributions. Because it was not yet a managed care county, Pueblo was not able to keep any
unspent funds of its own. As a result, this created an incentive for Pueblo County to spend all of
its child welfare block funding, but not to go over the amount, creating a partial managed care
incentive that was less clear in its effects that the full risk shift for managed care counties.

Pueblo County, in turn, contracted with various community agencies for services. In addition to
funding streams flowing through DSS, Pueblo County’s child welfare system also depended on
services paid for by other agency funding. These included:
§     Medicaid (Title XIX) funding paid through the HCPF to RTC providers for the treatment
    component of the services provided in these facilities—The county DSS directly pays for only
    room and board. Treatment costs are paid directly from the state Department of Human
    Services to HCPF but are then charged back to the county at year end against its allocation.
    The county has no control over the rates paid to providers since they are determined through a
    cost-based approach within available appropriations.




William M. Mercer, Incorporated                 196               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
§    Medicaid (Title XIX) funding for mental health services transferred through the HCPF to the
    Mental Health arm of the DHS Office of Health and Rehabilitation Services—In Pueblo
    County, funding for all Medicaid inpatient and outpatient mental health services has been
    capitated and, as a result, flows through the entity awarded the MHASA contract for that area
    of the state, SyCare-Options Colorado Health Networks (CHN).
§    DYC SB 94 funding for preventing detention placements for which some youth served by DSS
    are eligible—Services funded may be similar to CORE services.
§    ADAD funding through the DHS Office of Health and Rehabilitation Services—In Pueblo
    County, these funds are managed by Signal, the managed services organization (MSO) for
    substance abuse services.

As can be seen in the Blueprint, child welfare funds were the least managed of the primary
funding streams for expensive out-of-home placements such as inpatient, RTC, and CPA. In
1998, inpatient and other out-of-home placements through mental health were subject to
managed care controls and RTC placements through DYC were subject to the placement
restrictions of the juvenile justice system.




William M. Mercer, Incorporated                 197              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
William M. Mercer, Incorporated   198            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Flow of Funds—Post-Managed Care—Spring 2001
The Blueprint depicted on page 201 (entitled “Pueblo County—Spring 2001—Funding Flow”)
shows the flow of funds under the managed care pilot as they exist in the Spring of 2001. Since
most of the major changes in child welfare funding streams happened in 1997, counties (such as
Pueblo), that initiated their managed care pilots in the Summer and Fall of 1998 experienced no
changes in funding streams other than (1) the ability of the county to now retain unspent funds
from its child welfare funding block and (2) any additional changes initiated at the county level.
Pueblo is now able to retain any unspent funds from the seven combined funding streams, thus,
taking on the upside risk for the services it provides. To do this, Pueblo has to allow
non-managed care counties to use unspent funds from other areas of the state to cover their
overages before being eligible for these. The county is still eligible for excess Title IV-E
distributions it helps earn.

In the first two years of operating under a capped allocation, the CDHS returned unspent general
funds to the legislature. During that period of time, as well as during the third year, funding for
the child welfare block was based on “population increase.”

During the third year, SFY 2000, all Colorado counties collectively overspent their budgets by
approximately $20 million. The CDHS requested and received a $14 million supplemental
appropriation in SFY 2001, based on projected caseload growth. Because counties overexpended
so significantly in SFY 2000, the Child Welfare Allocation Committee approved allowing
managed care counties a choice to either participate in the statewide surplus distribution or retain
their own unspent funds.

Pueblo County has chosen the MOU option where they have the ability to retain unspent funds
and forego participation in unspent funds distributions from the state. Pueblo County has chosen
to be eligible for the benefits of upside risk rather than have the state partially protect it from
downside risk. This can be seen in both the base Blueprint and the State Funding Detail
Blueprint for the county.




William M. Mercer, Incorporated                 199               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   200            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
State Funding Detail. Looking at the State Funding Detail Blueprint (see the Blueprint depicted
on page 203 entitled “Pueblo County—Spring 2001—State Funding Detail”), another systemic
issue related to the cost overruns can be seen. This is the area of financial reporting. All counties
track their internal costs for staff and overhead. Pueblo County currently relies on the state for
their primary data on utilization for most purchased services. The county generally tracks only
internal department costs, not purchased services. Overall, Pueblo County staff reports
satisfaction with the accuracy and timeliness of state reports.

Only one additional funding stream change is shown. Pueblo DSS now earns a federal Medicaid
(Title XIX) match on the treatment portion of its CPA costs through a CPA Medicaid transfer
that sent these funds through the state HCPF to SyCare-Options CHN, to the CPA providers.
Pueblo County takes advantage of this in order to receive a roughly 50 percent federal match on
the treatment component of the care provided in CPAs. Part of the increased revenue realized
through this change has been used to increase the rates paid to county-administered foster care
homes to help the county better compete with CPAs to attract foster parents. The CPA treatment
contract specifies that SyCare-Options CHN is fully at risk (both downside and upside) for CPA
treatment costs. No other contractually-based provider incentives are used.




William M. Mercer, Incorporated                  201              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
William M. Mercer, Incorporated   202            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Provider Network Detail—Funding Flow. Looking at the Provider Network Detail Blueprint
(see the Blueprint depicted on page 205 entitled “Pueblo County—Spring 2001—Provider
Network Detail—Funding Detail”), it can be seen that the other major funding development in
Pueblo County has been the development of a new set of services called SOS. Established in
2000, it was developed using first year unspent funds and is delivered under contract with La
Familia Fuerte.




William M. Mercer, Incorporated               203              Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
William M. Mercer, Incorporated   204            Colorado Child Welfare Evaluation
                                        Second Interim Implementation Status Report
Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. When counties implement managed care in a way that is consistent with
such expectations, we call that “process fidelity.”

Fidelity is scored in each of four domains—funding flexibility, interagency coordination,
integration of informal networks and non-traditional services, and implementation of early
intervention and prevention programs—on a four-point scale, from one to four. Once a score is
calculated, it is placed within one of four fidelity categories—early, emerging, maturing, and
mature stages of fidelity.

Funding Flexibility
Pueblo County was rated in the Emerging Stage of fidelity in this domain. The county has not yet
added managed care techniques related to financing of services to its pilot, other than passing on
financial risk for CPA treatment funds through its MHASA CPA pass-through. As do all
Colorado counties, Pueblo also makes available flexible funds to families.

Interagency Coordination
In this domain, Pueblo County was given a rating of Maturing Stage fidelity. Pueblo has
developed formal, contractual relationships with some child-serving agencies and has developed
service planning processes and administrative coordination processes with child-serving agencies
as well. The number of agencies with which they have collaborated, however, has been more
limited than in most other counties. Pueblo County appears to have some capacity to integrate
data with other child serving agencies in the county. Pueblo reports that most of its relationships
with other agencies operate on an informal basis and that they believe that this works well.

Integration of Informal Networks and Non-Traditional Services
Pueblo County’s fidelity score placed it in the low end of the Emerging Stage of fidelity in this
domain. Pueblo County has developed relationships with informal networks of support in the
community but the extent of the integration of informal networks and non-traditional services
appears to be somewhat limited. They utilize paraprofessionals on a limited basis. Informal
networks of support were not reported to be involved in service planning nor are there
organizations outside of the formal child-serving system that are involved with the county in
funding services.




William M. Mercer, Incorporated                 205              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Implementation of Early Intervention and Prevention Programs
Pueblo County was found to be performing at the Emerging Stage of fidelity in this domain. The
county currently is delivering limited prevention programming and did not report any new
developments in the area of prevention and early intervention.




William M. Mercer, Incorporated              206              Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
Outcome Findings
Outcome findings are divided into three sections. The first shows trends in Pueblo County in the
nine outcome indicators in the past several state or federal fiscal years. The second section
compares findings on the outcome indicators in SFY 2000 in Pueblo to other managed care
counties, to the four comparison non-managed care counties, to national and comparison state
benchmarks, and to criterion reference standards. The final section provides a narrative summary
of the findings.

Annual Outcome Trends
Indicator 1: Recurrence of Abuse and Neglect

                                           Figure 96
              Percentage of Children Who Experienced a Confirmed Recurrence of
                   Abuse and/or Neglect within a 12-Month Period after being
                                   Abused and/or Neglected

          8
          7
          6
          5                              4.2
          4
          3
                                                            1.4                 1.6
          2
                      0.6
          1
          0
                    SFY 97             SFY 98              SFY 99              SFY 00




William M. Mercer, Incorporated                 207            Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect

                                      Figure 97
        Confirmed Abuse and Neglect per 1,000 Children in the County Population

  10
    9
    8
    7
    6                               5.4
                4.6
    5                                                 4.1                  3.7
    4
    3
    2
    1
    0
              SFY 97              SFY 98            SFY 99               SFY 00



Indicator 3: Referrals and Investigations

                                          Figure 98
                 Referrals of Alleged Abuse and Neglect per 1,000 Children
                                  in the County Population

   80
   70
   60                              55.1
                                                    46.4
   50           43
   40                                                                   31.3
   30
   20
   10
    0
              FFY 97              FFY 98           FFY 99               FFY 00




William M. Mercer, Incorporated             208             Colorado Child Welfare Evaluation
                                                   Second Interim Implementation Status Report
                                             Figure 99
                      Investigations of Alleged Abuse and Neglect per 1,000
                               Children in the County Population

    50
                                     41.9                 39.6
    40           34
                                                                                28
    30


    20


    10


     0
               FFY 97               FFY 98               FFY 99               FFY 00




Indicator 4: Discharge to Reunification
It should be noted that this indicator does not directly measure reunification in that it represents,
among children who had ever been in out-of-home placement, the percentage who were placed
with parents at case closure. This does not represent reunification, per se, in that there may be
youth who were placed with kin, guardians, and others whom they were also living with prior to
out-of-home placement. This indicator may then underestimate reunification because of the way
it is defined. We could not measure reunification directly because the data do not indicate whom
the children were living with immediately prior to their most recent out-of-home placement.




William M. Mercer, Incorporated                  209              Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
                                        Figure 100
                Percentage of Children placed with Parents at Case Closure

    100
     90
     80
                                         67.4
     70          59.5                                       59                 59.6
     60
     50
     40
     30
     20
     10
      0
                SFY 97                  SFY 98            SFY 99              SFY 00




Indicator 5: Children in Out-of-Home Placement

                                       Figure 101
               Children in Out-of-Home Placement at some Point during the
                Fiscal Year, per 1,000 Children in the County Population
                                                                   18              18.8
     20
     18
     16
     14        12.7
     12                           9.6             9.7
     10
      8
      6
      4
      2
      0
             SFY 96           SFY 97             SFY 98          SFY 99          SFY 00




William M. Mercer, Incorporated                  210               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Indicator 6: Length of Out-of-Home Stay

                                    Figure 102
           Out-of-Home Placement Length of Stay (in Months) for Closed Cases

   16

   14
                                             11.6
   12
                              9.8
   10
                                                               7.7           7.9
    8        6.9

    6

    4

    2

    0
           SFY 96           SFY 97           SFY 98          SFY 99        SFY 00




Indicator 7: Percentage in Positive Placements

                                           Figure 103
                       Percentage of Positive Placements at Case Closure
                        (Parents, Kin, Relatives, Adoption, Guardians)

   100                                                                    85.8
                                     85.3             84.9
    90          80.4
    80
    70
    60
    50
    40
    30
    20
    10
     0
              SFY 97                SFY 98            SFY 99             SFY 00




William M. Mercer, Incorporated                211             Colorado Child Welfare Evaluation
                                                      Second Interim Implementation Status Report
Indicator 8: Number of Adoptions

                                        Figure 104
                   Adoptions per 1,000 Children in the County Population

   1.8
   1.6
   1.4
                    1.14
   1.2
    1
                                             0.79
   0.8                                                                    0.61
   0.6
   0.4
   0.2
    0
                   FFY 98                   FFY 99                       FFY 00




Indicator 9: Time to Adoption

                                       Figure 105
          Average Months from Termination of Parental Rights to Finalized Adoption




    20                                                                    16.2
    15               12.2                    12.6

    10

      5

      0
                    FFY 98                  FFY 99                       FFY 00




William M. Mercer, Incorporated             212               Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
     Outcome Comparisons
                                         Table 10
 Pueblo County SFY 2000 Outcome Results Compared to Other Counties, to Benchmarks, and
                  to the Criterion Reference Standards Set for SFY 2000
                                                    Non-
                                        Other
                                                  Managed                                       Criterion
                          Pueblo       Managed
 Outcome Indicator                                  Care                Benchmark               Reference
                          County         Care
                                                   Large                                        Standard
                                       counties
                                                  counties
                                                                 National: 11%
Indicator 1: Confirmed                                                                      Acceptable: <10%
re-abuse
                          1.6%          3.8%        3.0%         Comparison States: 12%
                                                                                            Outstanding: <2%
                                                                 (Utah)
Indicator 2: Confirmed
                                                                 National: 11.8             Acceptable: <10
abuse per 1,000 child      3.7          5.1          5.1         Comparison States: 8.8
population                                                                                  Outstanding: <4
Indicator 3a: Referrals
                                                                 National: 72.1             Acceptable: <45
of suspected abuse per    31.3          44.4            48       Comparison States: 53.4
1,000 child population                                                                      Outstanding: <40
Indicator 3b:
Investigations of                                                National: 43.5             Acceptable: [not set]
suspected abuse per
                          28.0          24.3            31       Comparison States: 28.3    Outstanding: [not set]
1,000 child population
Indicator 4: Percent
                                                                 National: 59%*             Acceptable:
discharged to             59.6%         42%        57.7%         Comparison States: 68%*    >52%
reunification
Indicator 5:
Children in out-of-home                                          National: 8.1**            Acceptable: <8
care per 1,000 child
                          18.8          6.7         10.6         Comparison States: 5.6**   Outstanding: <4
population
Indicator 6:                                                                                Acceptable: <10
                                                                 National: 10.8
LOS in out-of-home         7.9          8.4          6.2         Comparison States: 12.3    Outstanding: <7
placement (in mos.)
Indicator 7:                                                                                Acceptable: >82%
                                                                 National: 83%
Percent discharged to     85.8%         87%        83.8%         Comparison States: 90%     Outstanding: >90%
positive placements
Indicator 8: Adoptions                                                                      Acceptable: >.4
                                                                 National: .66
per 1,000 child            .61          .65          .66         Comparison States: .52     Outstanding: >.9
population
Indicator 9:                                                                                Acceptable:
Time from termination                                            National: N/A              <18 months
of parental rights to
                          16.2          11.8        20.6         Comparison States: N/A     Outstanding:
adoption                                                                                    <9 months
 * Colorado indicator not exactly same as benchmark—counts only children with parents at case closure.
**Benchmarks are for point in time (on September 30) versus at any time in fiscal year (Indicator 5).




     William M. Mercer, Incorporated              213                 Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Summary of Findings
Pueblo County has evidenced very low rates of confirmed abuse and neglect in SFY 2000 and in
SFY 1999. In particular, the rate of recurring abuse and neglect was extremely low (1.6 percent)
relative to the rates in other large Colorado counties. The rate is extremely low compared to the
national and comparison state benchmarks.

Pueblos’ rate of abuse and neglect in SFY 2000 per 1,000 children in the population, is also
much lower than its rate of confirmed cases in SFY 1998, prior to the initiation or the managed
care reforms. The rate for confirmed cases of abuse that year of 5.4 percent is closer to the
average for other managed care counties in Colorado (as was the rate of abuse recurrence that
year of 4.2 percent) although still well below the national average.

Interestingly, although Pueblo County was found to have a very low rate of referrals of abuse
and neglect per 1,000 children in the population, its rate of investigations was comparable to
other large counties in Colorado. Pueblo has a very high percentage of referred cases that are
investigated relative to other Colorado counties, and to other states around the nation.

Pueblo had an extremely high rate of children placed out of the home, as measured by the
number of children per 1,000 in the population who were in out-of-home placement at some
point in SFY 2000. It is uncertain as to why the rate of abuse and neglect would be so low while
the rate of out-of-home placement is so high.

Pueblo County’s rate of out-of-home placement nearly doubled from SFY 1998 to SFY 1999 and
remained high in SFY 2000. While Pueblo’s rate of out-of-home placement was high, the
average length of stay in out-of-home placement was slightly lower than the average for other
managed care counties and slightly higher than the average for non-managed care large counties
in Colorado.

Pueblo’s rate of placement with parents at case closure was very high compared to other
managed care counties and slightly higher than other large Colorado counties. The rate of
positive placements was about average compared to other large counties in Colorado. From prior
to the implementation of managed care (SFY 1997 to SFY 2000), Pueblos’ rate of positive
placement has increased by about 5 percent, while its rate of placement with parents has
remained steady (59.5 percent in SFY 1997 versus 59.6 percent in SFY 2000).

Pueblo has evidenced a declining rate of adoption in the past three years and, at the same time,
an increasing average length of time between termination of parental rights and adoption.
Compared to other large counties in Colorado, however, Pueblo County does not have a
particularly low rate of adoption or a particularly long average number of months from
termination of parental rights to adoption.




William M. Mercer, Incorporated                214               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Pueblo’s pattern of results for the outcome indicators is unique compared to other counties,
comparable only to that of Mesa County, the other managed care county, that is located far from
the Denver metropolitan area. Rates of confirmed abuse and neglect are extremely low, while
rates of out-of-home placement are extremely high compared to other large counties in Colorado.

Pueblo’s trends in positive placements and placements with parents at case closure do not
parallel those of other counties, which tend to show a steady decrease in placements with parents
but a steady increase in all positive placements combined. Pueblo shows a slightly less dramatic
increase in positive placements overall but an unusually steady rate of placement with parents.
This pattern is complex and difficult to evaluate without seeing how it changes over time. As
additional data are collected and multi-year trends come into focus over the next two years of the
evaluation, the reasons underlying this pattern will become clearer. The main observation
currently is that Pueblo is an outlier in these categories and they should be examined more
closely.




William M. Mercer, Incorporated                215              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Cost Findings
A major assumption behind the introduction of managed care is that costs will be contained. It is
expected that a key factor underlying the containment of costs will be a shift from the use of high
cost services, such as residential treatment centers, to services that are less costly.

We conducted a series of financial analyses to determine if costs have been contained in the
managed care counties. Additional analyses were conducted to examine the pattern of
expenditures for higher cost care, as well as trends over several years. The results of these
analyses are shown in this section for Pueblo County by itself and in a later section for all of the
managed care counties together.

Figures 106 and 107 compare Pueblo County’s allocation at the beginning of SFY 2000 to its
expenditures at the end of the year. Figure 106 provides a comparison of the two main child
welfare funding streams, CORE services and the child welfare block. Figure 107 provides detail
regarding the funding streams that compose the child welfare block.

Figure 106 shows that Pueblo County experienced a small deficit in SFY 2000 in both CORE
services and the child welfare block. Figure 107 shows that Pueblo overspent its allocation for
80/20 and CHRP. It appears that Pueblo County overspent its allocation for SFY 2000.

                                       Figure 106
             CORE Services and Child Welfare Block Allocation vs. Expenditure*
                                Pueblo County, SFY 2000
         $16,000,000                                                $14,688,777 $15,049,008

         $13,500,000

         $11,000,000

          $8,500,000

          $6,000,000

          $3,500,000
                        $1,009,204 $1,044,884
          $1,000,000
                                  CORE Services                             Child Welfare Block
                                                  $(35,680)




                                                                                                  $(360,231)




   * Before Year End Close Out Process                              Allocation     Expenditure                 Difference




William M. Mercer, Incorporated                               216                Colorado Child Welfare Evaluation
                                                                        Second Interim Implementation Status Report
                                                            Figure 107
                                          Child Welfare Block Allocation vs. Expenditure*
                                                     Pueblo County, SFY 2000
      $30,000,000




                                        $15,049,008
                          $14,688,777




                                                                    $11,397,567
      $25,000,000




                                                       $9,919,678
      $20,000,000




                                                                                    $3,172,473

                                                                                                 $2,028,040
      $15,000,000




                                                                                                                                            $974,687

                                                                                                                                                       $974,687
                                                                                                                 $621,939

                                                                                                                            $648,714
      $10,000,000
       $5,000,000
                $0
                      Total CW Block                  Total 80/20                 RTC Medicaid                CHRP Medicaid               100% Admin.

                                                              Allocation                            Expenditure

    * Exclused Medicaid Fund and before Year End Close




Figures 108 and 109 provide a picture of how Pueblo County spent its child welfare block in
SFY 2000. Figure 108 shows the percentage of funds expended by budget item while Figure 109
shows the actual amount expended. Pueblo County was among the lower tier of managed care
counties in the proportion of expenditures for RTC. Of particular interest is a comparison of
Figure 108 with the same figure (89) for Mesa County as the proportions in each item are very
similar

                                           Figure 108
                Percentages of Child Welfare Block Expenditure* by Budget Items
                                    Pueblo County, SFY 2000

               100% Admin.
               6%                                                                                                                      Out of Home
                                                                                                                                       Placement Services
                                                                                  Other                                                41%
                                                                                  78%
      CHRP Medicaid                                                                                                                    Subsidized Adoptions
      4%                                                                                                                               9%
                                                                                                                                       Special
                                                                                                                                       Circumstances
         RTC Medicaid                                                                                                                  Child Care (SCCC)
         13%                                                                                                                           1%
                                                                                                                                        Case Services
                                                                                                                                        0%
                                                                                                                                        80/20 Admin.
       * Includes Medicaid Fund and before Year End Close Out Process                                                                   26%




William M. Mercer, Incorporated                                                   217                            Colorado Child Welfare Evaluation
                                                                                                        Second Interim Implementation Status Report
                                                 Figure 109
                              Child Welfare Block Expenditure* by Budget Items
                                          Pueblo County, SFY 2000
   $7,000,000
                                                                        $6,040,227
   $6,000,000

   $5,000,000
                                                                                                                                                  $3,766,903
   $4,000,000

   $3,000,000
                 $2,028,040
   $2,000,000                                                                             $1,366,587
                                                       $974,687
   $1,000,000                       $648,714
                                                                                                           $64,187         $159,662
           $0
                 R T C M edicaid   C H R P M edicaid   100% A d m in.   Out o f H o m e    Subsidized     Case Services        Special            80/20 A dmin.
                                                                         P lacement        A do ptio ns                   C ircum s t a n c e s
                                                                          Services                                           Child Care
                                                                                                                               (SCCC)
   * Before Year End Close Out Process




Figures 110–112 illustrate the trends in out-of-home expenditures in Pueblo County from
SFY 1998 through SFY 2000. Pueblo has the highest proportion of out-of-home placement
expenditures devoted to family foster care of all of the managed care counties. At the same time,
it has one of the lowest proportions spent on RTC.


