ferry by zhangyun

VIEWS: 1 PAGES: 16

									     Expanding Medicaid May
    Cost Less Than You Think:
         Evidence from Health
    Insurance Churning in the
             1998-2002 MEPS
                         Danielle H Ferry, BA
                           Kathryn Klein, BA
                           Sherry Glied, PhD



Supported by a grant from The Commonwealth Fund
Many low-income individuals
lack health coverage
   Ineligible for Medicaid
   Eligible for Medicaid, but not enrolled b/c:
       Enrollment procedures difficult/confusing
       Recertification procedures difficult/confusing
       2002: Reporting required every 3 or fewer months
        (2 states), 4-11 months (11 states), 12 or more
        months (37 states)
   Lack of/instability in coverage  unmet
    health needs
Study Objective
   To examine the health care spending &
    utilization patterns of those entering & exiting
    Medicaid in order to inform estimates of the
    cost of expanding Medicaid or increasing
    coverage through simplified enrollment &
    recertification procedures
Current cost estimates may be
flawed
   Most estimates use per-enrollee costs of
    current Medicaid enrollees
   BUT: Evidence of adverse selection
       Holahan (2001)
       Birnbaum et al. (2004)
       Long, Coughlin, & King (2005)
Data
   MEPS Panels 3-6 (1998-2002)
   Sample
       Adults ages 19-62 in first year on panel
       Determinable insurance status in all 24 months on
        panel
        45,650 respondents
   Monthly health expenditures & service
    utilization
   Monthly health insurance transitions
Health Insurance Transitions



  Insurance       X            Y            Y
  Month          m-1          m           m+1
  Time        1st month    month of     1st month
                prior to   transition      post
              transition                transition
Study Design: Cost of
Expanding Medicaid
   Compare average monthly health
    expenditures of:
       Individuals on Medicaid 24 months
       Individuals who transitioned: uninsured 
        Medicaid - in 1st month prior to transition (while
        uninsured)
       Individuals who were uninsured 24 months
Study Design: Cost of
Simplified Recertification
   Compare average monthly health
    expenditures of:
       Individuals on Medicaid 24 months
       Individuals who transitioned: Medicaid  No
        coverage or Medicaid  Private group insurance
        - in 1st month prior to transition (while on
        Medicaid)
Health Insurance Expenditures
   Inflation-adjusted
   “Unadjusted” mean expenditures
       Simple means
   “Adjusted” mean expenditures
       Two-part model
         I: Pr(Y>0)  logit
         II E(Y|Y>0)  GLM w/ Gamma distribution & log link
       Adjustments: Age, sex, race/ethnicity, education, poverty
        status, health status, region of residence
   All SEs adjusted for single-stage random sampling
    design of the MEPS & possibility of individuals
    having multiple transitions
Medicaid Enrollees vs. the
Uninsured: Total Expenditures


Mean monthly                   UNINSURED
expenditures          Always                               Always
($):                 Uninsured  Medicaid                   Medicaid
Unadjusted                48               207*               266*

Adjusted                   9               136*               223*


* Statistically different from “Always Uninsured” at the 5% level
Medicaid Enrollees vs. the
Uninsured: Expenditures by Type

Mean monthly           UNINSURED
expenditures  Always                                       Always
($):         Uninsured  Medicaid                           Medicaid
Physician                 23                61*               114*

Hospital                  22               133*               160*

Rx                         4                 13                17*

* Statistically different from “Always Uninsured” at the 5% level
+ Unadjusted expenditures
Medicaid Enrollees vs. the
Uninsured: Utilization by Type

                                 UNINSURED
Mean monthly            Always             Always
utilization (%):       Uninsured  Medicaid Medicaid
Physician visits            7.8             22.6*             31.0*
Hospital admits             0.4              2.7*              1.9*
ER visits                   1.2              5.0*              3.1*
Rx’s                        3.6             11.2*             10.1*
* Statistically different from “Always Uninsured” at the 5% level
+ Unadjusted shares
Medicaid Dis-Enrollees vs.
Long-term Medicaid Enrollees:
Total Expenditures


Mean monthly                  MEDICAID   MEDICAID
expenditures          Always                 
($):                 Medicaid Uninsured Private Group
Unadjusted               266            172*                151

Adjusted                 223            160*                116*


* Statistically different from “Always Medicaid” at the 5% level
Robustness Checks
   Looked backward/forward 2 & 4 months from
    transition
   Limited analysis to those with chronic
    conditions
   Incl’d cost of uncompensated care in
    expenditures of any part-year uninsureds
Results
   Adj. health care exp. among the persistently
    uninsured < 1/10 of pre-transition exp. of
    those uninsured who switched to Medicaid
   Pre-transition adj. health care exp. of those
    who dropped Medicaid & became uninsured
    < 3/4 as high as those of long-term Medicaid
    enrollees
   Adj. Medicaid exp. of those who switched to
    private group ins. = 1/2 those w/ stable
    Medicaid coverage
Conclusions
   There is adverse selection into Medicaid & it cannot
    be readily attributed to observable characteristics
   Enrollment & recertification procedures likely
    represent greater barriers to Medicaid coverage
    among those with fewer health needs
   Extrapolation from current per-enrollee Medicaid
    costs leads to overestimates of the cost of
    expanding Medicaid coverage to the uninsured
    and/or improving coverage rates among the current
    eligible population

								
To top