Correcting a Texas Birth Certificates

Document Sample
Correcting a Texas Birth Certificates Powered By Docstoc
					       The Culture of Data: Bridging the
        gap between health and socio-
                economic status
<<!-- PICOTITLE= "Health Disparities in
Colorado 2005: A Call to Action American
Indian data, issues and experiences as
examples" --> <!-- PICODATESET
mmddyyyy=10212005 -->




          Health Disparities in Colorado
             2005: A Call to Action
         American Indian data, issues and
            experiences as examples

                                      October 21, 2005
                                    8:30 a.m. – 9:30 a.m.
                      The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   1
 Linda Burhansstipanov, MSPH, DrPH, CHES
          (Cherokee Nation of OK)
             Executive Director
     Native American Cancer Research
           3022 South Nova Road
            Pine, CO 80470-7830
Phone: 303-838-9359; Fax: 303-838-7629
Native Cancer Survivor‟s Support Network:
               1-800-537-8295
      Privacy Broker: 1-877-838-9359
    Web Page: http://NatAmCancer.org

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   2
This presentation is dedicated to:




    Please add her to your prayers.
    New Primary Cancer: stomach

  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   3
        Culture of Data Objectives

1) Describe two examples of how racial
   misclassification occurs regardless of how
   careful data collectors implement quality
   control strategies.
2) Describe examples of American Indian
   cancer data issues.
3) Describe at least two strategies that are
   successful in improving the accuracy of
   American Indian cancer data.


     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   4
  We (communities, epidemiologists,
researchers, etc.) collaboratively need
   to address the barriers through
    culturally competent strategies
       (i.e., the Call to Action)
  Cancer




The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   6
Why it is Important to Collect
Accurate Native American Data

• Documents the cancer problem (which
  subsequently affects funding)

• Helps document the number and
  characteristics of Natives who take
  part in cancer programs (prevention,
  early detection, treatment, survival,
  palliative care)

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   7
Why it is Important to Collect
Accurate Native American Data
• Helps document the number of
  Natives who are diagnosed with
  cancer and may help figure out why
  our survival is so much poorer than
  other ethnicities, even when cancer
  stage / histological grade are
  matched



  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   8
Why it is Important to Collect
Accurate Native American Data

• Helps prove the successfulness of the
  tribal program in reaching un /
  underserved Natives (i.e., increases
  the likelihood of continued funding)
• Helps identify areas within the tribal
  community that are not being reached
  by the program


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   9
               Cancer Data Issues


Racial misclassification (ends up
undercounting) ...

According to Dr. Ed Sondik, Director,
NCHS, AIAN undercounts ~ 36%)

Coding errors (universal to people of all
races)

 The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   10
                Cancer Data Issues

Collapsing the diverse smaller population
groups into an “other” racial category ...
“Other” racial data

         Lose all racially specific information
         and cultural relevance
         “Are of no use when attempting to
         develop, assess, and monitor public
         health programs and services”
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   11
Most Common Reasons for Racial
       Misclassification

1.          Use of Spanish surnames to
            determine race / ethnicity

2.          Subjective use of personal
            observation by the data collector

3.          AIAN not a response category in
            medical records (e.g., hospital,
            health clinic)

     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   12
Most Common Reasons for Racial
       Misclassification

4.          Imprecise and inconsistent
            definitions of AI

5.          Changing self-identification

6.          Tribe formerly “unrecognized”


     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   13
Most Common Reasons for Racial
       Misclassification

7.          Tribal enrollment blood %
            ordinances changing

8.          Tribal enrollment ordinances re
            Paternal or Maternal lineage




     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   14
Every National and State Database
         has Limitations
Database: U.S. Census
Purpose: To identify descriptive
  demographic information for all
  population units
Issues (varies by geographic areas):
  – Undercount of Native Americans
  – Racial misclassification
  – Insensitive to annual fluctuations
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   15
Every National and State Database
         has Limitations
Database: U.S. Census (continued)
Effects the of Limitations:
  –Affects accuracy of the data used as
   the denominator to generate rates
  –Underestimates the number of
   American Indians and Alaska Natives
  –In comparison, it overcounts other
   racial groups
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   16
Every National and State Database
         has Limitations
Database: CDC National Center for
  Health Statistics (NCHS)
Purpose: To identify the frequency of
  deaths by cause (based on death
  certificate information provided by
  each state)
Issues (varies by geographic areas):
  – Racial misclassification