                                                       Figure 110
                                           Out-of-Home Placement Expenditure
                                                Pueblo County, SFY 1998

                Specialized Group                                                  Independent
                                                  RCCF
                   Care 5.0%                                                          Living
                                                  4.9%          RTC                            Home Based Program
                                                                6.7%                   0.4%          0.0%
                                                            Relative Foster
                                                              Care 0.8%
                                                                                                                      Transitional
                                                                                                                       Program
                                                                                                                         0.0%
                    Family Foster Care                     Other
                          82.2%                                                                        Receiving
                                                           0.5%
                                                                                                        Home
                                                                                          Shelter Care   0.0%
                                                                                             0.1%




William M. Mercer, Incorporated                                              218                     Colorado Child Welfare Evaluation
                                                                                            Second Interim Implementation Status Report
                                            Figure 111
                                Out-of-Home Placement Expenditure
                                     Pueblo County, SFY 1999

                                  Specialized Group Care              Independent Living
                                           7.3%                             0.6%
                                                    RCCF
                                                                                            Home Based Program
                                                    3.2%
                                                                                                  0.0%
                                                      RTC
                                                     7.3%                                    Transitional
                                       Excludes
                                         RTC        Relative Foster Care                       Program
                                       Medicaid             0.7%                                0.0%

                      Family Foster Care                                                     Shelter Care
                            81.0%                                                               0.0%
                                                   Other
                                                   0.6%                                    Receiving Home
                                                                                                0.0%




                                            Figure 112
                                Out-of-Home Placement
                                Expenditure County, SFY 2000
                                     Pueblo
                  Specialized Group        RCCF
                     Care 6.6%             2.5%                        Independent Living Home Based
                                                                             0.7%         Program 0.0%
                                                          RTC
                                                          7.1%
                                                                                                       Transitional
                                           Excludes
                                             RTC
                                                                                                        Program
                                           Medicaid                                                       0.0%

                                                                                              Shelter Care
                  Family Foster Care
                                                                                                 0.0%
                        82.6%

                                                  Other     Relative Foster
                                                                                      Receiving Home
                                                  0.7%        Care 0.5%
                                                                                           0.0%




William M. Mercer, Incorporated                             219                 Colorado Child Welfare Evaluation
                                                                       Second Interim Implementation Status Report
Figures 113 and 114 illustrate the extent of privatization of family foster care and group
homes. Managed care entities tend to manage services through contracts with private
agencies rather than through direct operation. Figure 113 shows that Pueblo County has
contracted out most of its family foster care and group homes to CPAs. Figure 114 shows
that family foster care homes are the primary contracted service.

                                          Figure 113
               Percentages and Amounts of Expenditure for Family Foster Care and
                       Group Home Services delivered by County vs. CPA
                             Pueblo County, SFY 1998 to SFY 2000
        100%
         90%
         80%
         70%
         60%
                         $4,715,069                     $5,000,003                    $5,186,436
         50%
         40%
         30%
         20%
         10%
          0%
                          SFY1998                        SFY1999                       SFY2000



     County Family Foster Care Homes and Group Homes          CPA Family Foster Care Homes and Group Homes




                                           Figure 114
               Percentages of Expenditure for Family Foster Care and Group Home
                           Services delivered by County and/or CPA
                              Pueblo County, SFY 1998 to SFY 2000
       100%
        90%
        80%
        70%
        60%
        50%
        40%
        30%
        20%
        10%
         0%
                         SFY1998                        SFY1999                       SFY2000


   CPA Family Foster Care Homes       CPA Group Homes     County Family Foster Care Homes    County Group Homes




William M. Mercer, Incorporated                         220                   Colorado Child Welfare Evaluation
                                                                     Second Interim Implementation Status Report
Multiple County Findings
Previous sections focus on individual county findings. This section examines counties as a
group, examining their implementation of key managed care principles and fidelity to the goals
of the CDHS for the managed care initiative. It also presents outcome findings for both waves of
the managed care counties and the group of non-managed care counties.

Process Findings
Two types of process findings are presented. First, the implementation of key managed care
processes is examined across counties. Second, the fidelity of managed care counties, as a group,
to managed care principles is described.

As noted previously, Mesa County decided not to continue as a managed care county. As a
result, it is included in the first section below but SFY 2000 data are used. It is not included in
the fidelity analysis as the necessary data for this analysis could not be collected.

Introduction of Managed Care Principles
Although the primary analyses of the process evaluation focus on the changes within each county
over time and their effects, a secondary level of analysis involves looking at process changes in
key categories across counties. Given differences between counties and the multiple levels and
the complexity of the variables involved in the managed care pilots, it cannot be definitively
stated which of the various county approaches are most effective. Instead, this analysis focuses
on describing the differences between counties and noting the degree to which they fit with some
general principles of managed care.

Financial Relationship with State
A central component of the managed care counties has been their unique financial relationship
with the state, specifically their ability to retain unspent funds resulting from their managed care
initiatives. All Colorado counties, including managed care counties, are responsible for any
overages from budgeted expenditures. Through this mechanism, the CDHS has shifted full
downside risk to local counties. Up until this year, managed care counties have had the
distinction of being able to retain any unspent funds from the child welfare block, thus taking on
the upside risk for the services they provide. To do this, however, managed care counties had to
allow non-managed care counties to use unspent funds from other areas of the state to cover their
overages before being eligible for these.




William M. Mercer, Incorporated                  221               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
In 2001, in response to pilot counties’ concerns about having to give up eligibility for statewide
surplus distributions, the CDHS offered counties a choice between keeping any of their own
unspent funds or being eligible for participation in the distribution of unspent funds from other
counties at the end of the year. Offered part way through the year, this gave counties the ability
to predict how their finances were likely to come out and to choose the option in their best
interest. All counties, except Pueblo, have chosen the MOU option where they give up the ability
to retain their own unspent funds and, instead, are able to participate in surplus distributions from
the state, as seen in Table 11.

                                          Table 11
                            Unspent funds or surplus distribution
                           Unspent funds                             Surplus
                 (take on upside and downside risk)          (downside risk protection)
 Arapahoe
 Boulder
 El Paso
 Jefferson
 Pueblo
 Mesa                                          Data not available

Management Information System
A key issue is the quality of financial reporting that each county relies upon for managing
utilization of purchased services. Managed care, or for that matter, the management of any model
of care, requires an information system that can provide the appropriate and timely reports
needed to properly manage a system of care.

As Table 12 shows, all counties track their internal costs for staff and overhead. Pueblo County
relies entirely upon the state for financial tracking and reporting of purchased services. Four
counties—Arapahoe, Boulder, El Paso, and Jefferson—track their own costs for purchased
services. In three of these counties—Arapahoe, El Paso, and Jefferson—it takes specialized,
labor-intensive queries to use those data and because of the difficulty aggregating the data, these
counties also rely on CDHS reporting to track provider expenditures and manage utilization.
Data on financial reporting were not available from Mesa County.




William M. Mercer, Incorporated                 222               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
                                           Table 12
                 Data reporting source relied upon for financial management
                                           State                    County
                Arapahoe
                Boulder
                El Paso
                Jefferson
                Pueblo
                Mesa                               Data not available


El Paso and Jefferson counties cite the accuracy and timeliness of CDHS financial reports as a
key reason for choosing the MOU option to protect them from downside risk. Both counties have
created new financial staff positions in the last year to begin to address their local data needs and
decrease reliance on the CDHS. Boulder County tracks its own expenditures and uses these data
throughout the year to monitor their financial position and reallocate any unspent funds to new
projects so that the use of funds is maximized.

Utilization Management
Utilization management refers to formal processes for ensuring that the right service is offered in
the right amount at the right time. Utilization management is an especially useful process for
containing the cost of the most expensive services, such as residential treatment. In managed care
programs, it is usually one of the first processes introduced when a program is implemented.

One utilization management approach is prior authorization of services. Through prior
authorization, services are not paid for unless the county has endorsed the service as necessary in
advance of the service being provided. All counties except El Paso have implemented some level
of prior authorization.

El Paso County, however, does review all out-of-home referrals prior to placement. El Paso
County has not given its review process the authority to deny the necessity of a placement
referral by a caseworker. Instead, the review process is consultative and the regular supervisory
process is used to resolve disagreements.

All prior authorization in the other five pilot counties is conducted by multi-agency teams.
Boulder and Mesa counties subject most services, including community-based referrals, to prior
authorization. Jefferson and Pueblo counties subject all out-of-home referrals to prior
authorization, and Arapahoe County targets its prior authorization process to RTC referrals only.
Table 13 shows the services requiring prior authorization in each county.




William M. Mercer, Incorporated                 223               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
                                           Table 13
                                 Scope of prior authorization
                                        All           Select
                     Most services out-of-home out-of-home              No prior authorization
 Arapahoe
 Boulder
 El Paso
 Jefferson
 Pueblo
 Mesa


Another utilization management tool is concurrent review. This refers to ongoing review of
placement appropriateness after a placement has been reviewed. This allows counties to monitor
the direction of treatment in placements, better ensure a focus on child welfare goals, and
promote more appropriate lengths of stay in placements.

As Table 14 shows, four counties—Arapahoe, Boulder, Jefferson, and Mesa—perform
concurrent review on out-of-home placements. Mesa County uses its multi-agency CORE Team
process for most services and dedicated RTC reviewers on the Front Range to conduct
concurrent reviews. Boulder County conducts concurrent reviews for most services through
IMPACT. Jefferson County conducts concurrent reviews of out-of-home placements through its
SURT. Arapahoe County uses dedicated reviewers to conduct concurrent reviews for RTCs.

                                          Table 14
                                  Concurrent review process

                    Multi-agency team          Dedicated staff/unit         Supervisor only

Arapahoe
Boulder
El Paso
Jefferson
Pueblo
Mesa

Case coordination between agencies can also be an important utilization management tool. Table
15 shows that all managed care counties have forums for multi-agency planning.

Multi-agency planning occurs at two levels. At the system level, agencies can come together to
plan joint initiatives and ensure administrative coordination. At the individual case level,
agencies can come together to ensure integrated services. Multi-agency planning efforts at either
level can be limited to a single, integrated forum or spread across multiple forums.

William M. Mercer, Incorporated                224              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
 In Boulder County, both levels of multi-agency planning occur in a single, integrated forum,
 IMPACT. In Mesa County, separate forums are available for system- and individual-level
 planning but those forums are each integrated (i.e., there is a single forum for each level of
 planning). In Arapahoe, El Paso, and Jefferson counties, multiple forums exist for planning at
 both levels. In Pueblo County, multiple forums exist for individual case planning but there are
 not any formal mechanisms for multi-agency system-level planning. Instead, this is conducted on
 an ad hoc and informal basis.

                                           Table 15
                                    Multi-agency planning
            Integrated multi-  Integrated
                                             Integrated multi- Multiple multi- Multiple multi-
             agency planning multi-agency
                                             agency planning - agency forums - agency forums -
              at system and    planning -
                                              individual level   system level individual level
             individual level system level
Arapahoe
Boulder
El Paso
Jefferson
Pueblo
Mesa

 Another level of case coordination is coordination within the local department of social or human
 services between child welfare and TANF services. Coordination can happen at both the point of
 access (i.e., are there single points of access to both programs or is access to each independent)
 and at the program level (i.e., are there jointly funded programs that can be accessed by persons
 from both programs or are programs in each area separate).

 Table 16 shows that El Paso County has integrated access and programming across multiple
 levels. Clients can access either program directly through either the child welfare or TANF
 intake units and there are many jointly funded programs available to either population. Boulder
 County has also integrated access and programs but to a lesser degree. Jefferson County has
 separate access but some level of joint programming. Arapahoe, Pueblo, and Mesa have distinct
 child welfare and TANF programs with only a minimal level of informal cross-referral between
 them.




 William M. Mercer, Incorporated                225              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                              Table 16
                                       Integration with TANF
             Integrated access
                               Integrated access and Separate access, Separate access and
            and programs—high
                               programs—low level some joint programs     programs
                   level
Arapahoe
Boulder
El Paso
Jefferson
Pueblo
Mesa

Quality Improvement
Quality improvement refers to formal processes through which specific areas for improvement
are identified, data are collected, solutions are developed and implemented, and data are
collected again to determine if solutions are working and to refine them as needed. All counties
have the quality of their compliance with settlement agreement indicators monitored by foster
care reviewers at the CDHS. For most Colorado counties, including managed care counties, this
limited level of externally imposed quality review is all that is available.

Although none of the managed care counties have a full quality improvement system, Table 17
shows that three counties have implemented quality improvement activities. El Paso has a formal
quality improvement plan and program tied to agency-wide and program-level goals and has a
major focus on using data to improve services over time. Boulder and Mesa counties have
specific quality-oriented initiatives of more limited scope.

                                            Table 17
                                  Quality improvement system
                            Formal written       Some level of    State foster care
                                plan            formal process      review team
          Arapahoe
          Boulder
          El Paso
          Jefferson
          Pueblo
          Mesa




William M. Mercer, Incorporated                226              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Provider Network
An adequate provider network contains services that are accessible, appropriate to the needs of
children and their families, of demonstrated effectiveness, and available in adequate quantity. To
achieve this, counties have adopted a variety of approaches.

At the level of network structure, Boulder County administers its own coordinated network
through IMPACT (see Table 18). Arapahoe and El Paso counties have focused upon broad-level
partnerships with provider networks managed by subcontractors. While all six pilot counties
have taken advantage of the option of passing CPA treatment funds through their local MHASA
and, thereby, use the MHASA’s CPA provider networks, Arapahoe and El Paso have initiated
relationships with additional networks: Arapahoe for RTC services and El Paso for CORE
services. In addition, El Paso, Jefferson, and Pueblo counties have specific partnerships with
individual provider organizations.

                                           Table 18
                                  Provider network structure
                                  Partnerships with Partnerships with Partnerships with
               Coordinated
                                  provider networks  CPA provider        individual
                network
                                     beyond CPA       network only        providers
Arapahoe
Boulder
El Paso
Jefferson
Pueblo
Mesa

Another difference between counties in relationship to their provider networks is whether
counties have allocated resources to centrally track provider capacity and support caseworkers
making out-of-home referrals. As Table 19 shows, all managed care counties, except Arapahoe
and Pueblo, have allocated dedicated resources to such efforts.

                                          Table 19
                           Central support to facilitate out-of-home
                                          referrals
                                             Yes               No
                           Arapahoe
                           Boulder
                           El Paso
                           Jefferson
                           Pueblo
                           Mesa


William M. Mercer, Incorporated                227               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Provider Financial Incentives
Provider financial incentives refer to payment methods through which providers are encouraged
to perform in ways that support the mission of the county child welfare system. Financial
incentives may range from bonuses to the use of risk-shifting strategies such as capped funding
allocations and case rates.

Boulder and El Paso counties have taken the most steps in this direction. Both counties have in
place formal contractual agreements that set capped funding levels and share risk between the
county and other agencies. Boulder does this through the IMPACT program and the
comprehensive services it funds. El Paso has done this through capped contracts with risk
corridors and case rates for CORE services. Boulder, El Paso, Jefferson, Mesa, and Pueblo
counties also have contractual risk-sharing agreements for CPA treatment services that place the
full risk for treatment costs on the local MHASA and the full risk for room and board costs on
the county child welfare department (see Table 20).

Arapahoe and El Paso counties have also built specific incentives for performance into some
provider contracts. Arapahoe County has developed special rates to promote quicker access and
shorter lengths-of-stay for services through a specific provider network (CCM). El Paso County
has developed incentives related to adoption for its CPA Project.


                                          Table 20
                                     Provider incentives
                   Contractually-based      Shared risk through CPA              Contracts with
                 shared risk beyond CPA         transfer and formal               performance
                         transfer              contractual language                incentives
Arapahoe
Boulder
El Paso
Jefferson
Pueblo
Mesa




William M. Mercer, Incorporated               228               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Financial incentives between child-serving agencies can also be better aligned through initiatives
that blend funding and resources to promote efforts of mutual benefit. Boulder County has gone
the farthest in this direction through IMPACT, which blends funding for essentially all out-of-
home placements as well as a host of placement alternatives for child welfare, juvenile justice,
mental health, substance abuse, and other agencies. Arapahoe, El Paso, and Mesa counties all
support joint programs with other agencies through blended funding for these specific projects,
to a greater (e.g., El Paso) or lesser (e.g., Arapahoe) degree. All six pilot counties also have
numerous initiatives to which they and other agencies contribute staff jointly (see Table 21).

                                        Table 21
                               Blended resources approach
             Blended funding across       Blended funding on         Blended staffing on
               multiple programs           specific projects          specific projects
 Arapahoe
 Boulder
 El Paso
 Jefferson
 Pueblo
 Mesa

Fidelity of Process Implementation
An important component of the evaluation is the assessment of the manner in which counties are
implementing managed care. Embedded within the MCIR are sets of items that measure the
extent to which counties have implemented managed care in a way that is consistent with the
CDHS’s expectations. The specific question of interest is, “Are changes in organizational
processes in line with intended goals?” When counties implement managed care in a way that is
consistent with such expectations, we call that “process fidelity.”

Fidelity is scored in each of four domains:

   Funding Flexibility. Are system dollars used flexibly and creatively and are counties taking
  advantage of managed care financing techniques in their contracts and arrangements with
  providers?

   Interagency Coordination. What is the level of coordination and integration of child welfare,
  mental health, youth corrections, substance abuse, developmental disability, and self-
  sufficiency services?

    Integration of Informal Networks and Non-Traditional Services. Is there integration of
  informal community services and non-traditional supports with traditional child welfare
  services to form a comprehensive service array for children and families? Is there greater
  “community ownership” for the well-being of children and families as evidenced by increased
  funding participation and community involvement in service design, delivery, and monitoring?

William M. Mercer, Incorporated                229              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
        Implementation of Early Intervention and Prevention Programs. Is there an increased
      investment in prevention and early intervention efforts at the community level?

   Each of the fidelity domains has scale scores attached. The scale scores allow for identifying the
   stage of managed care implementation for a given county. The four stages of fidelity
   implementation are:

       Early Stage Fidelity—The county has not yet achieved or is just beginning to achieve a
      minimum level of fidelity to the reform area in question.

        Emerging Stage Fidelity—The county has begun to demonstrate a significant adherence to
      the reform area in question.

        Maturing Stage Fidelity—The county has demonstrated fidelity in the majority of issues
      related to the reform area in question.

        Mature Stage Fidelity—The county has implemented the reform area in question in a way
      that would be comparable with the most advanced counties in the state and around the country.

   Table 22 shows the results across all five counties that participated in the MCIR this year. The
   table shows the scores and the stages of fidelity for each county within each of four fidelity
   domains.

                                              Table 22
         System Fidelity Results for the Five Present Managed Care Counties in Colorado
                       Arapahoe Co.  Boulder Co.   El Paso Co.   Jefferson Co.   Pueblo Co.
                            Fidelity      Fidelity      Fidelity        Fidelity     Fidelity
                      Score Stage Score Stage Score Stage Score Stage Score Stage
Funding Flexibility
(Range = 0–12)         2      Early      9    Mature      9     Mature     5   Emerging     4    Emerging

Interagency
Coordination           10   Maturing    15    Mature     15     Mature    14     Mature    12    Maturing
(Range = 0–16)
Informal Networks
(Range = 0–24)         9    Emerging    15   Emerging    17    Maturing   10   Emerging     8    Emerging

Prevention/Early
Intervention           0      Early      6    Mature      5     Mature     2   Emerging     2    Emerging
(Range = 0–6)




   William M. Mercer, Incorporated                230               Colorado Child Welfare Evaluation
                                                           Second Interim Implementation Status Report
As can be seen in the table, two of the counties are mature in their implementation of the system
reforms envisioned for the managed care initiative when assessed using the fidelity domains
outlined on the previous page. Boulder and El Paso counties each achieved mature levels of
implementation fidelity in all but one fidelity domain. The only domain in which they did not
achieve a mature stage of fidelity was the incorporation of informal networks into the service
system. Both counties, however, have achieved significant success in that domain and both were
found to be at (El Paso) or nearly at (Boulder) the maturing stage of fidelity in that domain.

The other three counties included in the analysis experienced less consistent levels in their
system process fidelity ratings. Jefferson County evidenced a mature level of implementation in
the interagency collaboration domain but received emerging stage fidelity ratings in the other
three domains. The highest stage of fidelity achieved in any domain by Arapahoe and Pueblo
counties was the maturing stage of fidelity. Both received that rating in the interagency
collaboration domain.

Across the four domains, the five counties, as a group, received the highest fidelity ratings in the
interagency collaboration domain. Three counties achieved a mature rating in that domain and
the remaining two were rated in the maturing stage of fidelity. As a group, the five counties’
fidelity of implementation was rated lowest in the funding flexibility and prevention/early
intervention domains. Although two counties (Boulder and El Paso) were rated as performing at
the mature level in these domains, the remaining counties were rated as only in the emerging or
early stages of fidelity.

Overall, the results indicate that all counties have achieved significant success in the state reform
goals in the interagency collaboration domain. The counties vary somewhat in their
implementation in the other three domains. These variations are discussed for each county in the
county-level results sections.




William M. Mercer, Incorporated                 231               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Outcome Findings
As mentioned earlier, our design uses managed care pilot counties as their own controls. We are
tracking key outcome indicators for all pilot counties over the course of the evaluation to see
how each county changes in its performance over time. We have separated the six counties into
two sets of three counties based on when the counties began implementing managed care and
taking on financial risk for their child welfare programs.

We have added to the basic design of using counties as their own controls a comparison group
made up of the four other largest counties in Colorado. Examining the trends in the other four
large counties alongside the trends for the six managed care counties helps us to control for the
effects of other events happening within the Colorado child welfare system.

In our presentation of the results of outcome indicators we typically will summarize the results
for three groups of counties. There groups are summarized in Table 23.




William M. Mercer, Incorporated                 232              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                        Table 23
                      Counties used in the Evaluation of Outcomes
  WAVE 1 Managed Care           WAVE 2 Managed Care           Non-Managed Care Large
         Counties                        Counties                   Counties
(began implementation 7/1/97) (began implementation 7/1/98)

         Boulder County                   Arapahoe County                   Adams County

        Jefferson County                   El Paso County                   Denver County

          Mesa County                      Pueblo County                   Larimer County

                                                                             Weld County

The results of our outcome analyses are presented below. Wherever data from state databases
were available and appeared to be complete and reliable, we analyzed the trends for each
outcome indicator from SFY 1996 through SFY 2000. In some cases, the federal fiscal year was
used because the data are reported to the federal government based on the federal fiscal year of
October 1 to September 30. In some cases, data from SFY 1996 were not complete and/or valid
enough to be used, whereas data from SFY 1997 forward were useful for a particular indicator.




William M. Mercer, Incorporated               233               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 1: Recurrence of Abuse and Neglect
This indicator is “Of all children who were victims of substantiated child abuse and/or neglect
during the reporting period, what percentage had another substantiated report within a 12-month
period?” In calculating this indicator, we examined the percentage of children who were abused
and/or neglected within 12 months of a previously confirmed abuse and/or neglect incident.
Children who were confirmed to have been abused and/or neglected during the reporting period
in question were tracked, using data from the Central Registry, to determine whether they had
suffered from another confirmed abuse and/or neglect incident not more than 12 months prior to
the incident from the current reporting period.

Figure 115 shows the percentage of children who experienced a recurrence of abuse and/or
neglect for the managed care counties compared to the non-managed care counties. It shows a
similar pattern for all three groups of counties with all evidencing a rise in rates of recurrence of
abuse and/or neglect in FFY 1998 and a return to near FFY 1997 levels of recurrence in FFY
1999. This pattern is generally true at the individual county level with one or two exceptions.

Wave 1 and Wave 2 managed care counties show slight increases (.1 percent and .6 percent)
from SFY 1997 to SFY 2000 while non-managed care counties actually evidence a slight
decrease (1.1 percent). Managed care counties are separated into Wave 1 counties, which began
managed care in SFY 1998, and Wave 2 counties, which began in SFY 1999.

The sharp increase in SFY 1998 is unusual. One might suspect that it is due to a contextual
factor, such as a highly publicized case of the recurrence of abuse that resulted in case workers
being more likely to confirm abuse during the year. If this were the case, however, an increase in
substantiated abuse and neglect would also be expected during the same year and this did not
occur (Figure 116). It could also be the result of a problem in reporting that year. We will
examine this year more closely in the next year of the evaluation.

                                      Figure 115
    Percentage of Children Who Experienced a Confirmed Recurrence of Abuse and/or
          Neglect Within a 12-Month Period after Being Abuse and/or Neglected

                10
                                               8
                 8
                                                        5.7
                 6                                                              4.9   4.4
                       4.3                                         4.4
                                    4.1             4
                                                                          3.5
                 4                                                                          2.9 3
                             2.3
                 2

                 0
                          FY 97                    FY 98             FY 99              FY 00
                                   MC Wave 1                  MC Wave 2               Non-MC


William M. Mercer, Incorporated                          234                 Colorado Child Welfare Evaluation
                                                                    Second Interim Implementation Status Report
    Data on rates of recurring abuse and neglect are available for benchmarking this indicator. They
    are drawn from “Child Maltreatment 1999,” which contains data on recurrence of abuse and
    neglect within 6 months of the initial substantiated incident of abuse or neglect, and “Child
    Welfare Outcomes 1998 Annual Report,” which contains the federal government’s most recent
    data on rates of recurring abuse and neglect within 12 months of the original incident of abuse or
    neglect. It should be noted that using a 6-month period to check for the recurrence of abuse and
    neglect will generally result in lower rates of recurring abuse and neglect than using a 12-month
    period. It should also be noted that not all states reported for these benchmarks; only two
    comparison states reported.

    The “Child Maltreatment 1999” and “Child Welfare Outcomes 1998” reports draw from the data
    on rates of recurrence of abuse and neglect that the federal government receives from states.
    states have different methods and processes for investigating and confirming referrals; therefore,
    the benchmarking data should be interpreted with some caution. The rates of confirmed abuse
    and neglect at the national level and for the few comparison states that reported are presented in
    Table 24.

    Benchmark comparisons show that all three groups of Colorado counties have a much lower rate
    of recurring abuse and neglect than is found at the national level and in the comparison states.
    This holds true across years and across the 6-month and the 12-month benchmark, except for
    Wave 1 counties in SFY 1998.

                                          Table 24
                 Benchmark Comparisons of Rates of Recurring Abuse and Neglect
                                Per 1,000 Child Population

                           Comparison State         Wave 1 County    Wave 2 County      Other Large
     National Rate             Means                  Average          Average         County Average

                                                     4.3 (1997)        2.3 (1997)         4.1 (1997)
11% (’99, w/in 12 mos)   12%* (’99, w/in 12 mos)     8.0 (1998)        4.0 (1998)         5.7 (1998)
7.5% (’98, w/in 6 mos)   9.4%** (’98, w/in 6 mos)    4.4 (1999)        3.4 (1999)         4.9 (1999)
                                                     4.4 (2000)        2.9 (1900)         3.0 (2000)
*    Utah
** Utah and Washington




    William M. Mercer, Incorporated                 235              Colorado Child Welfare Evaluation
                                                            Second Interim Implementation Status Report
Indicator 2: Substantiated Abuse and Neglect
This indicator is “What is the amount of substantiated abuse (confirmed) abuse and neglect per
1,000 children in the population?” To calculate this indicator for SFY 1997 through SFY 2000,
we first divided the number of cases of abuse and neglect by the number of children in the
population between the ages of 0 and 17 in each county during the SFY in question. This number
was then multiplied by 1,000 to obtain the rate of confirmed abuse and neglect per 1,000 children
in the population. Child population data were taken from the state Demographer’s office in
Colorado.

Figure 116 depicts the overall rate of abuse and neglect per 1,000 child population for the
Wave 1 and Wave 2 managed care county groups and for the other four large counties. It shows
that, after evidencing a slight increase in the rate of confirmed abuse and neglect from FFY 1997
to FFY 1999, all three groups of counties showed a decrease in the rate of abuse and neglect
from FFY 1999 to FFY 2000. The results still indicate a slight increase in the rate of confirmed
abuse and neglect for all three groups from FFY 1997 to 2000. This finding could be due to
actual increases in the rate of abuse and neglect or changes in the processes of investigating
referred cases of abuse and neglect in the field. We will explore this change in more detail in the
next year of the evaluation.

                                     Figure 116
        Confirmed Abuse and Neglect per 1,000 Children in the County Population


            8
                                                  7
                                   6                                  6
            6    5.3                                          5.4
                                             5         4.8                      5.1
                             4.7       4.4                                4.3
            4          3.3


            2


            0
                    FY                 FY              FY                 FY
                             MC Wave         MC Wave                Non-MC
                             1               2




William M. Mercer, Incorporated                  236                  Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Benchmarking data are available for this indicator. They are drawn from the federal
government’s most recent data on rates of abuse and neglect, reported in “Child Maltreatment
1999.”

“Child Maltreatment 1999” draws from the data on rates of abuse and neglect that is received
from the 50 states that provided data submissions to NCANDS and AFCARS. States have
different methods and processes for investigating and confirming referrals and the benchmarking
data should be interpreted with some caution. The rates of confirmed abuse and neglect at the
national level and for the group of comparison states are shown in Table 25.

Benchmarking findings indicate that large counties in Colorado have a lower rate of confirmed
abuse and neglect than other states. In 1999, all three groups’ rates of confirmed abuse and
neglect (7.0, 4.8, and 5.4) were lower than any of the benchmarking comparison states’ rates for
that same year (except for Wave 1 counties compared to Washington State, which had a rate of
5.4) and were far lower than the national average (11.9).

                                      Table 25
    Benchmark Comparisons of Rates of Abuse and Neglect per 1,000 Child Population
                 Comparison      Wave 1 County Wave 2 County            Other Large
  National Rate State Median         Average           Average        County Average
                                     5.3 (1997)       3.3 (1997)         4.7 (1997)
                                     6.0 (1998)       4.4 (1998)         5.0 (1998)
     11.8          8.8
                                     7.0 (1999)       4.8 (1999)         5.4 (1999)
                                     6.0 (2000)       4.3 (2000)         5.1 (2000)




William M. Mercer, Incorporated               237               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Indicator 3: Referrals and Investigations
This outcome indicator is “What is the number of referrals and investigations of alleged abuse
and neglect per 1,000 children in the population?” To calculate this indicator, we first divided the
number of referrals of alleged abuse and neglect by the number of children between the ages of
0 and 17 in each county during the SFY in question. This number was then multiplied by 1,000
to obtain the number of referrals per 1,000 children (0–17) in the population during the SFY. The
same method was used to calculate the rate of investigations of alleged abuse and neglect.

Figures 117 and 118 depict the overall rates of referrals and investigations of alleged abuse and
neglect per 1,000 children in the population for the Wave 1 and Wave 2 managed care counties
and for the other four large counties. The results generally indicate a decrease in the rates of
referrals and investigations of abuse and neglect from SFY 1997 to SFY 2000 across the large
counties in Colorado. The decreases are most consistent and most pronounced for the
non-managed care counties. The Wave 2 managed care counties evidenced a relatively flat trend
from SFY 1997 to SFY 2000.

All managed care counties combined received approximately the same number of referrals in
SFY 2000 that they did in SFY 1997 per 1,000 children in the population. Other large counties
saw a 20 percent drop in referrals from SFY 1997 to SFY 2000.

The rate of investigation of referrals was almost the same between SFY 1997 and SFY 2000 for
managed care counties (57 percent versus 56 percent) and rose from 60 percent to 65 percent for
non-managed care counties. Meanwhile, the percentage of investigated cases that were
confirmed rose by four percent from SFY 1997 to SFY 2000 for both managed care and
non-managed care counties. The results for the managed care counties suggest that managed care
has not led to a lessening of counties’ aggressiveness in investigating referred cases of abuse and
neglect.




William M. Mercer, Incorporated                 238              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                      Figure 117
   Referrals of Alleged Abuse and Neglect per 1,000 Children in the County Population
         100

          90

          80

          70
                                 60
          60                                           54
                                                                            49                  48
                   45                            46                                    46
          50               43            43                          43
                                                              40                 41
          40

          30

          20

          10

              0
                        SFY 97            SFY 98         SFY 99                       SFY 00
                                 Wave 1 MC       Wave 2 MC                  Non-MC



                                       Figure 118
 Investigations of Alleged Abuse and Neglect per 1,000 Children in the County Population

        100

         90
         80

         70
         60

         50
                                36                     34
         40                                                                 31                  31
                         25                                          26                  27
                  25                             25           22
         30                              22                                       21
         20

         10
          0
                       SFY 97                 SFY 98               SFY 99              SFY 00

                                     Wave 1 MC          Wave 2 MC            Non-MC


William M. Mercer, Incorporated                         239                 Colorado Child Welfare Evaluation
                                                                   Second Interim Implementation Status Report
Data on rates of abuse and neglect referrals and investigations are available nationally and for the
group of comparison states. The benchmarking data are taken from the federal government’s
most recent data on referrals and investigations of abuse and neglect from “Child Maltreatment
1999.”

The federal government’s report draws from the data on referrals and investigations of abuse and
neglect that it receives from states. states have different methods and processes for counting
referrals and investigations so the benchmarking data should be interpreted with some caution.
Nevertheless, especially when aggregated across states, they may provide a useful source of
comparison data for Colorado. The rates of referrals and investigations of abuse and neglect for
all states nationwide and for the group of comparison states are presented in Table 26, along with
the rates for Wave 1, Wave 2, and non-managed care counties.

With regard to the benchmarking results, and examining SFY 2000 in particular, it appears that
Colorado’s managed care counties receive a lower rate of both referrals and investigations,
compared to other states, but investigate a relatively comparable percentage of cases referred. In
SFY 2000, 56 percent of referred cases were investigated in Colorado’s managed care counties.
Among comparison states, approximately 53 percent of referred cases were investigated.

It is uncertain whether Colorado’s particularly low rate of confirmed abuse and neglect is due to
a lower rate of actual occurrence of abuse and neglect, a lower rate of reporting and referral of
suspected abuse and neglect cases, or more stringent standards used in confirming abuse or
neglect among investigated cases. It appears, however, that there is not evidence for the
possibility, raised in last year’s evaluation report, that the lower rate of confirmed abuse and
neglect could be due to a tendency to investigate a lower percentage of the referrals received.

                                      Table 26
   Benchmark Comparisons of Abuse and Neglect Referrals and Investigations per 1,000
                        Children in the County Population
               National       Comparison       Wave 1         Wave 2       Other Large
                 Rate        State Average     County         County         County
                                (Mean)         Average       Average         Average
                                                45 (1997)     43 (1997)       60 (1997)
                 72                53           43 (1998)     46 (1998)       54 (1998)
Referrals
                                                40 (1999)     43 (1999)       49 (1999)
                                                41 (2000)     46 (1900)       48 (1900)
                                                25 (1997)     25 (1997)       36 (1997)
                  44               28           22 (1998)     25 (1998)       34 (1998)
Investigations
                                                22 (1999)     26 (1999)       31 (1999)
                                                21 (2000)     27 (2000)       31 (2000)




William M. Mercer, Incorporated                 240              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Indicator 4: Discharge to Reunification
This indicator is “What is the percentage of children who were placed with parents at case
closure?” For this indicator, we calculated the percentage of children who were residing with
their parents at case closure. These data were available for children whose cases were closed in
each time period (SFY 1997, SFY 1998, SFY 1999, and SFY 2000) and who had been placed out
of the home at some point during their tenure as an open case.

It should be noted that reunification is a bi-directional variable. That is to say, both low and high
rates of reunification can be problematic. The comparison data in this section imply that higher
reunification is always a better outcome. It is clear, however, that a single-minded focus on
reunification would represent a compromise of child safety in many cases, as well as other goals
(such as permanency). As a result, while differences between Colorado and other states are
important, they must not be compared simplistically.

Figure 119 shows comparisons between the managed care counties and non-managed care
counties. The results generally indicate a continued decrease in the rates of children residing with
parents at case closure from SFY 1997 to SFY 2000 for Wave 1 counties (9 percent drop in three
years, 6 percent drop in 2) and a relatively steeper continued decrease for Wave 2 counties (12
percent drop in 3 years, 17 percent drop in last 2 years). At the same time, the non-managed care
counties, which saw a decrease of nearly 18 percent from SFY 1998 to SFY 1999 in the
percentage of children residing with parents at case closure, evidenced an increase of three
percent from SFY 1999 to SFY 2000.

                                         Figure 119
                 Percentage of Children Placed with Parents at Case Closure

        100
         90
         80
                              67                 67
         70                               61
                57     56                                             55                  58
         60                        54                   51      53
                                                                            48
                                                                                    44
         50
         40
         30
         20
         10
          0
                     SFY 97             SFY 98               SFY 99              SFY 00
                               MC Wave 1          MC Wave 2            Non-MC




William M. Mercer, Incorporated                   241                 Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Data are available for benchmarking this indicator. Data on rates of reunification with parents
and caregivers following discharge from out-of-home placement are available through the
federal government’s AFCARS data set. The most recent data available at the national level were
found in the federal government’s AFCARS Report, October 2000. The most recent data
available at the comparison state level were found in the report, “Child Welfare Outcomes
1998.”

It is important to note that the benchmark indicators are not exactly the same as the indicators
from the Colorado managed care evaluation. In the Colorado evaluation, children’s status at case
closure is examined, whereas the benchmark data examine children’s status at discharge from
foster care. In addition, the Colorado data only include parents and not other types of caregivers
with whom children could have been reunified, whereas the AFCARS indicators appear to
include parents and other caregivers. For these reasons, the results of the benchmark
comparisons should be interpreted with caution and are likely to underestimate Colorado’s
reunification rate in comparison to other states.

Benchmark comparisons in Table 27 show that all three groups of counties were lower than both
the national average and the average for comparison states in SFY 2000. All three groups have
consistently moved further away from the national and comparison states means over time.

                                          Table 27
                          Mean Rates for Reunification with Parents
 National Mean        Comparison         Wave 1         Wave 2 counties’             Other Large
                      States’ Mean   counties’ Mean           Mean                  counties’ Mean

         59%                68%             57% (1997)              56% (1997)          67% (1997)
                                            54% (1998)              61% (1998)          67% (1998)
                                            51% (1999)              53% (1999)          55% (1999)
                                            48% (2000)              44% (2000)          58% (2000)




William M. Mercer, Incorporated                242                Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 5: Children in Out-of-Home Placement
This indicator is “What is the number of children in out-of-home care at any point during the
SFY per 1,000 children in the population?” It should be noted that the method used to calculate
this indicator last year has been refined. In last year’s analysis, children who did not yet have an
“end service” date were excluded from the analysis. These children are now included for each
fiscal year of the analysis.

Figure 120 shows the results on this analysis. In SFY 2000, each of the three groups evidenced a
slight to moderate decrease in the rates of out-of-home placement as measured by this indicator.
The rates of out-of-home placement continue to be lowest for Wave 1 managed care counties,
next lowest for Wave 2 managed care counties, and highest for non-managed care counties.

The pattern of change over time is very similar for each group of counties, with each showing a
decrease in SFY 1997 and SFY 1998, and then an increase to above SFY 1996 rates in
SFY 1999, followed by another decrease in SFY 2000. Non-managed care counties returned
closer to their SFY 1996 rate in SFY 2000 (only .1 child per 1,000 higher in 2000 than in 1996)
than did the managed care counties, which were 1.1 children per 1,000 higher than in SFY 1996.

                                       Figure 120
         Children in Out-of-Home Placement at Some Point during the Fiscal Year
                       Per 1,000 Children in the County Population
         20
         18
         16
         14
         12                                                                         11.6
                           10.5                                                                        10.6
                                                                              9.1
         10                                                       8.7                            8.6
                                               8.5
                     7.5                                    6.8
          8                              6.4                            6.2
                                                                                           6.1
          6     5                                    4.7
                                   4.1
          4
          2
          0
                    SFY              SFY                   SFY            SFY                SFY
                                  MC Wave                  MC Wave              Non-




William M. Mercer, Incorporated                      243              Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Although benchmark data on the number of children in out-of-home placement at some point
during the fiscal year per 1,000 children in the population are not available, data on the number
of children in out-of-home placement on September 30 are available from the AFCARS October
2000 Report and “Child Welfare Outcomes 1998.” Such a comparison should overstate
Colorado’s out-of-home placement rate in comparison with other states, although trends in rates
should be comparable over time.

Benchmark comparisons are shown in Table 28. These data indicate that the national rate of
out-of-home placement, per 1,000 child population, on September 30 was 8.1. It is interesting to
note that the rate of out-of-home placement at any point during the fiscal year is actually lower
than that for the Wave 1 managed care counties (6.1), even though their statistic includes any
child or children in out-of-home placement at any time during the fiscal year rather than on a
single day.

                                       Table 28
              Children in Out-of-Home Placement per 1,000 Child Population
                                                         Wave 2
     National      Comparison      Wave 1 counties’                    Other Large
                                                        counties’
      Rate        States’ Median       Mean                          counties’ Mean
                                                          Mean

                                             5.0 (1996)           7.5 (1996)         10.5 (1996)
                                             4.1 (1997)           6.4 (1997)          8.5 (1997)
         8.1                5.6              4.7 (1998)           6.8 (1998)          8.7 (1998)
                                             6.2 (1999)           9.1 (1999)         11.6 (1999)
                                             6.1 (2000)           8.6 (2000)         10.6 (2000)




William M. Mercer, Incorporated                244                 Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Indicator 6: Length of Out-of-Home Stay
This indicator is “What was the length of stay in all out-of-home placements for children closed
to the system?” We used the CWEST data provided by the CDHS to examine the total number of
months that children placed out of the home were in out-of-home care while they were opened to
the child welfare system.