    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   17
Every National and State Database
         has Limitations
Database: CDC NCHS (continued)
Effects the of Limitations:
  –Undercounts of numerator data is
   likely to have greater impact in
   underestimating incidence and
   mortality (death) rates in American
   Indian and Alaska Native populations
  –Lesser impact of overestimating rates
   in the general population.
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   18
Every National and State Database
         has Limitations
Database: National Cancer Institute
  (NCI) Surveillance, Epidemiology, and
  End Results (SEER)
Purpose:
   – To analyze trends in cancer incidence,
     mortality (based on NCHS data), and
     patient survival in the United States


    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   19
Every National and State Database
         has Limitations
Database: NCI SEER (continued)
Purpose (continued):
  – To analyze trends in cancer incidence,
    mortality (based on NCHS data), and
    patient survival in the United States
Issues (varies by geographic areas):
  – Racial misclassification
  – Most data are from SW tribes
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   20
Every National and State Database
         has Limitations
Database: NCI SEER (continued)
Effects the of Limitations:
  –Affects the accuracy of the numerator
   by underestimating the number of
   American Indians and Alaska Natives;
  –Southwestern American Indian data are
  NOT generalizable to Indigenous
   Peoples from other regions of U.S.

    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   21
Every National and State Database
         has Limitations
Database: Indian Health Service (IHS)
Purposes:
  – To calculate IHS "user population"
  – To calculate IHS patient care rates;
  – NOTE: IHS "user population" figures
    are used in calculating vital event
    rates since state birth and death
    certificates do not provide
    information on use of IHS services
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   22
Every National and State Database
         has Limitations
Database: IHS (continued)
Issues (varies by geographic areas):
  –Quality control related to medical records,
   e.g., data entry and coding of diagnosis
  –Limited to 33 reservations states
  –Partial and/or lack of urban American Indian
   and Alaska Native cancer data
  –Limited to its "user" population (i.e., used a
   facility within the last three years)

    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   23
Every National and State Database
         has Limitations
Database: IHS (continued)
Effects the of Limitations:
  –Data represent small portion of
   reservation American Indians and not
   the majority of the Indigenous
   populations (contrary to IHS
   estimations of percentage who use)

 –Ex. CA=largest numbers of AIAN of
   any state, but less than 10% use IHS
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   24
    So what does this mean for our
           tribal programs?
• We need to collect and store accurate
  information about our own communities
• Use selected national databases when
  necessary, but recognize the limitations
• Of the four National databases, for
  cancer in Indian Country, IHS appears
  to be the “best”


     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   25
     AIAN Cancer Data Key Points

Unavailable accurate national / regional /
local cancer surveillance data for AI/AN

Data from one geographic region does
not accurately reflect cancer patterns
for Natives in another region

Example: The most accurate AIAN
cancer data resources: AK and NM
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   26
The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   27
Table 1. Age Adjusted Cancer Mortality Data
                              New Mexico AI Females (♀) and Males (♂)                                           Alaska AN Females (♀) and Males (♂)
Cancer Site              IHS ABQ ♀ a        SEER NM ♀ b          IHS ABQ ♂ a        SEER NM ♂ b         IHS AK ♀ a         SEER AK ♀ b         IHS AK ♂ a          SEER AK ♂ b