Figure 121 shows the results from SFY 1996 through SFY 2000. Two trends are noteworthy.
First, the non-managed care counties consistently have evidenced lower lengths of stay from
before the implementation of managed care to after the implementation of managed care. At the
same time, the number of children per 1,000 in the county population who receive out-of-home
placement during the fiscal year has been consistently higher in non-managed care counties.
There is often a relationship between higher rates of out-of-home placement and lower lengths of
stay as the higher number averages out the more severe cases, which tend to predominate when
out-of-home placement rates drop.

Although Wave 1 managed care counties had a 43 percent higher average length of stay in
out-of-home placements than non-managed care counties in 1996, they had only a 27 percent
higher average length of stay in out-of-home placement in SFY 2000. The gap in length of stay
appears to be closing, suggesting that managed care counties are decreasing their length of stay
relative to non-managed care counties. While the pattern of change over time has been similar
for all groups (rising initially, then decreasing), the rate of increase and subsequent decrease has
been slightly sharper for managed care counties.

Second, for all large counties, the trend toward an increase in out-of-home length of stay that was
observed from SFY 1996 to SFY 1998 has reversed for managed care counties and has leveled
off, or slightly reversed, for non-managed care counties.




William M. Mercer, Incorporated                 245               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
                                    Figure 121
           Out-of-Home Placement Length of Stay (in Months) for Closed Cases

          20
          18
          16
          14
          12                                      10.1 9.9
          10                                                                             8.5
                                  7.7   8                          7.7 8.2         7.9
           8      7                                          6.8             6.9
                      6.1                   6.4                                                6.2
           6                4.9

           4
           2
           0
                  SFY 96           SFY 97           SFY 98          SFY 99          SFY 00

                            MC Wave 1              MC Wave 2                 Non-MC

Data are available for benchmarking this indicator from AFCARS October 2000, which contains
national benchmarks, and “Child Welfare Outcomes 1998,” which only contains data on the
median (as opposed to the mean) length of stay for children exiting care. In Table 29, at the
comparison state level, we present both the median length of stay (3.4 months), as reported in
Outcomes 1998, and the mean length of stay (LOS), as reported in 1997. Note that the mean, not
the median, was used in calculating the outcome indicator for Colorado counties. The median
tends to be lower than the mean for this variable because very long out-of-home episodes affect
(increase) the mean more than the median.

It should be noted that the benchmark indicators are not exactly the same as the indicators from
the Colorado managed care evaluation. In the Colorado evaluation, children’s length of stay in
out-of-home placement for their entire tenure as an open case was used, whereas the benchmark
data are based on specific foster care episodes (this is one episode only). This means that the
comparison between Colorado counties’ average lengths of stay and national and other state
benchmarks is biased in favor of other states.

Also of note is that benchmarks are based on data supplied by state child welfare agencies. States
may have different methods and processes for tracking placements and length of stay at
discharge so the results should be interpreted with some caution.




William M. Mercer, Incorporated                    246                Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Table 29 shows that the Colorado counties’ average length of stay in out-of-home placement
appears to be lower than the mean length of stay in out-of-home placement reported by other
states in the West and around the nation. This finding is surprising, since the benchmarks only
include one episode of out-of-home placement, whereas Colorado’s data include all of the
historical episodes of out-of-home placement for each child closed to the system.

                                        Table 29
                     Length of Stay in Months in Out-of-Home Care
                                         Wave 1          Wave 2
    National      Comparison States’                                              Other Large
                                        counties’       counties’
     Rate            Averages                                                    counties’ Mean
                                          Mean            Mean

                                             7.0 (1996)           6.1 (1996)          4.9 (1996)
  10.8 (Mean)     12.3 (Mean – 1997)         7.7 (1997)           8.0 (1997)          6.4 (1997)
                   3.4 (Median – 1998)      10.1 (1998)           9.9 (1998)          6.8 (1998)
                                             7.7 (1999)           8.2 (1999)          6.9 (1999)
                                             7.9 (2000)           7.5 (2000)          6.2 (2000)




William M. Mercer, Incorporated                247                 Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Indicator 7: Percentage in Positive Placements
The indicator is “For all children who were closed to the system, what percentage left to positive
placements?” We used CWEST data provided by CDHS to examine the percentage of children
who were in positive placements at case closure. We used the state’s definition of positive
placements, which includes the following: Parents, Certified Kinship Care, Guardian, Adoptive
Parent, Adoptive Parent with Subsidy, Foster Adoptive Parent, Kinship Guardianship, Kinship
Adoption without Subsidy, Kinship Subsidized Adoption, Kinship Custody, and Relatives.

Figure 122 shows the rates of positive placement at case closure for the managed care counties
and the other four large counties. The trends in the percentages of children who were in positive
placements at case closure are somewhat different for each of the three groups of counties. For
Wave 1 counties, the trend has remained flat from SFY 1997 to SFY 2000, increasing by only
.2 percent during that time (from 84.4 percent to 84.6 percent). For Wave 2 counties, the
percentage of positive placements at case closure increased from 81.6 percent in SFY 1997 to
87.8 percent in SFY 2000. Non-managed care counties evidenced a slight decrease in the
percentage of positive placements from SFY 1997 to SFY 2000 (86.2 percent to 83.8 percent).

                                           Figure 122
                       Percentage of Positive Placements at Case Closure
                        (Parents, Kin, Relatives, Adoption, Guardians)

        90
                                                                                       87.8
        88
        86                   85.2                                 85.3
               84.4                                84.5                         84.6
                                            83.5           83.5                               83.8
        84                           82.8
                      81.6
        82                                                               80.8
        80
        78
        76
                   SFY 97               SFY 98                SFY 99               SFY 00

                                    MC Wave 1        MC Wave 2           Non-MC




William M. Mercer, Incorporated                      248              Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
Benchmark data are available for this indicator. “Child Welfare Outcomes 1998” reports rate of
“reunification,” “guardianship,” and “adoption” at case closure and reports a national rate of
83 percent. We calculated the average rate for the group of seven comparison states used for this
study as 90 percent, using the Outcome 1998 data reported for each of those states. We believe
that the rates reported in “Child Welfare Outcomes 1998” are comparable to the Colorado data
and are useful as benchmarks for positive placement rates.

Benchmark comparisons are shown in Table 30. Managed care counties’ positive placement
rates appear slightly better than the national rate and slightly lower than the comparison counties’
rate.

                                         Table 30
                    Percentage of Positive Placements at Case Closure
                                          Wave 1          Wave 2
    National      Comparison States’                                              Other Large
                                         counties’       counties’
     Rate            Averages                                                    counties’ Mean
                                           Mean            Mean

                                            84.4 (1997)          81.6 (1997)         85.2 (1997)
       83%                 90%              82.8 (1998)          83.5 (1998)         84.5 (1998)
                                            83.5 (1999)          85.3 (1999)         81.8 (1999)
                                            84.6 (2000)          87.8 (2000)         83.8 (2000)




William M. Mercer, Incorporated                 249                Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
Indicator 8: Number of Adoptions
This indicator is “What is the number of adoptions per 1,000 children in the population?” In year
one of the evaluation, we used the AFCARS and CWEST data provided by the CDHS to
examine the number of adoptions per 1,000 children in the population. Data from the state
Demographer’s Office was again used as the source of the child population data. The adoption
data from AFCARS are reported on a FFY basis (October 1 to September 30). Only data from
FFY 1998 and FFY 1999 were available since July–September, 1997 data were not available to
make a complete SFY 1998 year. Only data from FFY 1998, FFY 1999, and FFY 2000 were
available.

Figure 123 shows the number of adoptions per 1,000 children in the population for the managed
care counties and the other four large counties in Colorado. The rates of adoptions per 1,000
children in the population are somewhat different for the three groups of Colorado counties.
Wave 2 counties continue to have more adoptions per 1,000 children in the population than do
the other two groups of counties. Wave 2 counties’ higher rate of adoptions is due entirely to
El Paso County’s high rate of 1.22 adoptions per 1,000 children in the population. Wave 1
counties’ adoption rate rose slightly in FFY 2000, while non-managed care counties continued a
slight increase in the rate of adoptions.

                                        Figure 123
                   Adoptions per 1,000 Children in the County Population

            1.2
                                                          0.98
              1             0.89
                                                                                     0.8
            0.8                                                  0.64                      0.66
                                   0.61
            0.6                                                               0.43
                     0.42                          0.38
            0.4
            0.2
              0
                         SFY 98                       SFY 99                     SFY 00

                                          Wave 1            Wave 2             Wave 3




William M. Mercer, Incorporated                     250                   Colorado Child Welfare Evaluation
                                                                 Second Interim Implementation Status Report
 Data on the number of adoptions per 1,000 children in the population are available through
 federal government reports. National level data are available through the AFCARS October 2000
 Report. The most recent data on a state-by-state basis are available through “Child Welfare
 Outcomes 1998.” States may vary in the ability of their information systems to report adoption
 data accurately and the results should be interpreted with some caution.

 Benchmark comparisons are shown in Table 31. The rate of adoptions per 1,000 children in the
 population is below the national rate for Wave 1 counties, higher for Wave 2 counties, and
 equivalent for non-managed care counties.

                                           Table 31
                         Adoptions per 1,000 Children in the Population
                           Comparison        Wave 1                Wave 2       Other Large
              National       States’      Counties’ Mean        Counties’ Mean Counties’ Mean
               Rate         Median*       (FFY 1998 and 1999)   (FFY 1998 and 1999)   (FFY 1998 and 1999)

Adoptions
                 .66            .52         .42 (1998)            .89 (1998)            .61 (1998)
 Per 1,000
               (1999)         (1998)        .38 (1999)            .98 (1999)            .64 (1999)
  Child
                                            .43 (2000)            .80 (2000)            .66 (2000)
Population




 William M. Mercer, Incorporated              251                 Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
Indicator 9: Time to Adoption
This indicator is “What is the time from the termination of parental rights to adoption?” We used
the AFCARS and CWEST data provided by CDHS to examine the time from termination of
parental rights to adoption. The fields in the data sets pertaining to the date of termination of
parental rights (TPR) and to the date of the adoption finalization were used. Only data from
FFY 1998, FFY 1999, and FFY 2000 were available.

Figure 124 summarizes the average time from termination of parental rights to finalized adoption
for children in the managed care counties and the other four large counties in Colorado. The
average number of months that children had to wait to be adopted decreased slightly for all three
county groups between FFY 1999 and FFY 2000.

                                     Figure 124
        Average Months from Termination of Parental Rights to Finalized Adoption

          25
                                                      21.6                     20.6
          20                  18.5
                                          17.4                   16.2

          15       12.7
                                               11.6
                       10.6                                             10.4
          10


            5


            0
                       FFY 98                 FFY 99                 FFY 00

                                  Wave 1 MC      Wave 2 MC        Non-MC


There was a large variability in time from termination of parental rights to adoption across the
managed care counties in FFY 2000. Boulder and El Paso counties had average waits of only 9.9
months and 9.2 months, respectively, while Mesa’s average wait was 22.5 months. Managed care
counties average number of months to adoption continued to be lower than the non-managed
care counties. It is interesting to note that El Paso had both the highest number of adoptions as
well as the lowest average months from termination of parental rights to adoption among all
large counties in Colorado in FFY 2000.




William M. Mercer, Incorporated                252              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Summary of Outcome Indicator Results
Trends in the outcome indicators often were consistent between the county groups. For the nine
outcome indicators analyzed for this report, the trends for the three groups of Colorado large
counties were very similar on most of them. Where the trends were different, the differences
were slight. On the basis of outcomes, it appears that managed care has not yet had a discernable
effect on outcomes, as measured in this evaluation, for children and families in Colorado.

The benchmarking results generally revealed favorable performance for all Colorado large
counties relative to national averages and comparison state averages. For example, there is a
lower rate of confirmed abuse and neglect in Colorado and Colorado children are less likely to be
in out-of-home placement during the year. Furthermore, Colorado children spend less time in
out-of-home placement. These findings are as true for the Wave 1 managed care counties as for
the other county groups studied. On this basis, it appears that managed care has not impacted,
either positively or negatively, Colorado’s apparent superior performance on outcomes,
compared to other states.

Outcome Indicator Results for the Prevention of Abuse
Our analyses indicated a slight rise in the rate of abuse and neglect per 1,000 children in the state
(Figure 116) for all three county groups studied. (This trend has also been seen nationally.) At
the same time, the rates of referrals and investigations for abuse and neglect have declined
(Figures 117 and 118). It is not yet clear whether abuse and neglect of children has increased in
Colorado’s large counties or whether the counties have become more likely to confirm abuse and
neglect cases that are investigated.

There was no discernable rise in the rates of recurrence of confirmed abuse and neglect of
children in the Wave 1 managed care counties (Figure 115). The rate increased only .1 percent
from SFY 1997 to SFY 1999 in the three Wave 1 counties. On this basis, it appears that managed
care has not yet had a discernable effect on the rates and recurrence of abuse and neglect among
Colorado children.

Outcome Indicator Results for Family Connection
Three different outcomes were used as indicators of family connection. The results on the first
indicator showed a decrease in the percentages of children residing with parents at case closure
in all three Colorado large county groups (Figure 119). The decrease was most pronounced for
the non-managed care counties.




William M. Mercer, Incorporated                 253               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
The other two indicators addressed out-of-home placement; both the number of children in out-
of-home placement per 1,000 children in the population (Figure 120) and the average length of
stay in out-of-home placement for children during their tenure in the child welfare system
(Figure 121). The results indicated a steady increase in the number of children placed out of the
home at some point during the fiscal year for all three county groups. Wave 1 managed care
county children had the largest percentage increase by a slight margin. The mean length of stay
in out-of-home care for children generally increased, also, except for the Wave 1 managed care
group, which saw no change.

The trends for all three outcome indicators in this domain are not encouraging, yet there is no
indication that managed care has anything to do with the trends. Subsequent years of the
evaluation should help reveal whether these trends begin to reverse, or not, and whether any
changes in the trends are different or are more or less pronounced for managed care counties in
comparison to the non-managed care large counties. On the basis of findings in this area, it
appears that managed care has not yet had a discernable effect on indicators of family connection
for children and their families in Colorado.

Outcome Indicator Results for Staying in a Permanent Home
The outcome indicators in this domain address the rates of positive placements at case closure
and adoption. Trends in the percentage of children in positive placements at case closure (Figure
122) were somewhat more favorable for the managed care counties, largely as a result of the
contribution of Wave 2 counties.

The adoption-related indicators evidenced greater differences between county groups than had
been seen, generally, with the other indicators. The Wave 2 managed care counties had better
performance on the two adoption indicators; they had both a higher rate of adoption per 1,000
children in the population (Figure 123) and a shorter number of average months from termination
of parental rights to adoption (Figure 124). In addition, their trends were more positive than for
the other two county groups. The Wave 2 counties’ performance was substantially boosted by El
Paso County, which put a particularly high priority on adoption services. It is noteworthy that the
Wave 1 managed care counties, while higher than Wave 2 counties, had a shorter number of
average months from termination of parental rights to adoption than non-managed care counties.

Although there is some evidence in this domain that managed care has had a favorable effect on
outcomes, the evidence is mixed. Managed care counties, as a group, do not show significantly
superior performance in rates of positive placements and adoptions when compared to non-
managed care counties. However, they are clearly better in time to adoption. Given that the
results are mixed, it cannot be concluded at this point that managed care has had a demonstrable
effect on the domain of maintaining a permanent living arrangement for children in Colorado.




William M. Mercer, Incorporated                254               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Cost Findings


Senate Bills 97-218 and 98-165 were drafted in response to sharp increases in child welfare
costs, primarily in the area of out-of-home care. A major assumption underlying these bills is
that costs will be contained. We tested this assumption through two series of analyses.

In the first series of analyses, we compared each managed care county’s allocation at the
beginning of SFY 2000 to its actual expenditures at the end of the year. These analyses were
conducted for both the child welfare block and CORE services funding. The results are displayed
in the individual county sections of this report in the first figure in the cost findings section.
More detail is provided in the second figure in that section.

These analyses show that all managed care counties, except Mesa County, spent more of their
child welfare block than their original allocations. Within the block, all counties, except for
Arapahoe and Jefferson, overspent their 80/20 allocations. All counties, except for Jefferson,
overspent CORE services in SFY 2000.

This picture changes with the close out at the end of SFY 2000. In the close out, funds are
transferred between budget categories to adjust for overexpenditures. For example, a county may
transfer TANF funds to Title XX expenditures within the child welfare block to bring the block
into balance. After the close out process, managed care counties must pay for any remaining
deficits with county funds.

After the close out, Arapahoe, Jefferson, and Mesa have either balanced their budgets or shown a
savings (Figure 125). The other three counties show a deficit. Deficits range from 1.37 percent of
the child welfare block to 14.62 percent of the block (see Figure 126).
                                         Figure 125
                           Managed Care Savings* by State Fiscal Year
                                                                                         $1,326,151
                         Breakeven




                                                                                                                                                                    $417,606
                                     Breakeven




                                                                                                                                                         $278,915




                                                                                                                                                                                              $268,578
                                                                                                                                  $174,604
                                                                                                                      Breakeven


                                                                                                                                             Breakeven




     $2,500,000
                                                           $69,111
                                                 $30,287




                                                                                                                                                                               $16,503




     $1,500,000
                    NA




                                                                                    NA




                                                                                                                                                                                         NA




       $500,000
      $(500,000)    Arapahoe                         Boulder                         El Paso                            Jefferson                                   Mesa                  Pueblo
                                                                                                                                                                                                         $(201,455)




     $(1,500,000)
                                                                     $(1,449,936)




     $(2,500,000)
     $(3,500,000)
                                                                                                      $(3,313,282)




     $(4,500,000)

      NA = Prior to Managed Care Implementation
            * Savings exclude CORE Service funds                                                                        SFY1998                               SFY1999                     SFY2000




William M. Mercer, Incorporated                                                          255                                  Colorado Child Welfare Evaluation
                                                                                                                     Second Interim Implementation Status Report
                                      Figure 126
               Managed Care Savings as a Percentage of Child Welfare Block
                                 Allocation (SFY2000)
           10.00%


            5.00%

                    Breakeven                          Breakeven    0.20%
            0.00%
                    Arapahoe      Boulder   El Paso    Jefferson    Mesa       Pueblo
           -5.00%                                                              -1.37%


          -10.00%
                                            -10.85%
          -15.00%
                                  -14.62%

          -20.00%



It is apparent that managed care counties, as a group, did not have unspent funds in SFY 2000.
They did have unspent funds in previous years but this was more than offset by deficits in FY
2000. Of note, though, is that two of the six managed care counties, Boulder and El Paso, were
responsible for most of the deficit in FY 2000.

A key factor in containing or reducing cost is a shift from the use of high cost services, such as
residential treatment centers, to services that are less costly. There are two key elements for
managing the use of services. First is utilization management, the close tracking of high cost
services to ensure that they are used only when needed, and in no more than the amount that is
required. Second is an array of lower cost services that can provide an effective alternative to
high cost care. We conducted three series of cost analyses to study these elements.

In our first series of analyses, we compared the ratio of CORE service expenditures to selected
out-of-home placement expenditures for each county. This ratio is meant to be a rough index of
community service effort. As counties develop more community programs, this ratio would be
expected to increase. It should be cautioned that this ratio does not account for community
programs that might be developed with other sources of funds; over the course of the evaluation,
it is hoped that it can be refined to include these funds.




William M. Mercer, Incorporated                 256               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
The ratios for each managed care and each non-managed care county for FY 2000 are shown in
Figures 127 and 128. Actual expenditures are displayed in the graphs as well as the ratio. The
amount of variation is striking, ranging from Boulder at 55 percent (highest index of community
service effort) to Pueblo at 15 percent (lowest index of community service effort). Associated
with this is that Boulder’s placement rate and length of stay in out-of-home care are well below
average (Table 6), while Pueblo’s placement rate is almost three times that of other counties and
its length of stay is slightly below average (Table 10). In addition, Boulder’s utilization
management process is the most sophisticated of the managed care counties.
                                    Figure 127
            Selected Out-of-Home Placement Compared to CORE Service
                   Expenditures, Managed Care Counties (FY2000)
       Pueblo                                   $6,141,055                                    $1,044,884


        Mesa                               $2,910,274                                     $951,309


    Jefferson                      $4,567,144                                    $3,075,451


       El Paso                       $11,613,167                                    $6,052,659


      Boulder                $2,043,731                                   $2,459,513


    Arapahoe                         $6,526,749                                    $3,698,242


               0%    10%     20%      30%       40%     50%        60%     70%      80%       90%    100%
  Out-of-Home expenditures include Family Foster Care, Specialized Group Care, RCCF, RTC, and Relative
  F t C
                                     Selected OOH            CORE Services


                                       Figure 128
                 Selected Out-of-Home Placement Compared to CORE Service
                    Expenditures, Non Managed Care Counties (SFY2000)
        Weld                              $3,925,646                                   $1,326,669


      Larimer                         $2,995,983                                    $1,385,418


      Denver                              $17,123,427                                  $5,883,332


      Adams                        $6,510,261                                    $4,532,911

             0%     10%      20%      30%       40%     50%        60%     70%      80%       90%    100%


  Out-of-Home expenditures include Family Foster Care, Specialized Group Care, RCCF, RTC, and Relative Foster C

                                      Selected OOH           CORE Services




William M. Mercer, Incorporated                              257                  Colorado Child Welfare Evaluation
                                                                         Second Interim Implementation Status Report
Figures 129 and 130 show changes in the ratio of CORE service expenditures to selected out-of-
home placement expenditures over the course of three years. In examining these trends, SFY
1999 should be viewed with caution. There is a large decrease in SFY 1999 expenditures for
CORE services that would not be expected and that cannot be explained by the CDHS. This
decrease reduces all of the ratios for SFY 1999. As a result, we have limited our discussion in
this section to SFY 1998 and SFY 2000.

                                       Figure 129
               Ratio of CORE Service Expenditures to Selected Out-of-Home
                        Placement Expenditures by State Fiscal Year,
                                 Managed Care Counties


       0.70
                                             0.56


                                                           0.55




       0.60
                                                    0.44




                                                                                            0.42


                                                                                                          0.40
                               0.36




       0.50
                0.35




                                                                                                   0.33
                                                                                   0.34




                                                                                                                                                                                     0.34
                                                                                                                                                                       0.30
       0.40




                                                                                                                                 0.25
                       0.21




                                                                                                                                                                              0.21
                                                                     0.23




                                                                                                                   0.22
                                                                            0.17




                                                                                                                          0.17
       0.30




                                                                                                                                                         0.15
                                                                                                                                           0.13
                                                                                                                                                  0.09
       0.20
       0.10
       0.00
                Arapahoe                      Boulder                    El Paso             Jefferson                Mesa                   Pueblo                    Managed
                                                                                                                                                                        Care
                                           Out-of-Home expenditures include Family Foster Care,                                                                        Counties
                                        Specialized Group Care, RCCF, RTC, and Relative Foster Care.


                                                           SFY1998                        SFY1999                    SFY2000


                                       Figure 130
               Ratio of CORE Service Expenditures to Selected Out-of-Home
              Placement Expenditures by State Fiscal Year, Non-Managed Care
      0.60                              Counties
      0.50
                                      0.41
               0.34




                                                                                          0.34




      0.40
                                                                                                            0.32




                                                                                                                                                                0.31
                                                      0.30




                                                                                                                                                                                     0.30
                                                                                                   0.26
                                                                            0.26




                                                                                                                                             0.25




      0.30
                        0.23




                                                                                                                          0.20




                                                                                                                                                                          0.20
                                                                  0.19




                                                                                                                                    0.14




      0.20

      0.10

      0.00
                       Adams                                 Denver                              Larimer                          Weld                     All Other Top 10
                                                                                                                                                           Large Counties
       Out-of-Home expenditures include Family Foster Care, Specialized Group
       Care, RCCF, RTC, and Relative Foster Care.

                                                              SFY1998                        SFY1999                       SFY2000


William M. Mercer, Incorporated                                                              258                          Colorado Child Welfare Evaluation
                                                                                                                 Second Interim Implementation Status Report
In comparing SFY 1998 and SFY 2000 in Figures 129 and 130, Boulder and Pueblo’s positions
remain the same each year. Of note, is El Paso’s significant increase in SFY 2000. A relationship
would be expected between this change and placement rate and length of stay but this does not
appear to be the case (Figures 44 and 45). Also of note, is that there is a slight increase in the
ratio for all managed care counties, while there is a decrease for non-managed care counties.

The second series of analyses examines the utilization and cost of selected out-of-home
placements over time. Specifically, we examined the effect of managed care on the cost of:

   CPA foster care;
   CPA specialized group care;
   county foster care;
   county specialized group care;
   RCCF;
   RTC;
   relative foster care; and
   subsidized adoption.

We examined utilization and costs for five years (SFY 1996 through SFY 2000) for the six
managed care counties and the four non-managed care comparison counties. To provide as
complete a picture as possible, annual expenditures, average monthly caseload, and average
monthly cost per child were analyzed.

Tables 32 and 33 illustrate the results of these analyses, while Tables 34 through 41 provide
detailed results. The tables introduce a new term, FPE, which is a “full-time placement
equivalent.” It is the equivalent of one child staying in a residential placement for one month
(30.5 days). If two children stay in a residence for a little over 15 days each during a month, they
would be counted as one FPE, as both combined stayed for 30.5 days.

Table 32 shows that the large increases in client FPEs that were occurring prior to the advent of
managed care in counties have, in general, diminished. For example, the rate of increase in RTC
FPEs has decreased in five counties and remained the same in the other. Of note is that increases
in relative foster care and subsidized adoption are positive findings.

Table 33 is of particular interest. It shows that the rate of growth in cost per FPE for CPA foster
care and specialized group care is less than that for the same services operated by counties. RTCs
show a high rate of growth in cost across almost all counties after the introduction of managed
care.




William M. Mercer, Incorporated                 259               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
         Table 32: Comparison of Annual Trend of Monthly FPE Client1 for Selected Out-of-Home
                                  Services, Managed Care counties
                           Arapahoe *               El Paso *               Pueblo *                Boulder **         Jefferson **           Mesa **
       Service/
                           Pre        Post        Pre        Post         Pre         Post          Pre      Post       Pre      Post       Pre       Post
       County            Managed     Managed    Managed     Managed     Managed      Managed      Managed   Managed   Managed   Managed   Managed    Managed
                          Care        Care       Care        Care        Care         Care         Care      Care      Care      Care      Care       Care


CPA Foster Care             ٨           ↑         ↑↑           ■           ■           ■            ٧        ↑↑         ↑↑         ■         ■         ■
CPA Specialized
Group Care                  ■          ↓↓          ↓           ■          ↓↓           ■            ↓         ↓         ↓↓        ↓↓         ■         ■

County Foster Care          ■           ■          ■           ↓          ↓↓          ↑↑            ■         ■         ↓          ■         ■         ■
County Specialized
Group Care                  ■           ■         ↓↓           ↑          ↓↓          ↓↓           ↑↑         ↓         ↓          ■         ■         ↓
RCCF                        ■           ■          ↓           ↓           ■           ■           ↓↓         ↑         ■          ↓        ↑↑        ↑↑
Residential
Treatment Center           ↑↑           ↓          ↑           ■           ↑           ■            ↑         ↑         ↑↑         ↑        ↑↑         ■
Relative Foster
Care                        ↑           ■         ↑↑           ↑          ↑↑           ↓            ↑         ↑         ↓↓        ↑↑        NC         ■
Subsidized
Adoption                    ↑           ↑          ↑          ↑↑           ↑           ■           ↑↑         ↑         ↑↑         ↑         ↑         ↑
1
     Based on available monthly data for the period from 7/1/95 to 6/30/00.
*
     Managed care implementation started in SFY 1999 for Arapahoe, El Paso, and Pueblo counties.
**
     Managed care implementation started in SFY 1998 for Boulder, Jefferson, and Mesa counties.

↑↑ or ↓ = An annual increase/decrease of 30% or more
↑ or ↓ = An annual increase/decrease between 10% - 29%
■ = An annual increase/decrease of less than 10%
NC = Not calculated due to insufficient data

William M. Mercer, Incorporated                                                260                                    Colorado Child Welfare Evaluation
                                                                                                             Second Interim Implementation Status Report
Table 33: Comparison of Annual Trend of Average Monthly Cost Per FPE Client1 for Selected
Out-of-Home Services, Managed Care Counties
                           Arapahoe *               El Paso *               Pueblo *                Boulder **         Jefferson **           Mesa **
       Service/
                           Pre        Post        Pre        Post         Pre         Post          Pre      Post       Pre      Post       Pre       Post
       County            Managed     Managed    Managed     Managed     Managed      Managed      Managed   Managed   Managed   Managed   Managed    Managed
                          Care        Care       Care        Care        Care         Care         Care      Care      Care      Care      Care       Care


CPA Foster Care             ■           ■          ■           ■           ↓           ■            ■         ■         ■          ↓         ■         ■
CPA Specialized
Group Care                  ■           ■          ■           ■           ■           ■            ■         ↓         ■          ■         ■         ■

County Foster Care          ■           ↑          ↑           ↑           ↓           ↑            ↑         ↑         ■          ↑         ↑         ■
County Specialized
Group Care                  ■           ■          ↑           ■           ↑           ■           ↑↑         ↑         ↑          ■        ↑↑        ↑↑
RCCF                       ↓↓           ■         ↓↓           ■           ↓          ↑↑           ↓↓         ■         ↓          ■        ↓↓         ↑
Residential
Treatment Center            ■           ↑          ■           ↑           ■           ↑            ■         ■         ■          ■         ↓         ↑
Relative Foster
Care                        ■           ↑          ↑           ■           ■           ↓            ↓         ↑         ■          ■        NC         ↑
Subsidized
Adoption                    ■           ■          ↑           ↑           ■           ■            ■         ■         ↑          ■         ■         ■

1
     Based on available monthly data for the period from 7/1/95 to 6/30/00.
*
     Managed care implementation started in SFY 1999 for Arapahoe, El Paso, and Pueblo counties.
**
     Managed care implementation started in SFY 1998 for Boulder, Jefferson, and Mesa counties.

↑↑ or ↓↓ = An annual increase/decrease of 30% or more
↑ or ↓ = An annual increase/decrease between 10% - 29%
■ = An annual increase/decrease of less than 10%
NC = Not calculated due to insufficient data

William M. Mercer, Incorporated                                                261                                    Colorado Child Welfare Evaluation
                                                                                                             Second Interim Implementation Status Report
Table 34: CPA Family Foster Care by State Fiscal Year

                                                       Annual Expenditure                                        Average Monthly FPE Clients                  Average Monthly Cost Per FPE Client
        County
                        SFY96            SFY97              SFY98             SFY99             SFY00         SFY96   SFY97   SFY98   SFY99   SFY00       SFY96         SFY97         SFY98         SFY99         SFY00
 Arapahoe           $    1,457,280   $     2,138,372    $     2,420,381   $     2,577,608   $     2,762,907      68     102     125     141     161   $     1,777   $     1,742   $     1,612   $     1,525   $     1,428
 Boulder            $     563,087    $      466,014     $      223,156    $      212,602    $      442,530       30      25      12      13      23   $     1,564   $     1,561   $     1,545   $     1,367   $     1,564
 El Paso            $    5,348,368   $     7,719,516    $     7,831,447   $     7,012,258   $     7,644,237     251     381     441     434     442   $     1,774   $     1,700   $     1,482   $     1,347   $     1,441
 Jefferson          $     422,097    $     1,010,317    $     1,004,229   $      693,408    $      851,803       20      43      48      44      52   $     1,732   $     1,942   $     1,744   $     1,317   $     1,361
 Mesa               $    1,415,047   $     1,240,789    $     1,308,964   $     1,076,423   $     1,185,570      60      61      70      57      59   $     1,959   $     1,690   $     1,556   $     1,587   $     1,693
 Pueblo             $    5,904,008   $     5,694,005    $     4,533,369   $     4,659,167   $     4,874,025     254     292     257     260     280   $     1,939   $     1,627   $     1,471   $     1,495   $     1,451
 All Managed Care
                    $   15,109,887   $    18,269,014    $    17,321,546   $    16,231,466   $    17,761,073     114     151     159     158     170   $     1,791   $     1,710   $     1,568   $     1,440   $     1,490
 Counties
 Adams              $    1,795,674   $     2,518,064    $     3,224,190   $     3,372,726   $     3,595,125      91     129     170     186     210   $     1,648   $     1,623   $     1,585   $     1,511   $     1,425
 Denver             $    7,661,031   $     8,933,983    $     9,244,989   $     9,875,190   $    10,028,168     405     446     474     509     505   $     1,575   $     1,669   $     1,624   $     1,618   $     1,654
 Larimer            $     841,287    $      726,323     $     1,155,160   $     1,493,202   $     1,276,096      40      35      57      70      63   $     1,737   $     1,721   $     1,685   $     1,767   $     1,697
 Weld               $     793,883    $     1,471,208    $     1,487,369   $     1,872,925   $     2,181,407      34      64      66      84      93   $     1,906   $     1,917   $     1,864   $     1,860   $     1,961
 All Other Top 10
                    $   11,091,874   $    13,649,577    $    15,111,708   $    16,614,043   $    17,080,795     142     169     192     212     218   $     1,716   $     1,732   $     1,689   $     1,689   $     1,684
 Counties

 Total/Average      $   26,201,761   $    31,918,591    $    32,433,253   $    32,845,509   $    34,841,869     125     158     172     180     189   $     1,761   $     1,719   $     1,617   $     1,539   $     1,567



Table 35: CPA Specialized Group Care by State Fiscal Year

                                                       Annual Expenditure                                        Average Monthly FPE Clients                  Average Monthly Cost Per FPE Client
        County
                        SFY96            SFY97              SFY98             SFY99             SFY00         SFY96   SFY97   SFY98   SFY99   SFY00       SFY96         SFY97         SFY98         SFY99         SFY00
 Arapahoe           $     304,711    $      332,903     $      246,140    $      192,098    $       77,277       14      15      12       9       3   $     1,785   $     1,924   $     1,728   $     1,749   $     1,924
 Boulder            $     273,133    $      199,679     $       97,007    $       65,575    $       36,357       15      11       8       7       4   $     1,552   $     1,482   $      977    $      756    $      630
 El Paso            $    1,081,653   $      979,513     $      850,886    $     1,050,235   $     1,102,384      52      43      36      40      40   $     1,731   $     1,888   $     1,961   $     2,175   $     2,310
 Jefferson          $     450,843    $      239,734     $      251,731    $       76,407    $       96,129       20      11      13       5       5   $     1,874   $     1,783   $     1,567   $     1,161   $     1,511
 Mesa               $     210,501    $      247,000     $      227,312    $      257,696    $      222,146       10      12      12      15      11   $     1,724   $     1,672   $     1,574   $     1,398   $     1,723
 Pueblo             $     638,186    $      316,430     $      181,700    $      340,836    $      312,411       29      13       9      15      15   $     1,826   $     1,966   $     1,772   $     1,899   $     1,724
 All Managed Care
                    $    2,959,028   $     2,315,258    $     1,854,776   $     1,982,846   $     1,846,704      23      18      15      15      13   $     1,749   $     1,786   $     1,597   $     1,523   $     1,637
 Counties
 Adams              $     421,621    $      467,791     $      468,642    $      358,457    $      372,695       20      21      22      17      17   $     1,775   $     1,851   $     1,767   $     1,761   $     1,779
 Denver             $    1,287,353   $      993,757     $      818,554    $      747,717    $      432,496       62      48      39      35      19   $     1,731   $     1,713   $     1,729   $     1,777   $     1,903
 Larimer            $     119,165    $       74,208     $       74,313    $       91,489    $       39,438        7       4       4       4       2   $     1,493   $     1,652   $     1,421   $     1,676   $     1,445
 Weld               $     157,358    $      161,892     $      166,989    $      177,430    $      252,158        8       7       7       6       9   $     1,666   $     1,827   $     2,019   $     2,411   $     2,400
 All Other Top 10
                    $    1,985,497   $     1,697,648    $     1,528,497   $     1,375,093   $     1,096,787      24      20      18      15      12   $     1,666   $     1,761   $     1,734   $     1,906   $     1,882
 Counties

 Total/Average      $    4,944,525   $     4,012,906    $     3,383,274   $     3,357,939   $     2,943,491      24      19      16      15      13   $     1,716   $     1,776   $     1,652   $     1,676   $     1,735




William M. Mercer, Incorporated                                                                   262                                                      Colorado Child Welfare Evaluation
                                                                                                                                                  Second Interim Implementation Status Report
Table 36: County Family Foster Care by State Fiscal Year

                                                          Annual Expenditure                                           Average Monthly FPE Clients                      Average Monthly Cost Per FPE Client
        County
                        SFY96              SFY97               SFY98             SFY99             SFY00         SFY96     SFY97     SFY98   SFY99   SFY00       SFY96           SFY97           SFY98         SFY99         SFY00
 Arapahoe           $     836,933      $      843,271      $     1,082,529   $     1,167,437   $     1,404,377     159       152       185     188     195   $         441   $         458   $      487    $      519    $      601
 Boulder            $     398,582      $      428,530      $      515,495    $      490,148    $      635,392       96        90        95      95      97   $         347   $         396   $      446    $      426    $      548
 El Paso            $     921,208      $      902,060      $     1,052,963   $     1,173,692   $     1,079,758     170       140       138     132     105   $         451   $         537   $      627    $      738    $      859
 Jefferson          $     766,580      $      748,809      $      930,109    $     1,383,131   $     1,587,673     171       146       147     173     162   $         374   $         428   $      518    $      664    $      816
 Mesa               $     456,590      $      586,857      $      526,906    $      532,739    $      576,526      112       122       110     108      98   $         338   $         402   $      398    $      412    $      492
 Pueblo             $     480,889      $      229,000      $      130,848    $      119,323    $      236,023       37        27        17      14      22   $     1,071     $         686   $      652    $      673    $      935
 All Managed Care
                    $    3,860,782     $    3,738,526      $     4,238,849   $     4,866,470   $     5,519,749     124       113       116     119     113   $         504   $         485   $      521    $      572    $      709
 Counties
 Adams              $    1,204,761     $    1,117,459      $     1,105,863   $     1,077,257   $     1,026,599     247       213       198     180     155   $         407   $         437   $      465    $      499    $      551
 Denver             $    1,806,968     $    1,597,890      $     1,540,503   $     1,845,815   $     2,288,205     411       347       300     293     284   $         366   $         384   $      429    $      524    $      671
 Larimer            $     508,138      $      586,738      $      621,433    $      524,434    $      739,118      114       112       115      87      90   $         373   $         438   $      453    $      503    $      683
 Weld               $     624,118      $      454,060      $      419,893    $      468,054    $      516,310      124        85        78      91      89   $         422   $         443   $      447    $      429    $      478
 All Other Top 10
                    $    4,143,985     $    3,756,147      $     3,687,693   $     3,915,559   $     4,570,232     224       189       173     163     155   $         392   $         426   $      449    $      489    $      596
 Counties

 Total/Average      $    8,004,766     $    7,494,673      $     7,926,542   $     8,782,029   $    10,089,980     164       143       138     136     130   $         459   $         461   $      492    $      539    $      663




Table 37: County Specialized Group Care by State Fiscal Year

                                                          Annual Expenditure                                           Average Monthly FPE Clients                     Average Monthly Cost Per FPE Client
        County
                        SFY96              SFY97               SFY98             SFY99             SFY00         SFY96     SFY97     SFY98   SFY99   SFY00       SFY96           SFY97           SFY98         SFY99         SFY00
 Arapahoe           $     336,023      $      383,366      $      344,934    $      336,071    $      371,252       37        37        35      31      33   $         754   $         855   $      824    $      895    $      948
 Boulder            $     149,226      $      167,152      $      103,592    $      126,941    $      193,242       17        11        10      11      12   $         783   $     1,227     $      899    $      981    $     1,303
 El Paso            $     143,232      $      222,191      $       44,019    $       58,921    $       47,636       13        20         4       4       3   $         922   $         924   $     1,054   $     1,272   $     1,327
 Jefferson          $     501,375      $      521,430      $      457,224    $      346,213    $      306,593       56        50        47      36      27   $         752   $         868   $      809    $      796    $      935
 Mesa               $         45,672   $         94,057    $       88,502    $      250,156    $      375,802          6         9      10      21      22   $         610   $         855   $      735    $     1,010   $     1,402
 Pueblo             $         74,878   $         78,853    $      100,228    $       88,140    $       93,503          8         8       9       8       8   $         735   $         786   $      890    $      896    $     1,022
 All Managed Care
                    $    1,250,406     $    1,467,050      $     1,138,499   $     1,206,440   $     1,388,028      23        23        19      18      18   $         759   $         919   $      869    $      975    $     1,156
 Counties
 Adams                   NC                 NC             $       27,144    $       24,349    $       27,720     NC        NC           2       1       1        NC              NC         $     1,575   $     1,948   $     2,031
 Denver             $     172,315      $      166,572      $      178,357    $      151,085    $      150,855       18        19        18      17      14   $         776   $         743   $      812    $      743    $      896
 Larimer            $     106,907      $      104,502      $      107,658    $      116,810    $      141,550       12        11        10      11      10   $         754   $         806   $      859    $      895    $     1,187
 Weld               $         94,181   $         74,784    $      107,055    $      218,505    $      163,256          7         7      10      19      15   $     1,115     $         894   $      897    $      947    $      885
 All Other Top 10
                    $     373,403      $      345,858      $      420,214    $      510,750    $      483,381       12        12        10      12      10   $         882   $         814   $     1,036   $     1,133   $     1,250
 Counties

 Total/Average      $    1,623,809     $    1,812,908      $     1,558,713   $     1,717,190   $     1,871,409      19        19        15      16      15   $         800   $         884   $      935    $     1,038   $     1,193




William M. Mercer, Incorporated                                                                       263                                                         Colorado Child Welfare Evaluation
                                                                                                                                                         Second Interim Implementation Status Report
Table 38: Residential Child Care Facility (RCCF) by State Fiscal Year