All Sites                   100.0                  99.0             134.4                123.0            187.0                179.0            218.6                  225.0
Breast                        10.3                   8.7                   #                    #           21.4                 16.0                   #                       #
Cervix                          4.0                  8.0                                                      5.2                      *
Ovarian                         9.6                  7.3                                                      6.5                      *
Prostate                                                              21.0                 16.2                                                      7.4                        *
Colon &                         5.5                      *            15.8                   8.5            24.3                 24.0             27.0                   27.2
Rectum
Kidney & Renal                  7.9                      *              9.1                      *           11.9                  7.4            14.1                   13.4
Pelvis
Lung &                        11.5                       *              9.2                10.4             46.5                 45.3             70.3                   69.4
Bronchus
Stomach                         6.7                      *              8.9                11.4               5.2                      *          18.9                   18.9
Gallbladder                     4.8                  8.9                 12                     #             3.2                      *             2.1                        #
a.    Cobb N, Paisano RE. Cancer Mortality among American Indians and Alaska Natives in the United States: Regional Differences in Indian Health, 1989-1993. Indian Health
       Service. IHS Pub. No. 97-615-23. Rockville, MD, 1997.
b.    Miller B.A., LN Kolonel, L. Bernstein, JL Young, GM Swanson, DW West, CR Key, JM Liff, CS Glover, GA Alexander, L. Coyle, BF Hankey, C Percy. Racial/Ethnic Patterns of
       Cancer in the United States 1988-1992. National Cancer Institute. Pub. No. (NIH) 96-4104. 1996.
*     rate not calculated when fewer than 25 deaths
#     data not available

Burhansstipanov L, Hampton JW, Wiggins C. Issues in Cancer Data and Surveillance for American Indian and Alaska
Native Populations. Journal of Registry Management. 1999:29:4:153-157.
   Tribal and Geographic
         Variability

Higher incidence and mortality among
Northern Plains tribes than in the
Southwest
           comparable to whites and African-
           American rates
                             lung, breast, prostate, cervix


 The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   29
Which tribal Nations are living in CO?

   43% of Natives live in greater Denver
   Metropolitan Ares
   11% are CO Reservations (Ute Mountain
   Ute and Southern Ute)

   Majority of Natives living in CO are
   from tribal Nations based in other
   states


     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   30
Which tribal Nations are living in CO?

   Greater Denver Metropolitan Area:
   ~30% Northern Plains
   ~25% Oklahoma (data are grouped in
   “East” data on the IHS Mortality slides)
   ~25% Southwest (AZ, NM)

   CO data are grouped with the “SW”
   data on the IHS Mortality slides
     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   31
               Updated data from IHS


Roberta Paisano, MHSA and David
Espey, M.D., Indian Health Service
National Epidemiology Program

“Best” source of data from any federal
agency

Clearly illustrates regional variations


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   32
                     Background

Cancer continues to escalate among
AIAN (i.e., we are not benefiting from the
decreases in cancer mortality observed in
other populations and our incidence is still
increasing)
Cannot rely on NCI, NCHS, CDC
Registries for accurate AIAN data
Note on “IHS mortality slides”, the location
of SW on the rankings
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   33
                        Background
 Cancer continues to escalate among
 AIAN (i.e., we are not benefiting from the
 decreases in cancer mortality observed in
 other populations and our incidence is still
 increasing)
 AIAN continue to have the poorest 5-year
 relative survival from “all cancers” in
 comparison to all other ethnicities in the US
               Data from Indian Health Service (IHS)
Special thank you to Roberta Paisano, MHSA and David Espey, MD
     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   34
IHS Age-adjusted Malignant Neoplasm Death Rates




      IHS Trends, 1999, Chart 4.33, p. 116
      AI/AN cancer rates for all cancers, by
     region, both sexes, compared to US all
              race rates, 1999-2002
     600

     500

     400

     300

     200

     100

       0




                                                                  t
                               ns




                                                       t
           ka




                                                                   es
                                        t
                    US




                                                    as
                                         s
                                ai




                                      Ea
           as




                                                                 hw
                                                 Co
                             Pl
        Al




                                                              ut
                                              P.
                          N.