                                                       Annual Expenditure                                         Average Monthly FPE Clients                  Average Monthly Cost Per FPE Client
        County
                        SFY96             SFY97              SFY98             SFY99             SFY00         SFY96   SFY97   SFY98   SFY99   SFY00       SFY96         SFY97         SFY98         SFY99         SFY00
 Arapahoe           $      992,418    $      485,537    $       406,874    $      356,101    $      362,748       28      25      29      26      24   $     2,952   $     1,613   $     1,111   $     1,130   $     1,290
 Boulder            $      551,233    $      536,238    $       519,240    $      461,847    $      233,362       14      19      18      19      10   $     3,529   $     2,347   $     2,304   $     1,987   $     1,899
 El Paso            $     1,655,084   $      887,410    $       356,283    $      376,760    $      460,987       50      38      24      18      21   $     2,783   $     1,898   $     1,221   $     1,819   $     1,779
 Jefferson          $     1,055,540   $      739,662    $       364,302    $      518,568    $      389,365       25      21      23      20      21   $     3,541   $     2,923   $     1,353   $     2,125   $     1,376
 Mesa               $     1,049,238   $      565,544    $       248,375    $      154,027    $      229,292       19      19      13      10       9   $     4,647   $     2,435   $     1,643   $      977    $     2,221
 Pueblo             $      696,335    $      207,136    $       277,214    $      186,816    $      155,486       19       8      11       8       5   $     3,037   $     2,229   $     2,213   $     1,929   $     2,647
 All Managed Care
                    $     5,999,849   $    3,421,527    $      2,172,288   $     2,054,119   $     1,831,241      26      22      20      17      15   $     3,415   $     2,241   $     1,641   $     1,661   $     1,869
 Counties
 Adams              $      799,868    $      304,151    $       248,620    $      215,530    $      190,410       17      11      10      10       8   $     3,873   $     2,291   $     2,059   $     1,769   $     1,799
 Denver             $     2,498,279   $    1,295,661    $       671,545    $      681,506    $      864,630       76      52      37      33      39   $     2,822   $     2,040   $     1,461   $     1,712   $     1,832
 Larimer            $      870,654    $      615,136    $       294,405    $      258,269    $      193,427       24      17      13      13       9   $     3,052   $     2,966   $     1,827   $     1,706   $     1,752
 Weld               $      473,703    $      226,498    $       131,891    $       39,905    $       20,346       19      10      10       7       5   $     2,126   $     1,862   $     1,073   $      390    $      307
 All Other Top 10
                    $     4,642,504   $    2,441,446    $      1,346,461   $     1,195,210   $     1,268,813      34      22      18      16      15   $     2,968   $     2,290   $     1,605   $     1,394   $     1,422
 Counties

 Total/Average      $    10,642,353   $    5,862,973    $      3,518,748   $     3,249,329   $     3,100,053      29      22      19      16      15   $     3,236   $     2,260   $     1,626   $     1,554   $     1,690




Table 39: Residential Treatment Center (RTC) by State Fiscal Year (Room and Board only)

                                                       Annual Expenditure                                         Average Monthly FPE Clients                  Average Monthly Cost Per FPE Client
        County
                        SFY96             SFY97              SFY98             SFY99             SFY00         SFY96   SFY97   SFY98   SFY99   SFY00       SFY96         SFY97         SFY98         SFY99         SFY00
 Arapahoe           $      640,604    $    1,006,050    $      1,232,868   $     1,290,748   $     1,253,116      64      93     133     149     128   $      825    $      901    $      775    $      719    $      815
 Boulder            $      193,097    $      217,087    $       204,358    $      170,818    $      357,319       26      30      24      19      35   $      598    $      590    $      671    $      720    $      868
 El Paso            $      698,216    $      884,158    $       976,120    $      571,957    $      655,039       78     103     110      72      69   $      734    $      717    $      739    $      666    $      792
 Jefferson          $      471,205    $      587,741    $       707,156    $      739,994    $     1,231,189      48      66      84     103     126   $      811    $      742    $      699    $      598    $      813
 Mesa               $      124,498    $      221,761    $       190,185    $      265,028    $      224,718       13      25      25      25      23   $      836    $      738    $      621    $      897    $      817
 Pueblo             $      276,039    $      377,875    $       378,648    $      428,575    $      438,352       29      41      41      48      43   $      796    $      758    $      776    $      744    $      851
 All Managed Care
                    $     2,403,659   $    3,294,671    $      3,689,335   $     3,467,119   $     4,159,734      43      60      70      69      71   $      767    $      741    $      713    $      724    $      826
 Counties
 Adams              $      507,988    $      844,519    $       826,771    $      733,913    $     1,037,983      47      87     100      91     106   $      900    $      815    $      691    $      673    $      815
 Denver             $     2,035,779   $    2,074,452    $      2,197,826   $     2,270,301   $     2,551,473     177     218     240     254     290   $      958    $      791    $      766    $      744    $      740
 Larimer            $      425,213    $      645,385    $       383,146    $      388,763    $      436,152       49      74      53      47      45   $      726    $      728    $      602    $      686    $      808
 Weld               $      338,189    $      564,038    $       462,994    $      428,202    $      704,972       34      60      53      50      77   $      849    $      783    $      723    $      712    $      762
 All Other Top 10
                    $     3,307,169   $    4,128,395    $      3,870,738   $     3,821,179   $     4,730,580      77     110     111     110     130   $      858    $      779    $      696    $      704    $      781
 Counties

 Total/Average      $     5,710,828   $    7,423,066    $      7,560,073   $     7,288,298   $     8,890,314      57      80      86      86      94   $      803    $      756    $      706    $      716    $      808




William M. Mercer, Incorporated                                                                    264                                                     Colorado Child Welfare Evaluation
                                                                                                                                                  Second Interim Implementation Status Report
Table 40: Relative Foster Care by State Fiscal Year

                                                            Annual Expenditure                                              Average Monthly FPE Clients                   Average Monthly Cost Per FPE Client
        County
                        SFY96               SFY97                 SFY98             SFY99               SFY00         SFY96     SFY97     SFY98   SFY99   SFY00       SFY96        SFY97        SFY98        SFY99        SFY00
 Arapahoe           $      202,304      $      305,195       $       335,623    $      291,673    $        295,072       46        66        75      63      57   $      362   $      385   $      375   $      390   $      429
 Boulder            $       51,048      $       53,876       $        78,593    $      102,247    $        145,529       15        20        17      20      22   $      278   $      219   $      391   $      416   $      552
 El Paso            $       43,086      $      198,258       $       346,176    $      477,034    $        623,126          9      43        55      57      72   $      363   $      389   $      517   $      688   $      724
 Jefferson          $       61,826      $       28,306       $        42,576    $      111,029    $        104,391       16           7       8      15      19   $      310   $      315   $      401   $      600   $      475
 Mesa               $             -     $             -      $        69,189    $       68,283    $         96,219      -         -          16      16      18   $      -     $      -     $      352   $      350   $      455
 Pueblo             $       26,675      $       33,578       $        44,100    $       43,685    $         31,255          6         8       9       9       7   $      356   $      356   $      391   $      410   $      340
 All Managed Care
                    $      384,938      $      619,212       $       916,258    $     1,093,952   $       1,295,591      16        24        30      30      32   $      278   $      277   $      404   $      476   $      496
 Counties
 Adams              $      109,076      $      191,143       $       200,590    $      247,555    $        259,728       26        41        43      55      53   $      348   $      386   $      388   $      373   $      410
 Denver             $      453,605      $      696,269       $       821,281    $      754,675    $        807,601      106       153       181     165     150   $      356   $      379   $      378   $      381   $      449
 Larimer            $           5,223   $           7,616    $        27,845    $       70,813    $        170,203          1         2       6      16      32   $      312   $      359   $      354   $      370   $      449
 Weld               $       61,360      $       50,377       $        34,138    $       47,911    $         87,198       14        11         8      10      16   $      355   $      384   $      365   $      386   $      453
 All Other Top 10
                    $      629,264      $      945,405       $      1,083,855   $     1,120,953   $       1,324,729      37        52        60      62      63   $      343   $      377   $      371   $      377   $      441
 Counties

 Total/Average      $     1,014,202     $    1,564,618       $      2,000,113   $     2,214,905   $       2,620,319      24        35        42      43      44   $      304   $      317   $      391   $      436   $      474




Table 41: Subsidized Adoption Services by State Fiscal Year

                                                            Annual Expenditure                                              Average Monthly FPE Clients                   Average Monthly Cost Per FPE Client
        County
                        SFY96               SFY97                 SFY98             SFY99               SFY00         SFY96     SFY97     SFY98   SFY99   SFY00       SFY96        SFY97        SFY98        SFY99        SFY00
 Arapahoe           $      779,101      $    1,078,632       $      1,319,390   $     1,598,587   $       1,930,761     186       240       282     334     376   $      348   $      374   $      390   $      399   $      428
 Boulder            $      315,197      $      456,574       $       592,846    $      678,368    $        791,793       82       114       139     155     175   $      321   $      335   $      354   $      365   $      377
 El Paso            $      845,311      $    1,047,463       $      1,488,520   $     3,198,577   $       5,273,488     226       257       311     498     713   $      311   $      340   $      395   $      530   $      615
 Jefferson          $      508,854      $      778,818       $       959,054    $     1,104,039   $       1,294,771     146       196       228     256     292   $      289   $      331   $      351   $      359   $      369
 Mesa               $      260,977      $      302,619       $       382,232    $      480,650    $        636,361       67        74        91     108     135   $      323   $      339   $      351   $      369   $      392
 Pueblo             $      634,334      $      825,904       $       982,133    $     1,140,244   $       1,346,503     155       188       222     250     273   $      342   $      365   $      369   $      380   $      411
 All Managed Care
                    $     3,343,773     $    4,490,010       $      5,724,174   $     8,200,465   $      11,273,678     144       178       212     267     327   $      322   $      347   $      368   $      400   $      432
 Counties
 Adams              $      892,892      $    1,081,314       $      1,315,501   $     1,514,732   $       1,935,156     232       263       304     348     413   $      321   $      343   $      360   $      362   $      390
 Denver             $     1,958,468     $    2,688,553       $      3,419,536   $     4,291,859   $       5,791,288     453       594       711     835     986   $      360   $      377   $      401   $      427   $      489
 Larimer            $      204,104      $      282,043       $       351,119    $      494,065    $        611,755       61        76        89     114     136   $      278   $      310   $      327   $      361   $      375
 Weld               $       53,647      $       69,539       $       117,308    $      160,587    $        226,278       18        21        29      39      53   $      254   $      270   $      330   $      340   $      356
 All Other Top 10
                    $     3,109,110     $    4,121,449       $      5,203,464   $     6,461,242   $       8,564,477     191       238       283     334     397   $      303   $      325   $      354   $      373   $      402
 Counties

 Total/Average      $     6,452,883     $    8,611,460       $     10,927,638   $    14,661,707   $      19,838,155     162       202       241     294     355   $      315   $      338   $      363   $      389   $      420




 William M. Mercer, Incorporated                                                                  265                                              Colorado Child Welfare Evaluation
                                                                                                                                          Second Interim Implementation Status Report
The third series of analyses is presented in Table 42. There is one caution that should be
observed with the table. SFY 1999 Medicaid expenditures show an unexpected decrease when
compared to SFY 1998. According to the CDHS, there was a problem in Medicaid close out in
SFY 1999 and $6 million was not assigned to the counties. This problem is being corrected. In
the meantime, while it may be appropriate to compare counties within a year, it is probably not
appropriate to make comparisons across years.

The table consists of three sections. The first shows Medicaid allocation and expenditures, the
second shows room and board expenditures and the total of Medicaid and room and board
expenditures. The final section makes comparisons of expenditures and days used on the basis of
children in poverty in the county’s population.

The table shows that the group of managed care and non-managed care counties did not spend
the entire Medicaid allocation in SFY 1998 and SFY 1999. They were over by 8 percent in
SFY 2000.

This finding has both a positive and negative side. The Medicaid allocations in the table consist
of approximately 50 percent state-county funds and 50 percent federal funds. In SFY 1998, the
10 counties did not use approximately $2.7 million of their allocations, meaning that about $1.35
million (the state-county share) could have been used for other purposes such as less costly
alternatives to RTC. This is the positive side. The negative side is that the other $1.35 million
(the federal share) would probably be lost to the service system unless used to match
state-county funds for some other Medicaid-reimbursable service.

In Table 42, a comparison of each county’s allocation to its child population in poverty shows
that allocations are not distributed on the basis of this variable. As a result, it is not possible to
make comparisons among counties, or sets of counties, in how efficiently funds were expended;
there is no apparent common denominator. We adjusted both total expenditures and total days of
RTC by dividing both by the child population in poverty of each county, a common
denominator. This provides an index of how much was spent and how many days were used per
child in poverty in each county’s population. The results are displayed in the lower section of
Table 42.

We selected children in poverty, rather than the total population of children in each county, as
there is a relationship between poverty and child abuse and neglect. There is also a relationship
between serious emotional disturbances and poverty. These children should have a higher
probability of using more intensive services, such as RTC. Of interest is that the size of the child
population in poverty in each set of counties is almost identical.

The table shows that managed care counties expend less funds on RTC per child in poverty than
non-managed care counties for each year in the table. They also use less days each year per child
in poverty. Finally, managed care counties show more variability among counties than the non-
managed care counties. It appears that managed care counties are performing better on this index
than non-managed care counties.



William M. Mercer, Incorporated                  266               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
          Table 13: Residential Treatment Center Service (RTC) Medicaid Allocation,
                          Actual Expenditures, SFY1998 to SFY2000
                                                                                                                                                                Percent Difference of Medicaid
                                        RTC Medicaid Allocation                                                 RTC Medicaid Expenditures
      County                                                                                                                                                          Allocation Spent
                              SFY98                   SFY99                   SFY00                  SFY98               SFY99                   SFY00           SFY98       SFY99      SFY00

  Arapahoe           $            4,029,313    $          4,101,749    $           4,259,698   $       5,344,558     $      4,758,173   $           5,657,697     33%        16%         33%
  Boulder            $            1,321,744    $          1,362,346    $           1,451,557   $           986,951   $       597,762    $           1,592,854    -25%        -56%        10%
  El Paso            $            5,929,456    $          6,055,350    $           6,332,717   $       4,101,420     $      2,239,503   $           3,206,825    -31%        -63%       -49%
  Jefferson          $            2,861,510    $          2,938,011    $           3,104,634   $       3,494,757     $      3,399,437   $           5,832,875     22%        16%         88%
  Mesa               $            1,584,951    $          1,621,102    $           1,700,512   $           996,942   $       936,425    $           1,001,899    -37%        -42%       -41%
  Pueblo             $            2,968,265    $          3,032,226    $           3,172,473   $       1,570,047     $      1,622,870   $           2,028,040    -47%        -46%       -36%
 All Managed Care
                     $           18,695,239    $      19,110,784       $          20,021,591   $      16,494,675     $     13,554,170   $         19,320,189     -12%        -29%        -4%
 Counties
  Adams              $            3,318,099    $          3,404,736    $           3,595,599   $       3,766,956     $      3,090,253   $           4,704,479     14%         -9%        31%
  Denver             $           10,597,147    $       10,795,612      $          11,232,417   $       9,154,963     $      8,412,584   $         12,415,774     -14%        -22%        11%
  Larimer            $            1,934,814    $          1,976,476    $           2,068,097   $       2,134,262     $      1,433,304   $           2,091,724     10%        -27%        1%
  W eld              $            1,853,336    $          1,903,385    $           2,013,204   $       2,151,294     $      1,733,903   $           3,480,153     16%         -9%        73%
 All Non-Managed
                     $           17,703,396    $      18,080,209       $          18,909,317   $      17,207,475     $     14,670,044   $         22,692,131      -3%        -19%        20%
 Care Counties

 Total/Average       $           36,398,635    $      37,190,993       $          38,930,908   $      33,702,150     $     28,224,214   $         42,012,320      -7%        -24%        8%


                                              RTC Room & Board Expenditure                                                  Total Expenditures (RTC Medicaid & R&B)
          County
                                      SFY98                           SFY99                        SFY00                 SFY98                       SFY99                     SFY00

  Arapahoe                $               1,232,868         $              1,290,748     $           1,253,116       $       6,577,426       $            6,048,921      $           6,910,813
  Boulder                 $                   204,358       $               170,818      $            357,319        $       1,191,309       $              768,580      $           1,950,173
  El Paso                 $                   976,120       $               571,957      $            655,039        $       5,077,540       $            2,811,460      $           3,861,863
  Jefferson               $                   707,156       $               739,994      $           1,231,189       $       4,201,913       $            4,139,431      $           7,064,064
  Mesa                    $                   190,185       $               265,028      $            224,718        $       1,187,127       $            1,201,453      $           1,226,617
  Pueblo                  $                   378,648       $               428,575      $            438,352        $       1,948,695       $            2,051,445      $           2,466,393
  All Managed Care
                          $               3,689,335         $              3,467,119     $           4,159,734       $     20,184,010        $           17,021,289      $          23,479,923
  Counties
  Adams                   $                   826,771       $               733,913      $           1,037,983       $       4,593,727       $            3,824,166      $           5,742,463
  Denver                  $               2,197,826         $              2,270,301     $           2,551,473       $     11,352,789        $           10,682,885      $          14,967,247
  Larimer                 $                   383,146       $               388,763      $            436,152        $       2,517,408       $            1,822,067      $           2,527,876
  W eld                   $                   462,994       $               428,202      $            704,972        $       2,614,288       $            2,162,105      $           4,185,125
  All Non-Managed
                          $               3,870,738         $              3,821,179     $           4,730,580       $     21,078,213        $           18,491,223      $          27,422,711
  Care Counties

  Total/Average           $               7,560,073         $              7,288,298     $           8,890,314       $     41,262,223        $           35,512,512      $          50,902,634

                                                                                                                                   Total Days Used Per Child in Poverty in County
                         Estim ated # of Children Total Expenditure Per Child in Poverty in County Population                                       Population
          County
                         (<18) in Poverty (1997)
                                                         SFY98              SFY99               SFY00                                       SFY98                  SFY99               SFY00

  Arapahoe                                12,318      $                     534    $                 491    $              561                       3.94                    4.43         3.78
  Boulder                                     6,234   $                     191    $                 123    $              313                       1.41                    1.14         2.02
  El Paso                                 18,880      $                     269    $                 149    $              205                       2.13                    1.39         1.34
  Jefferson                                   9,484   $                     443    $                 436    $              745                       3.23                    3.97         4.83
  Mesa                                        5,571   $                     213    $                 216    $              220                       1.67                    1.61         1.50
  Pueblo                                      9,455   $                     206    $                 217    $              261                       1.57                    1.85         1.64
 All M anaged Care
                                          61,942      $                     326    $                 275    $              379                       2.46                    2.45         2.49
 Counties
  Adam s                                  15,180      $                     303    $                 252    $              378                       2.40                    2.18         2.55
  Denver                                  31,572      $                     360    $                 338    $              474                       2.77                    2.94         3.34
  Larim er                                    6,188   $                     407    $                 294    $              409                       3.13                    2.78         2.67
  W eld                                       7,899   $                     331    $                 274    $              530                       2.47                    2.29         3.57
 All Non M anaged
                                          60,839      $                     346    $                 304    $              451                       2.67                    2.65         3.10
 Care Counties

 Total/Average                           122,781      $                     336    $                 289    $              415                       2.56                    2.55         2.79

Note: Child population in poverty is based on estim ates in 1997 by the US Census Bureau, Sm all Area Incom e and Poverty Estim ates Program


William M. Mercer, Incorporated                                                              267                      Colorado Child Welfare Evaluation
                                                                                                             Second Interim Implementation Status Report
Unlike residential placements, adoption is viewed as a permanent, positive placement. We
conducted a series of cost analyses to examine the progress of the managed care counties in this
area.

Figure 131 shows the expenditures for three years for subsidized adoption in the managed care
counties. Figure 132 provides the same information for non-managed care counties. Both figures
show that all counties have increased their expenditures each year in this area. Of particular
interest is the large increase in expenditures in El Paso County. Associated with this is El Paso’s
lead among managed care counties in adoptions per 1,000 children (Table 7).
                                       Figure 131
           Subsidized Adoption Expenditures by State Fiscal Year, Managed Care
                                        Counties




                                                                                                                                           $5,273,488
                                                                                                                           $3,198,577
         $6,000,000
                                                    $1,930,761




         $5,000,000
                                       $1,598,587




                                                                                                          $1,488,520
                        $1,319,390




                                                                                                                                                                                                                                                                     $1,346,503
                                                                                                                                                                                   $1,294,771




                                                                                                                                                                                                                                                     $1,140,244
                                                                                                                                                                      $1,104,039
         $4,000,000


                                                                                                                                                         $959,054




                                                                                                                                                                                                                                      $982,133
                                                                                         $791,793
                                                                              $678,368




                                                                                                                                                                                                                       $636,361
                                                                 $592,846




                                                                                                                                                                                                            $480,650
                                                                                                                                                                                                $382,232
         $3,000,000
         $2,000,000
         $1,000,000
               $-
                        ARAPAHOE                                   BOULDER                                     EL PASO                                   JEFFERSON                                          MESA                          PUEBLO
                                     SFY1998                                             SFY1999                                                        SFY2000



                                          Figure 132
              Subsidized Adoption Expenditures by State Fiscal Year, Non-Managed
                                       Care Counties
                                                                                                                                        $5,791,288
                                                                                                              $4,291,859
                                                                                            $3,419,536




          $6,000,000
          $5,000,000
                                                                 $1,935,156
                                                    $1,514,732
                               $1,315,501




          $4,000,000
                                                                                                                                                                                                 $611,755




          $3,000,000
                                                                                                                                                                                   $494,065
                                                                                                                                                                $351,119




                                                                                                                                                                                                                                                                  $226,278
                                                                                                                                                                                                                           $117,308

                                                                                                                                                                                                                                          $160,587




          $2,000,000
          $1,000,000
                $-
                                              ADAMS                                                      DENVER                                                            LARIMER                                                    WELD


                       SFY1998                                     SFY1999                                                 SFY2000




William M. Mercer, Incorporated                                                                                        268                                       Colorado Child Welfare Evaluation
                                                                                                                                                        Second Interim Implementation Status Report
Figures 133 and 134 display the average monthly client FPE for subsidized adoption. El Paso
and Denver are the leaders in this area.