                                                           So
Special thanks to David Espey, MD and Roberta Paisano, MHSA, IHS National
                           Epidemiology Program
  Leading AI/AN Cancer Incidence Rates - Both Sexes,
    Northern Plains, 1999-2002 (cases/100,000 pop)




David Espey, MD; CDC/DCPC/EARB, IHS, IHS Research Conf 6-10-05
 Leading AI/AN Cancer Incidence Rates -
               Both Sexes,
  Alaska, 1999-2002 (cases/100,000 pop)




David Espey, MD; CDC/DCPC/EARB, IHS, IHS Research Conf 6-10-05
IHS Five Geographic Regions
     WA
                         MT            ND
                                                 MN
 OR
                                       SD                  WI                           ME
           ID             WY                                         MI

                                       NE             IA
                                                                                   NY     MA
                                                                 IN                       RI
           NV                                                                 PA        CT
CA                UT
                              CO
                                             KS

                AZ                                                            NC
                                                                     TN
                         NM                      OK
                                                                             SC
                                                                      AL
                                                                MS
      AK
                                            TX
                                                           LA

                                                                             FL




                ALASKA             NORTHERN PLAINS

                EAST               PACIFIC COAST                 SOUTHWEST
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
                  Breast, By Region, 1994-98




                          Both Sexes                    Males          Females
                         N     Rate               N        Rate      N    Rate

U.S. All Races                      13.4                       0.3         24.0
All IHS Regions            385       7.9**             2       0.1   383   14.2**

Alaska                      33      11.1               0       0.0    33   20.8
East                       109       7.5**             1       0.2   108   13.4**
No. Plains                  75      12.0               0       0.0    75   21.6
Pacific Coast               92       8.8**             0       0.0    92   16.2**
Southwest                   76       5.2**             1       0.2    75    9.2**



 ** Denotes a rate significantly different from the US rate.
                      Age-Adjusted Cancer Mortality Rates
                      Breast, By Region, Females, 1994-98




  Southwest
                                                            All IHS (14.2)
        East

Pacific Coast                                                  All US (24.0)

      Alaska

  No. Plains


                0         5          10          15          20          25    30
                Rate per 100,000 per year, adjusted to 1970 U.S. population

   LB -- sorry about scale changing on these bar graph slides
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
                 Cervical, By Region, 1994-98




                          Both Sexes                    Males          Females
                         N     Rate               N        Rate      N    Rate

U.S. All Races                        1.4                      0.0          2.6
All IHS Regions            108        2.0**            0       0.0   108    3.7**

Alaska                       3        0.7              0       0.0     3    1.5
East                        37        2.4**            0       0.0    37    4.3**
No. Plains                  19        2.6**            0       0.0    19    4.7**
Pacific Coast               15        1.3              0       0.0    15    2.4
Southwest                   34        2.1**            0       0.0    34    3.9**



 ** Denotes a rate significantly different from the US rate.
                        Age-Adjusted Cancer Mortality Rates
                       Cervical, By Region, Females, 1994-98




      Alaska
                                                 All US (2.6)
Pacific Coast
                                                         All IHS (3.7)
  Southwest

        East

  No. Plains


                 0      1      2      3      4      5      6      7       8   9   10
                Rate per 100,000 per year, adjusted to 1970 U.S. population
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
                  Ovary, By Region, 1994-98




                          Both Sexes                    Males              Females
                         N     Rate               N        Rate          N    Rate

U.S. All Races                   4.1                       0.0                 7.4
All IHS Regions            141         2.9**           0         0.0**   141         5.2**

Alaska                      13         4.7             0         0.0      13         8.9
East                        35         2.3**           0         0.0      35         4.1**
No. Plains                  27         4.2             0         0.0      27         7.5
Pacific Coast               20         1.9**           0         0.0      20         3.5**
Southwest                   46         3.3             0         0.0      46         5.9



 ** Denotes a rate significantly different from the US rate.
                        Age-Adjusted Cancer Mortality Rates
                        Ovary, By Region, Females, 1994-98




Pacific Coast
                                                     All IHS (5.2)
        East
                                                                  All US (7.4)
  Southwest

  No. Plains

      Alaska


                 0                      5                      10                15
                Rate per 100,000 per year, adjusted to 1970 U.S. population
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
              Colon/Rectum, By Region, 1994-98




                          Both Sexes                    Males             Females
                         N     Rate               N        Rate         N    Rate

U.S. All Races                 16.9                    20.2                 14.4
All IHS Regions            642     13.7**          320     15.6**       322     12.0**

Alaska                      83      32.5**            33       27.9      50    35.8**
East                       176      11.9**            89       13.9**    87    10.3**
No. Plains                 160      27.7**            85       34.4**    75    22.6**
Pacific Coast               94       9.2**            49       10.9**    45     7.7**
Southwest                  129       9.6**            64       10.8**    65     8.6**



 ** Denotes a rate significantly different from the US rate.
                     Age-Adjusted Cancer Mortality Rates
                    Colon/Rectum, By Region, Both, 1994-98




Pacific Coast

  Southwest
                                             All IHS (13.7)

                                              All US (16.9)
        East

  No. Plains

      Alaska


                0    5    10     15     20     25    30     35     40     45   50
            Rate per 100,000 per year, adjusted to 1970 U.S. population
                      Age-Adjusted Cancer Mortality Rates
                    Colon/Rectum, By Region, Males, 1994-98




  Southwest

Pacific Coast                               All IHS (15.6)

                                               All US (20.2)
        East

      Alaska

  No. Plains


                0     5   10     15    20     25     30     35     40     45   50
            Rate per 100,000 per year, adjusted to 1970 U.S. population
                         Age-Adjusted Cancer Mortality Rates
                     Colon/Rectum, By Region, Females, 1994-98




Pacific Coast

  Southwest                                 All IHS (12.0)

                                                 All US (14.4)
        East

  No. Plains

      Alaska


                 0      5     10     15     20     25     30      35     40   45   50
                Rate per 100,000 per year, adjusted to 1970 U.S. population
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
                 Prostate, By Region, 1994-98




                          Both Sexes                    Males             Females
                         N     Rate               N        Rate         N    Rate

U.S. All Races                   9.3                   23.4                   0.0
All IHS Regions            288         6.4**       288     15.4**         0         0.0

Alaska                      12       5.0**            12       11.4**     0         0.0
East                        88       5.6**            88       14.0**     0         0.0
No. Plains                  74      13.8**            74       34.6**     0         0.0
Pacific Coast               41       4.4**            41       10.2**     0         0.0
Southwest                   73       5.9**            73       14.1**     0         0.0



 ** Denotes a rate significantly different from the US rate.
                     Age-Adjusted Cancer Mortality Rates
                     Prostate, By Region, Males, 1994-98




Pacific Coast

                                                   All IHS (15.4)
      Alaska
                                                            All US (23.4)
        East

  Southwest

  No. Plains


                0     5      10       15      20       25      30         35   40
            Rate per 100,000 per year, adjusted to 1970 U.S. population
Total Number of Deaths and Age-Adjusted Cancer Mortality Rates,
                   Skin, By Region, 1994-98




                          Both Sexes                    Males              Females
                         N     Rate               N        Rate          N    Rate

U.S. All Races                   2.2                       3.1                 1.5
All IHS Regions             31          .6**          12          .6**    19          .7**

Alaska                       3          .9             1          .3**     2         1.5
East                        12          .8**           4          .6**     8          .9**
No. Plains                   5          .7**           2         1.0       3          .7
Pacific Coast                2          .2**           1          .2**     1          .2**
Southwest                    9          .6**           4          .8**     5          .5**



 ** Denotes a rate significantly different from the US rate.
                     Age-Adjusted Cancer Mortality Rates
                       Skin, By Region, Both, 1994-98




Pacific Coast

  Southwest                        All IHS (.6)

  No. Plains                                All US (2.2)

        East

      Alaska


                0             1                 2                3        4
            Rate per 100,000 per year, adjusted to 1970 U.S. population
                     Age-Adjusted Cancer Mortality Rates
                       Skin, By Region, Males, 1994-98




Pacific Coast

                                  All IHS (.6)
      Alaska

        East                                               All US (3.1)

  Southwest

  No. Plains


                0             1                 2                3        4
            Rate per 100,000 per year, adjusted to 1970 U.S. population
                        Age-Adjusted Cancer Mortality Rates
                         Skin, By Region, Females, 1994-98