                                         Figure 133
              Average Monthly FPE Client for Subsidized Adoption by State Fiscal
                               Year, Managed Care Counties

      800
      700




                                                                             713
      600




                                                                       498
      500
                           376
                    334




      400
                                                                311




                                                                                                 292
              282




                                                                                                                                             273
                                                                                          256




                                                                                                                                       250
                                                                                    228




                                                                                                                                 222
      300
                                                    175
                                              155
                                       139




                                                                                                                      135
      200




                                                                                                                108
                                                                                                         91
      100
       0
              ARAPAHOE                  BOULDER                  EL PASO            JEFFERSON                 MESA                 PUEBLO


                                   SFY1998                        SFY1999                   SFY2000




                                             Figure 134
                  Average Monthly FPE Client for Subsidized Adoption by State Fiscal
                                 Year, Non-Managed Care Counties


       1,000
                                                                      835




         900
                                                                              986
                                                          711




            800
            700
            600
                                             413




            500
                                 348
                     304




            400
            300
                                                                                                          136
                                                                                                   114




            200
                                                                                          89




                                                                                                                                             53
                                                                                                                                   39
                                                                                                                            29




            100
              0
                           ADAMS                                DENVER                          LARIMER                          WELD


                                       SFY1998                    SFY1999                 SFY2000




William M. Mercer, Incorporated                                                     269                    Colorado Child Welfare Evaluation
                                                                                                  Second Interim Implementation Status Report
Figures 135 and 136 show that, while El Paso and Denver may be the leaders in subsidized
adoption client FPEs, they are paying more per FPE than the other counties. El Paso, in
particular, paid $126 more per client FPE in SFY 2000 than the second most expensive county,
Denver.

                                         Figure 135
            Average Monthly Cost Per FPE Client for Subsidized Adoption by State
                            Fiscal Year, Managed Care Counties

       $800




                                                                                  $615
       $700




                                                                           $530
       $600
                               $428




                                                                                                                                                                $411
                        $399




                                                                  $395
                $390




                                                        $377




                                                                                                                                  $392
       $500




                                                                                                                                                         $380
                                                                                                         $369




                                                                                                                           $369




                                                                                                                                                $369
                                                 $365




                                                                                                 $359
                                        $354




                                                                                          $351




                                                                                                                   $351
       $400
       $300
       $200
       $100
       $-
                ARAPAHOE                 BOULDER                    EL PASO              JEFFERSON                        MESA                    PUEBLO


                                      SFY1998                      SFY1999                        SFY2000




                                          Figure 136
              Average Monthly Cost Per FPE Client for Subsidized Adoption by State
                           Fiscal Year, Non-Managed Care Counties


        $600
                                                                                  $489
                                                                         $427




        $500
                                                           $401
                                          $390




                                                                                                                      $375
                                 $362
                       $360




                                                                                                            $361




                                                                                                                                                                 $356
                                                                                                                                                       $340
                                                                                                                                         $330
                                                                                                 $327




        $400

        $300

        $200

        $100

        $-
                               ADAMS                              DENVER                                LARIMER                                   WELD


                                        SFY1998                          SFY1999                    SFY2000




William M. Mercer, Incorporated                                                          270                       Colorado Child Welfare Evaluation
                                                                                                          Second Interim Implementation Status Report
In summary, results are mixed. On the one hand, it appears that managed care counties have not
contained costs, as expenditures have exceeded allocations for the managed care counties as a
group. On the other hand, the degree of increases in cost per FPE for different types of
residential care appears to have been reduced from the period prior to the advent of managed
care. Managed care counties also appear to perform better than non-managed care counties on an
index based on expenditures and days of RTC compared to the population of children in poverty.




William M. Mercer, Incorporated              271              Colorado Child Welfare Evaluation
                                                     Second Interim Implementation Status Report
Title IV-E Waiver Demonstration
Colorado’s Title IV-E waiver demonstration project is aimed at reducing involvement in
out-of-home placement and improving other outcomes for children 10 years of age and older,
and their families, who are at risk for or are already experiencing multiple placement moves. The
project will utilize risk-based, performance-based, case rate contracts between counties and
service providers (or consortia of providers) to help achieve this overall objective. Arapahoe
County is the first county to take part in the demonstration.

The primary goal of Arapahoe County’s demonstration is to use new financing and contracting
mechanisms to encourage greater coordination of services among providers. The expected result
is that an experimental group of children, who are financed through the case rate, will experience
fewer unnecessary transitions and, in particular, faster placements into out-of-home care when it
is needed, less time in out-of-home care, and lower rates of re-entry into out-of-home care than a
control group of children whose care is funded through fee-for-service.

Arapahoe County intends to contract with a consortium of providers, Colorado Care
Management (CCM), to provide a complete array of residential, in-home, and follow-up
services. Arapahoe County is presently developing a case rate financing structure to use with
CCM.

The target population to be served in Arapahoe County’s demonstration includes children who
are assessed as being at high risk for, or who already are experiencing, multiple placement
moves and/or are at significant risk of aging out of the system without a permanent family
relationship or another appropriate permanent living arrangement. These are children who are in
high-cost residential care settings, have had multiple placements, are experiencing difficulty in
school (functioning below grade level, truant, poor peer relationships), and/or are experiencing
difficulty in reintegrating into family settings.

There are two phases to the evaluation of the Title IV-E demonstration in Arapahoe County. The
first phase, which started in SFY 2001, involves the collection of information on a pilot group of
children presently in residential care and their families. This is being done for two purposes.
First, it will study the difference in outcomes between the network fee that Arapahoe is presently
using with CCM and the case rate that is being developed. Second, it provides a test of the
procedures and instruments that are being used in the evaluation of the Title IV-E demonstration.

When the case rate is introduced in SFY 2002, the second phase of the design will begin. It will
involve the random assignment of children and their families to demonstration and control
conditions. Baseline and follow-up data will be collected on children and their families in both
the demonstration and control conditions. Individual-level data will include process (e.g.,
services received, costs) and outcomes (safety, permanency, well-being) data. In addition,
system-level and aggregate data will be collected through measurements of system and practice
changes.




William M. Mercer, Incorporated                272              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
The following sections describe the results of the first phase of the evaluation of the Title IV-E
demonstration, as well as some preliminary cost findings. The first section describes the practice
evaluation findings, which consist of results from the administration of the CWPA. The second
describes some findings related to one of the outcome measures, the North Carolina Family
Assessment Scale (NCFAS). The final section contains cost findings. Process findings are
described in previous sections of this report and will be supplemented in SFY 2002 with more
detail.




William M. Mercer, Incorporated                273               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Practice Findings
This component studies the impact of the initiative on the services that are provided to children
and their families. For example, it explores family members’ perceptions of the accessibility,
family-friendliness, and flexibility of services. The primary instrument used in this component of
the evaluation is the CWPA.

In SFY 2001, the CWPA was used to study the perceptions of parents and caregivers in the pilot
group in Arapahoe County. The parent/caregiver version of the CWPA, which was used last
year, was revised based on focus group feedback and a psychometric analysis of the instrument.
The revision was aimed at tailoring the CWPA-Parent/Caregiver Version to the goals of
demonstration program, which, in Arapahoe County, focuses heavily on children in RTC
placements.

The CWPA-Parent/Caregiver IV-E Version addresses the following domains of child welfare
practice related to the Title IV-E demonstration:

§     Family-Centered Practice. Family-centered practice in child welfare is rooted in the premise
    that a child’s family network has the ability to accurately assess the child’s needs and the
    responsibility to meet the identified needs. Family-centered practice is exemplified by a strong
    family-caseworker partnership, a leading role for families in the assessment process, and full
    family participation in service planning.
§     Strengths-Based Practice. Leading child welfare theorists and activists advocate a strengths-
    focused and empowering approach to working with families. A strengths-based approach
    builds on a family’s current successes and helps parents to develop additional skills and
    resources for the future. If strengths-based practices are successfully used with a family, the
    family network becomes better able, over time, to meet the needs of its children and keep them
    safe, even after they leave the child welfare system.
§    Access and Quality of Services. The ability of families to access needed services, and the
    quality of those services once accessed, are key service-related goals measured by the CWPA,
    Parent/Caregiver IV-E Version.
§     Collaboration with Community Support Systems and Other Agencies. Community supports
    include such entities as churches, health clinics, recreational organizations, extended families,
    neighbors, and other groups. Promoting their use in child welfare practice is seen as a way to
    expand traditional county services into a more comprehensive array of services and
    coordinated system of social supports. Collaboration with other human service systems is also
    an important feature of child welfare practice in a managed care environment. The prevalence
    of families within the child welfare system involved in multiple systems is widely recognized.
    A better integrated overall system of care is hoped to eliminate redundant services and costs,
    offer a better experience to families and, ultimately, lead to better outcomes.




William M. Mercer, Incorporated                   274              Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
The CWPA was administered by mail to all 32 parents and caregivers of children in RTCs
through Arapahoe County in February 2001. Returns were received from parents and caregivers
of eight children, out of a total of 26 that could have been delivered, for a response rate of 32
percent. While the response rate is low, it is comparable to last year’s overall rate and does allow
for some observations.

Family-Centered Practice Findings
The results of the survey indicate that caseworkers and service providers for these 8 children
have adopted a family-centered approach in their work. All 8 families reported that they were
satisfied or very satisfied with the overall process. The families reported knowing about the
family service plan (7 of 8), helping to develop the plan (8 of 8), and agreeing somewhat (2 of 8)
or strongly (6 of 8) with the terms of the plan. Most families strongly agreed with the idea of
having their child placed in an RTC (6 of 8) and strongly agreed with the choice of the facility (5
of 8). Even though some families disagreed with the choice of the particular RTC where their
child was placed (2 of 8), all of the respondents indicated that they received an explanation for
the selection.

Similar results were found for the interaction surrounding discharge planning. Survey
respondents indicated that they knew about the discharge plan (7 of 8), were involved in
developing the plan (5 of 7), and agreed somewhat (3 of 7) or strongly (2 of 7) with the discharge
plan.

The service planning and discharge planning processes for these eight families were
characterized by mostly strong family-professional partnerships in the various agencies involved
in providing services. First, parents and caregivers reported that residential providers and
families were meeting regularly while their children were in placement. Half of the families were
meeting with residential staff every other week or more frequently. Only one family reported that
they were meeting less than once a month. Regardless of the frequency of scheduled meetings,
every caregiver indicated that they were satisfied with the amount of time the RTC staff was
spending with them.

Second, service professionals appear to be respecting these parents and caregivers and listening
to their views. Survey respondents somewhat or strongly agreed that their caseworkers (6 of 8)
and probation officers (6 of 8) treated them with respect. Caregivers agreed strongly (2 of 8) or
somewhat (4 of 8) that their caseworkers listened to their views on what their children and
families needed. Three of seven strongly agreed and the remaining four of seven somewhat
agreed that the staff at the RTCs listened to their views.

Responses were more variable regarding communication between families and probation
officers. Two of five strongly agreed, one of five somewhat agreed, one of five somewhat
disagreed, and one of five strongly disagreed that the child’s probation officer listened to their
views. (The item related to communication with probation officers was not applicable for some
respondents.)

In a similar vein, six of eight respondents somewhat or strongly agreed that the RTC staff
explained what the parents or caregivers were expected to do. A strong majority of caregivers (6

William M. Mercer, Incorporated                 275               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
of 7) somewhat or strongly agreed that their ideas were taken into account when decisions were
actually made.

Another marker of family-centered practice is flexibility on the part of professionals to
accommodate the needs and preferences of a particular family. In general, caregivers felt that
both the RTC staff and the probation officers were flexible (8 of 8 and 6 of 8 agreed,
respectively). The results about department caseworkers were more variable. Three of eight
strongly agreed, two of eight somewhat agreed, two of eight somewhat disagreed, and one of
eight strongly disagreed that their caseworker was flexible.

Strengths-based Practice Findings
The results from the survey are less clear about the use of a strengths-based approach with the
families in the study than they were about family-centered practices. On the one hand, four of
five families responding stated that they were satisfied or very satisfied that the services they
received built on their families’ strengths. On the other hand, most of the study participants
indicated that they were not told what they were doing well in caring for their children (5 of 6),
an integral practice in a strengths-based approach to working with families. An additional three
of five disagreed that the RTC staff had pointed out their strengths. Three of four caregivers
strongly disagreed that their probation officer told them what they were doing well for their
children.

It may be that more can be done to implement a strengths-based focus in Arapahoe County.
However, it may be equally likely that this initial finding is due to the small number of people
who answered these questions. With more respondents, a clearer picture will emerge of the
prevalence of strengths-based practices.

Access and Quality of Services Findings
The eight caregivers provided a range of responses to questions concerning access to services
through the child welfare department. Generally, about half were pleased with the degree of
access they experienced and half were displeased. For example, three of seven caregivers said
they were satisfied or very satisfied with the ease of getting services and four of seven were
dissatisfied or very dissatisfied. When asked about the ease of traveling to the services offered,
three of seven caregivers expressed satisfaction and four of seven expressed dissatisfaction.

Most caregivers felt positively about the quality of the services. Six of seven were satisfied or
very satisfied with the match between the needs of their families and the actual services they
received. All of the participants who responded to the question (6 of 6) felt that the services
respected their families’ race and culture. And, finally, the majority of caregivers believed that
the services they received were helpful overall. Three of eight respondents somewhat agreed and
four of eight strongly agreed that the services were helpful, and one of eight strongly disagreed.




William M. Mercer, Incorporated                 276               Colorado Child Welfare Evaluation
                                                         Second Interim Implementation Status Report
The quality of the services may be high in the eyes of the caregivers because most of the service
professionals are reliably meeting their responsibilities. Most caregivers somewhat (1 of 8) or
strongly (4 of 8) agreed that they could count on their department caseworker. One of eight
somewhat disagreed and two of eight strongly disagreed. All of the caregivers somewhat (4 of 8)
or strongly (4 of 8) agreed that the RTC staff was reliable. Fewer caregivers thought highly of
their children’s probation officers. One of five strongly agreed, two of five somewhat agreed,
and two of five strongly disagreed that they could count on their probation officers.

Collaboration with Community Support Systems and Other Agencies Findings
The CWPA results suggest that caseworkers and RTC staff are not consistently taking advantage
of the community supports to which families may have access. Three of six caregivers said their
caseworkers asked about community supports and three of four said they did not ask. Only
somewhat more than half of caregivers (3 of 5) said the RTC staff asked about community
supports.

On the other hand, respondents indicated that caseworkers, staff and probation officers were
collaborating well with one another. A majority of caregivers agreed that their caseworkers (5 of
7), the RTC staff (7 of 7), and their children’s probation officer (3 of 5) worked well with other
agencies. Yet, this collaboration has not eliminated one of the perennial complaints of parents
involved with the child serving-system: the necessity of having to repeat their family history to
many different people. Seven of eight families indicated that repeating their story multiple times
is still a problem for them.

Conclusion
In the areas of family-centered practice, the analysis of this small pilot group suggests that
Arapahoe County caseworkers and service providers have adopted a family-centered approach in
their work with parents and caregivers who have a child in residential services. The results from
the survey seem less clear about the use of a strengths-based approach with the families in the
study than they were about family-centered practices. On the one hand, four of five families
stated that they were satisfied or very satisfied that the services they received built on their
families’ strengths. On the other hand, most of the study participants indicated that they were not
told what they were doing well in caring for their children. Caregivers in the study provided a
range of responses to questions concerning access to services through the child welfare
department. Generally, about half of the caregivers were pleased with the degree of access they
experienced, and half were displeased.




William M. Mercer, Incorporated                277               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
Item-by-item, results form the survey are presented in Table 43.

                                             Table 43
                            Title IV-E Pilot Group CWPA Responses

Survey Item                                  Strongly     Somewhat Somewhat      Strongly   Number of
                                              Agree         Agree   Disagree     Disagree     Valid
                                                                                            Responses

1) I was told what I do well in caring for
   my children by:
a) My caseworker                                1                        2           3           6
b) The staff at the treatment center
   where my child was placed                    2                        2           1           5
c) My child’s probation officer                 1                                    3           4
2) My views on what my child and
   family needed were listened to by:

a) My caseworker                                2            4                       2           8
b) The staff at the treatment center
   where my child was placed                    3            4                                   7
c) My child’s probation officer                 2            1           1           1           5
3) I was told why my child was placed in
   this particular facility.                    7            1                                   8
4) I agreed with the plan to:
a) Have my child receive residential
   services.                                    6            1           1                       8
b) Have my child served by this
   particular facility.                         5            1           1           1           8
5) The staff at the treatment center
   where my child was placed explained
   to me what I was expected to do.             3            3           1           1           8


6) My ideas were taken into account
                                                3            3                       1           7
   when making decisions.
7) I wish I had not had to tell my story
   so many times to so many different           3            2           1                       6
   people.
8) I was treated with respect by:

a) My caseworker                                4            2           1           1           8



William M. Mercer, Incorporated                     278               Colorado Child Welfare Evaluation
                                                             Second Interim Implementation Status Report
                                            Table 43
                           Title IV-E Pilot Group CWPA Responses

Survey Item                               Strongly     Somewhat Somewhat      Strongly   Number of
                                           Agree         Agree   Disagree     Disagree     Valid
                                                                                         Responses

b) The staff at the treatment center
   where my child was placed                 6            1                                   7

c) My child’s probation officer
                                             2            1           1                       4
9) When working with my family to get
   services:
a) My caseworker was flexible when
   needed                                    3            2           2           1           8
b) The staff at the treatment center
   where my child was placed were
   flexible when needed                      6            1                                   7


c) My child’s probation officer was
   flexible when needed                      2            1           1                       4

10) My caseworker worked well with
    other agencies (for example, the
    court, schools, probation, mental
    health, health, and drug/alcohol         4            1           1           1           7
    providers).
11) The staff at the treatment center
    where my child was placed worked
    well with other agencies (for
    example, the court, schools,
    probation, mental health, health,        4            3                                   7
    and drug/alcohol providers).


12) My child’s probation officer worked
    well with other agencies (for
    example, the court, schools,
    probation, mental health, health,        2            1                       2           5
    and drug/alcohol providers).


13) I was asked about community
     supports my family has found
     helpful (for example, churches,
     family, friends):
a) By my caseworker                          2            1           2           1           6

William M. Mercer, Incorporated                  279               Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
                                            Table 43
                           Title IV-E Pilot Group CWPA Responses

Survey Item                                Strongly     Somewhat Somewhat      Strongly   Number of
                                            Agree         Agree   Disagree     Disagree     Valid
                                                                                          Responses

b) By the staff at the treatment center
   where my child was placed                  3                        1           1           5

c)   By my child’s probation officer
                                              2                        1                       3
14) I knew about the family service plan
    that was put together to help my
    child and family.                         5            2           1                       8

15) I was involved in developing the
    family service plan that was put
    together to help my child and             5            3                                   8
    family.
16) I agreed with the family service
    plan.                                     6            2                                   8
17) I knew about the discharge plan that
    was put together to help my child
    and family.                               3            3           1                       7

18) I was involved in developing the
    discharge plan that was put
    together to help my child and             2            3                       2           7
    family.
19) I agreed with the discharge plan.
                                              2            3           1           1           7
20) I could count on my caseworker.
                                              4            1           1           2           8
21) I could count on the staff at the
    treatment center where my child           4            4                                   8
    was placed.
22) I could count on my child’s
    probation officer.                        1            2                       2           5
23) Overall, the services I got were
    helpful.                                  4            3                       1           8
24) With the time spent with you by:


a)   Your caseworker                          1            5                       2           8




William M. Mercer, Incorporated                   280               Colorado Child Welfare Evaluation
                                                           Second Interim Implementation Status Report
                                            Table 43
                           Title IV-E Pilot Group CWPA Responses

Survey Item                                 Strongly     Somewhat Somewhat      Strongly   Number of
                                             Agree         Agree   Disagree     Disagree     Valid
                                                                                           Responses

b)   The staff at the treatment center
     where your child was placed               2            5                                   7

25) With how easy it was to get services
    you needed?                                2            1           3           1           7
26) With how well services respected
    your family’s race and culture?            4            2                                   6

27) With how easy it was to travel to the
    services offered?                          1            3           3           1           8
28) With your level of involvement in
    the process?                               4            3                                   7
29) Did the services your family
    received match your family’s               3            3                       1           7
    needs?
30) Did the services your family
    received build on your family’s            1            2           2                       5
    strengths?




William M. Mercer, Incorporated                    281               Colorado Child Welfare Evaluation
                                                            Second Interim Implementation Status Report
Outcome Measure Findings
In this component, the pilot group of thirty-two children was used. These children entered an
RTC between February 1998 and December 2000. Most were new to an RTC when the study
began. Seventy-five percent (n=23) had been in an RTC for less than a month when the first
outcomes tracking form was completed. On the other hand, two of the children had been in an
RTC for almost two years when they entered the study.

For many of the pilot group of children (40 percent, n=13), their entry into RTC represented the
first change in their living arrangement in at least six months. The remaining children, however,
had moved anywhere from one to eight times in the previous six months. The average number of
moves for all of the children combined was 1.4.

The permanency goal for most of the children when they entered the study was to return home
(n=23, 72 percent). Five children were en route to independent living (16 percent), two were
slated for adoption (6 percent), and two held a long-term foster care placement agreement
(6 percent).

Information about the pilot group of children comes from two measures that are new to the
evaluation, the Outcomes Report Form (ORF) and the NCFAS, Version 2.0. The ORF is a one-
page sheet that may be completed by the caseworker assigned to a child in an out-of-home
placement or by a quality assurance manager responsible for tracking outcomes for children in
placement. The ORF contains case tracking data, such as the date of a child’s entry into
placement the type and restrictiveness of the placement, the permanency goal for the child, the
date the goal was achieved, and the date the case was closed.

The NCFAS is a standardized instrument that is used in clinical, case management, and research
settings in human services. It contains 36 items in 5 domains.

§   Family Environment,
§   Parental Capabilities,
§   Family Interactions,
§   Family Safety, and
§   Child Well-Being.

Each of the five domains on the NCFAS has an “overall” item as part of the subscale. There is an
overall Family Environment item, an overall Parental Capabilities item, and so on. The overall
item appears at the top of the subscale sheet on the instrument and is followed by the individual
items pertaining to that domain.




William M. Mercer, Incorporated                282              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Each item on the NCFAS is rated on a 6-point scale that rates different levels of strength or
problem in the area of functioning or well-being addressed by the item. Scores on individual
items range from 2 to –3 according to the following scale:

  +2     Clear strength
  +1     Mild strength
    0    Baseline/Adequate
   -1    Mild problem
   -2    Moderate problem
   -3    Serious problem

The NCFAS was completed for 30 of the 32 children in the pilot group on February 21, 2001.
One survey could not be used, reducing the number of children to 29. Because one intake
NCFAS was identified late in the analysis stage, some of the intake analyses that are reported
here, however, were based on 28 children.

Two of the 30 children were discharged from their RTC placements before the end of the pilot
study. Thus, both intake and closure NCFAS surveys were completed for them. These two
surveys were reviewed to see if any sensitivity to change during an RTC placement could be
observed. For the remaining 2 children out of the total of 32, a NCFAS was completed only
when they left the RTC, not at study entry.

The following sections profile the pilot group and provide information on changes in two
children who were discharged from RTC. The five domains of the NCFAS are used to provide
the profiles.

In order to describe the children and families in the pilot group, the mean of each item on the
NCFAS is reported. To calculate a mean of an item for a group of individuals and then to
compare that item mean to another item mean requires the assumption that the scale underlying
the items be interval in nature. As the NCFAS has not been determined to be an interval scale,
interpretation should be limited to statements about relative levels of average rating between
items/subscales. For example, an average score of –0.7 on one item could be compared to an
average score of –1.2 on another item and the observation made that the first score is less of a
problem on average for the families than the second. One could even label both scores as in the
range of a mild problem. However, one would not be able to state that the degree of difference
between the two was exactly 0.5 or a half-step on the scale.




William M. Mercer, Incorporated                283              Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Family Environment Findings
The environment of the family is a relative strength for the children in the pilot group (Table 44).
The overall rating for the family environment was -.5, exactly between the anchors of “adequate”
and “mild problem.” The nine individual items ranged from .04 for Personal Hygiene and Food
and Nutrition, to -.93 for Community Safety. Thus, according to the NCFAS ratings, the most
difficult challenge facing families in this domain appears to be disturbances in the neighborhood
that may, occasionally, prevent children from spending time outside in the community.

                                              Table 44
                                  Mean Scores for Environment Items
                                               N                Mean           Std. Deviation
        Overall Environment                           24               -.50                1.414
        Housing stability                             28               -.21                1.792
        Community Safety                              28               -.93                1.412
        Habitability of housing                       28               -.25                1.506
        Income/Employment                             28               -.54                1.527
        Financial Management                          28               -.71                1.512
        Food & Nutrition                              28                .04                1.290
        Personal Hygiene                              28                .04                1.401
        Transportation                                27               -.44                1.739
        Learning Environment                          27               -.56                1.502
        Item scores range from 2 to -3.



Parental Capabilities Findings
The parents of the children in the pilot group experience some mild problems overall in their
capacity to properly care for their children, according to the NCFAS ratings (mean = -1.09,
Table 45). Parents’ physical health is adequate on average but caseworkers rated parents as
having problems in disciplinary practices, child supervision, enrichment opportunities, and in the
area of mental health. No areas of strength, other than physical health (a very mild strength
overall), were identified in this domain.




William M. Mercer, Incorporated                 284                 Colorado Child Welfare Evaluation
                                                           Second Interim Implementation Status Report
                                            Table 45
                          Mean Scores for Parental Capabilities Items
                                                   N            Mean         Std. Deviation
       Overall Parental Capabilities                   23          -1.09               1.593
       Child Supervision                               28          -1.21               1.475
       Disciplinary Practices                          28          -1.39               1.423
       Enrichment Opportunities                        28          -1.11               1.524
       Parents' Mental Health                          28            -.89              1.595
       Parents' Physical Health                        28             .07              1.359
       Parents' Drug Use                               28            -.36              1.660
       Item scores range from 2 to -3.




Family Interaction Findings
Mild problems are also evident among the pilot families in the domain of family interactions
(mean = -1.09, Table 46). Some problems were noted in the amount of mutual support offered by
family members, the degree of bonding between parents and children, the clarity and
appropriateness of the parents’ expectations of the children, and the relationship between adult
caregivers.

                                            Table 46
                            Mean Scores for Family Interaction Items
                                                  N            Mean          Std. Deviation
        Overall Family Interaction                     23         -1.26                1.287
        Bonding with Children                          28         -1.18                1.467
        Expectations of Children                       28         -1.18                1.090
        Mutual Support within Family                   28         -1.21                1.134
        Relationship between Parents                   24         -1.13                1.454
       Item scores range from 2 to -3.


Family Safety Findings
Family safety is a problem for families in the pilot group, as shown by a mean score of –1.17 on
the overall safety item in Table 47. Problems have been identified in the areas of physical abuse,
sexual abuse, neglect, and domestic violence. Emotional abuse is a particular area of concern for
these families (mean = 1.39).




William M. Mercer, Incorporated                285               Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                                             Table 47

                               Mean Scores for Family Safety Items
                                               N                Mean           Std. Deviation
        Overall Family Safety                         23           -1.17                   1.403
        Physical Abuse                                28             -.79                  1.618
        Sexual Abuse                                  27             -.85                  1.486
        Emotional Abuse                               28           -1.39                   1.227
        Neglect                                       28             -.86                  1.484
        Domestic Violence                             24             -.83                  1.579
        Item scores range from 2 to -3.




Child Well-Being Findings


The well-being of children in the pilot group is the most significant problem identified on the
NCFAS surveys. The mean scores for the overall well-being item and four of the seven
individual items in the domain all either reached the ranking of “moderate problem” or were
closer to the “moderate” than the “mild” problem level (see Table 48). The children, as a group,
were rated by caseworkers as experiencing emotional difficulties, handling stress poorly, being
uncooperative and oppositional, fighting often with peers, being influenced negatively by peers,
having poor relationships with their parents, and having problems with law enforcement. The
ratings indicated they have problems to a lesser degree with school performance and sibling
relationships. Caseworkers did not rate children in the pilot group as having a large amount of
motivation to change their behaviors and to remain with their families.

                                             Table 48
                                 Mean Scores for Child Well-being
                                                   N               Mean         Std. Deviation
       Overall Child Well-Being                            23         -1.65                1.152
       Child Mental Health                                 28         -1.89                1.031
       Child Behavior                                      28         -2.07                1.086
       School Performance                                  27         -1.33                1.617
       Parent-Child Relationship                           28         -1.64                1.393
       Sibling Relationships                               27         -1.11                1.086
       Peer Relationships                                  27         -1.59                 .971
       Motivation to Maintain Family                       28          -.82                1.517
      Item scores range from 2 to -3.




William M. Mercer, Incorporated                 286                  Colorado Child Welfare Evaluation
                                                            Second Interim Implementation Status Report
Note on the Averaging of NCFAS Scores
As a group, the pilot families do not appear to have severe problems outside the domain of child
well-being. Averaging scores across a group, however, reduces the visibility of some extreme
cases. Caseworkers gave the lowest rating possible (serious problem) on at least some items to
almost every family in the group. The seriousness of these problems is moderated when they are
combined with families who have lesser difficulty in these particular item areas. To preserve the
focus on the clinical needs of individual families, group averages should be interpreted carefully.
As discussed above, group averages should also be interpreted carefully because the scaling for
the NCFAS is generally considered to be ordinal-level, not interval-level.

NCFAS Change Scores
Two children in the pilot group were discharged from the RTC during the course of the pilot
study. Although a group consisting of two children is too small to present definitive results, some
information about the utility of the NCFAS to measure change in the RTC population can be
obtained by comparing the children’s NCFAS scores at intake and closure.

Change was measured in two ways. First, the overall rating for each of the five NCFAS
subscales was compared for both children. The scores are presented in Table 49. The overall
domain scores indicate that there has been some pre-post change in a positive direction for most
domains for both children.

                                       Table 49
        Changes in Overall Ratings between Intake and Closure for Two Children

           NCFAS Domain              Child 1                        Child 2
                                        Intake          Discharge      Intake       Discharge
    Overall Family Environment            +1             missing         -3             -2
    Overall Parental Capabilities         -2               +1            -3             -2
    Overall Family Interactions           -3                -1           -3             -3
    Overall Family Safety               missing          missing         -3             -2
    Overall Child Well-Being              -2                0            -3             -2

The second method for measuring change is to count the number of times a family’s rating on an
individual item changed between the child’s intake and discharge on those items that were rated
as a problem at intake. Counting the number of items that demonstrated improved ratings over
time may be done with ordinal-level data and does not require any assumptions about the
orthogonality of the NCFAS subscales.

For the first child, 16 of the 36 NCFAS items were rated as a mild, moderate, or serious problem.
Fifteen of the same items showed improved ratings at discharge.

Improvement across multiple items was also evident in the case of the second child. Twenty-two
of 36 items were problematic at intake and 20 of them received higher scores at discharge.

Based on an examination of only two children, the NCFAS instrument does show promise in
being sensitive enough to measure change concurrent with treatment at an RTC. Moreover, the


William M. Mercer, Incorporated                   287              Colorado Child Welfare Evaluation
                                                          Second Interim Implementation Status Report
NCFAS results suggest that change can occur in a positive direction and across a variety of
problem areas.




William M. Mercer, Incorporated               288               Colorado Child Welfare Evaluation
                                                       Second Interim Implementation Status Report
Cost Findings
As mentioned earlier, Arapahoe County is presently using a network fee with CCM to expedite
admissions of children to RTC. The network fee was included as part of a contract in
March 2000. It will be replaced with a case rate in FY 2002.

To study the impact of the network fee on RTC, we conducted a cost analysis. In the analysis,
the average monthly FPEs in RTC and the average monthly cost per FPE in RTC for all managed
care counties were determined for the time period beginning when they became managed care
counties through February 2000. This information is displayed in Table 50 in the columns
labeled “baseline.”

Baseline data were trended forward for three months (March–June 2000) to form a projection of
what might have been expected during this period, based on historical data. Also, actual data
from this period were collected. The columns labeled “projected” and “actual” display these data
for each managed care county.

Finally, baseline data are compared to actual data as a percentage to study the direction and
amount of difference. The same comparison is made for baseline data and projected data.

The results show that Arapahoe County decreased average monthly FPEs by 9.2 percent for the
period of March–June compared to the baseline period. This decrease was greater than any other
managed care county.

Arapahoe County, based on projections, would have been expected to have reduced average
monthly FPEs by 27.4 percent. This is a greater decrease than any other managed care county
and is larger than what actually occurred (9.2 percent).

In regard to average monthly cost per FPE, Arapahoe increased more than any other county,
except Jefferson. This is not as great an increase as would be expected on the basis of projection,
which was the largest of any of the managed care counties (22.5 percent).




William M. Mercer, Incorporated                 289              Colorado Child Welfare Evaluation
                                                        Second Interim Implementation Status Report
                     Table 50: Effect of Arapahoe CCM Integrated Network on Average Monthly FPE and
        Average Monthly Cost Per FPE for Residential Treatment Center (Room & Board) as Compared to Other Counties

                                              Average Monthly FPE                                          Average Monthly Cost Per FPE

         County                                                   Percent      Percent                                                Percent      Percent
                                       Actual      Projected                                               Actual      Projected
                        Baseline                                (Baseline to (Baseline to    Baseline                               (Baseline to (Baseline to
                                    (3/00 - 6/00) (3/00 - 6/00)                                         (3/00 - 6/00) (3/00 - 6/00)
                                                                  Actual)     Projected)                                              Actual)     Projected)

     Arapahoe *           141           128          102          -9.2%        -27.4%         $750         $852           $919        13.6%        22.5%

     El Paso *             70           74            62          5.3%         -10.9%         $728         $730           $873        0.2%         19.9%

     Pueblo *              45           48            38          6.0%         -14.6%         $785         $858           $959        9.2%         22.2%

     Boulder **            25           38            32          52.9%         30.7%         $753         $755           $838        0.2%         11.3%

     Jefferson **          99           144          134          45.3%         34.8%         $685         $853           $782        24.6%        14.2%

     Mesa **               24           24            23          -3.2%         -7.9%         $772         $825           $909        6.8%         17.6%

*
     Baseline = Monthly data from 7/98 to 2/00 for Arapahoe, El Paso, and Pueblo that started managed care since SFY99.
**
     Baseline = Monthly data from 7/97 to 2/00 for Boulder, Jefferson, and Mesa that started managed care since SFY98.




Overall, the results indicate that Arapahoe County reduced its average monthly FPE more
than any other managed care county, compared to the baseline period, after the beginning
of the contract with CCM. During the same period, the average monthly cost increased
more than any other county except Jefferson. It appears, then, that Arapahoe County got
results, but had to pay for them.

These results need to be regarded as tentative, as only four months of data were compared
to the baseline period. A longer period is needed for more confidence.

Another potential issue is the baseline period that was selected. The baseline period was
based on the period from a county obtaining managed care status through February 2000.
Different baseline periods might be chosen, such as the four months immediately before
the agreement with CCM was signed, with different results.

Finally, the results might be interpreted to mean that the agreement with CCM had an
impact on average FPEs per month. This is about the same time, however, that the staff
member responsible for managed care in Arapahoe County took a more assertive role in
the utilization management of RTC cases. The results could have been a result of his
intervention, the contract with CCM, both, or some other factor that has not been
identified.




William M. Mercer, Incorporated                                              290                     Colorado Child Welfare Evaluation
                                                                                            Second Interim Implementation Status Report
Discussion and Recommendations

Colorado formally entered the managed care arena in 1997 with Senate Bill 97-218. This bill
capped allocations and increased the focus on performance and outcomes while providing counties
with a greater degree of flexibility in the use of funds than they had in the past. It allowed for the
establishment of three managed care counties and gave them the ability to negotiate a limited
number of rates and payment methods with providers, to retain unspent general funds and invest
them in additional services, and to waive certain rules and regulations. Senate Bill 98-165
authorized three more managed care pilot counties and allowed any county to develop an MOU
with the CDHS to become a managed care county beginning in SFY 2001.

Three counties—Boulder, Jefferson, and Mesa—became managed care pilot counties in SFY 1998.
Three others—Arapahoe, El Paso, and Pueblo—became managed care pilot counties in SFY 1999.
At the time of this writing, June 2001, the first three counties have been managed care counties for
almost four years while the second three have been managed care counties for almost three years.
The outcomes and cost data in this report, which go through SFY 2000, encompass three years
under managed care for the first set of counties and two years for the second set.

As was the case in last year’s report, this report shows that Colorado’s ten largest counties continue
to perform better than comparison states and the national average in a number of areas. For
example, there is a lower rate of confirmed abuse and neglect in Colorado and Colorado children
are less likely to be in out-of-home placement during the year. In addition, Colorado children spend
less time overall in out-of-home care.

Colorado also has programs that have drawn national attention. El Paso County’s creative use of
TANF funds and adoption efforts have been the subject of published articles. Boulder County’s
IMPACT program has also been recognized nationally. To the extent that those programs have been
supported through the spending and regulatory flexibility of the managed care pilot initiative,
Colorado can attribute those process achievements to the managed care initiative.

While it is clear that Colorado is performing well in many areas, there is little evidence to date, in
terms of outcomes and costs, that this performance is the result of the introduction of managed care
into Colorado’s child welfare system. Overall, managed care pilot counties show results that are
similar to those of a set of counties that are not managed care pilot counties. We arrived at the same
conclusion last year and attributed this to three possible factors. First, we noted that it takes
considerable time to implement a complex program like managed care. Second, we pointed out that
outcomes depend on processes and that early studies need to focus on determining that the
appropriate processes are in place and are of high quality. Finally, we noted that all counties did not
start from the same point and that some are more mature than others in the development of
particular processes.

To address these three factors, this year’s evaluation examined each county’s implementation of six
key aspects of managed care – utilization management, quality improvement, provider network
capacity, performance indicators, financial incentives, and management information systems. We
found considerable variations across counties in some processes such as utilization management. In
some areas, such as the use of financial incentives, we found a few counties using them or

William M. Mercer, Incorporated                 291                  Colorado Child Welfare Evaluation
                                                                   Interim Implementation Status Report
exploring their use, while most counties had not introduced them. Finally, some areas, such as
determining the network capacity required to appropriately meet the services needs of families and
having an adequate management information system, were found to be problems for all the
managed care counties.

Until the counties are able to successfully implement these key aspects of managed care, it is the
view of the evaluators that it will be difficult for them to generate a detectable level of
improvement, as a result of managed care, in outcomes and costs. It is our view that the child
welfare system, as a whole, needs to revisit its vision for managed care given what it has learned,
develop a strategic plan to carry out that vision in a logical sequence, and develop a detailed annual
work plan to guide the implementation. It is also our view that neither the CDHS nor the counties
have received the degree of support and technical assistance necessary for the successful
implementation of managed care; this needs to be provided.

These views provide the basis for the following recommendations. We have kept the number of
recommendations to a minimum as each one is substantial, requiring a significant investment of
resources.

§    Revisit and revise Colorado’s vision for managed care in child welfare and develop a strategic
    plan and annual work plan. The CDHS and counties should do this in partnership through the
    Managed Care Implementation Team (MCIT). The MCIT should also be used as the primary
    vehicle for guiding the systematic implementation of the work plan. The MCIT is one of the most
    valuable tools in Colorado for the implementation of managed care as long as it has a clear focus
    and a logical sequence of tasks.
§     Clarify the financial risk arrangement between the CDHS and the counties. The first step is to
    rationalize the child welfare allocation. The original financial mechanism of allowing counties to
    retain unspent funds for service development in return for foregoing participation in any year-end
    surplus distribution requires refining. The stop-gap measure this past year of allowing counties to
    be treated as non-managed care counties was important in helping them remain managed care
    counties despite financial difficulties. This approach does not seem to make sense over the longer
    term. Evolving this financing arrangement into a risk sharing arrangement with a financial
    incentive for containing costs while retaining some measure of protection through a common risk
    pool (e.g., the surplus distribution) should be a major goal of next year’s MOU development
    process.
§     Identify technical assistance and support needs in key managerial process areas and arrange
    for their provision. While all six process areas noted above (management information systems,
    utilization management, local financial incentives, performance indicators, quality improvement,
    and provider network capacity) are areas that could benefit from technical assistance, we
    recommend three target areas in the next year.




William M. Mercer, Incorporated                  292                  Colorado Child Welfare Evaluation
                                                                    Interim Implementation Status Report
§     Management information systems. Managed care requires a strong financial data
    management component and most counties do not currently possess this. Instead, most
    counties rely upon reporting from the state that was never envisioned as sufficient for a
    managed care environment with its premium upon real-time, tailored and flexible reporting.
    The CDHS has planned for TRAILS to provide some functionality in this area but it is
    recommended that the CDHS set a strategic goal of helping each county to develop its own,
    independent financial tracking and reporting capacity.
§     Utilization management. All managed care counties have tried various strategies to impact
    the use of expensive out-of-home placements. The CDHS and the counties are now in the
    position of being able to look across the various approaches that have been tried, pick those
    that have been most successful, refine them, and begin to standardize expectations and
    strategies across counties. While differences will remain, areas of consensus (e.g., review
    elements to include in the prior authorization process for out-of-home care) are emerging and
    should be disseminated and supported. It is recommended that the CDHS set a strategic goal of
    developing cross-county utilization management standards and support counties in
    implementing them.
§     Contract management functions that combine financial incentives and performance
    indicators. Counties such as El Paso and Boulder have developed contract management
    approaches and capacity that begin to move this area. Pilot counties such as Jefferson, which
    belong to the Northern Consortium, have begun to address these matters in their relationships
    with CPAs. It is recommended that the CDHS set a strategic goal to support contract
    approaches predicated on clear financial incentives and performance indicators and support
    counties in developing and managing them over time.

This evaluation report should be used as a tool in the coming year at both the state and county
level. The strengths as well as the areas needing improvement in each county are fairly evident in
this report. Where outcomes in a particular county are considerably better than those in other
counties, the processes related to those outcomes should be examined for possible adoption in
other counties, especially in counties with the lowest level of performance in an area. Harnessing
the evaluation as a major vehicle for disseminating findings in support of the strategic goals and
associated technical assistance recommended above should be a primary focus of evaluation
activities during the next year.

Colorado continues to make major changes in its child welfare system. The evaluation shows
that there are good outcomes in a number of areas but they cannot yet be clearly ascribed to
managed care. The CDHS and counties need to use the findings of this report to refine the
implementation of managed care so that its potential for improving outcomes for Colorado
children and families may be increased.




William M. Mercer, Incorporated                 293                 Colorado Child Welfare Evaluation
                                                                  Interim Implementation Status Report

								
To top