Pacific Coast

                                        All IHS (.7)
  Southwest
                                                      All US (1.5)
  No. Plains

        East

      Alaska


                 0                 1                 2                 3      4
                Rate per 100,000 per year, adjusted to 1970 U.S. population
QUERY: Why are the data so poor /
inaccurate for AI living in Colorado?
 Racial misclassification
 Multiple races (self-identification on
 Census) adds a new level of complexity
 The most common insurance carrier for
 AI living in Colorado does not collect
 racial data
 Which “formula” is the most accurate
 for our state?
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   56
QUERY: Why are the data so poor /
inaccurate for AI living in Colorado?
 People go home to reservation for care

 When they go home, they have to re-
 establish “residency” (typically 6
 months)

 Once residency is established, they can
 apply for Contracted Health Services
 (quick lesson next slides)

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   57
 IHS Contracted Health Services

IHS / Tribes / Urban programs “prioritize”
health conditions each fiscal year for
referral to care not available from local
services (called CHS)


IHS CHS under-funded by 40-60% of the
documented health care needs of the
local communities
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   58
                           IHS CHS


i.e., each year the community must
decide which types of health problems
experienced by 40-60% of their
community will go without care …


Not IHS’s fault, but it is in part due to
Congress for under-funding CHS

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   59
                          IHS CHS


Fiscal year begins October 1

During federal freezes, IHS CHS monies
are also “frozen” except for life and
death emergencies

Most tribes run out of CHS monies by
early summer


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   60
                           IHS CHS

Because of the under-funding, if a health
problem is listing lower than “5” for most
tribes, the patient is not referred out for
care
Of the 564 federally recognized tribes,
“cancer” is ranked among the top 5 in
less than 10 Nations



   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   61
                           IHS CHS


Not unusual to have 6 to 9 month
interval from diagnosis to initiation of
cancer care
Indian casinos have helped purchase
PRIVATE health insurance that usually
addresses the CHS “issues”



   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   62
                              2000 US Census:
                              Demographics
  American Indian and Alaska Native
  (AIAN) population is increasing at
  about 1.8% a year

  US Census 2000, 2.2 million reported
  AIAN Race “alone” (4.1 million AIAN
  alone or in combination with other
  race(s)
28% increase in “AIAN alone” since 1990 Census

     The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   63
                           Demographics


Average life span is about 5 years less
than for “US All Races”
AIAN Median household income =
$32,116

Approximately 24.5% of AIAN live
below the federal poverty level


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   64
Demographics: where do AIAN live?


 43% of AIAN lived in the West
 31% lived in the South
 17% lived in the Midwest

 9 % in the Northeast
 Approximately 60% live in urban areas


   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   65
Demographics: Largest Tribal Nations
  American Indian tribes with 50,000 or
  more individuals:
      Cherokee
      Navajo
      Choctaw
      Blackfeet
      Chippewa
      Muscogee
      Apache
      Lumbee
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   66
                Demographics: states with
 with more than 100,000 AIAN
 States
                       >100,000

  residents.
      California (628,000)
      Oklahoma (392,000)                                        North Carolina
      Arizona                                                   Michigan
      Texas
                                                                Alaska
      New Mexico
                                                                Florida
      New York
      Washington
    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   67
    Demographics: states with proportion
 of states where the AIAN
 Number
                  >1.5%

  population as a proportion of the total
  population exceeded the national
  average of 1.5%:
      Alaska (19 percent)
      Oklahoma (11 percent)
      New Mexico (10 percent)
      The other 16 states: AZ, CA, CO,
      ID, MT, NV, HI, OR, UT, WA, WY,
      KS, MN, ND, SD, NC
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   68
Demographics: states with >100,000
  The median age for AIAN population
  was 28.7 years
  The median age for the total U.S.
  population was 35.3 years.

  71% of AIAN alone over 24 years have
  at least a high school education

  11% of AIAN have a bachelor‟s degree
  or higher

    The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   69
Mortality compared with “US All Races”
    AIAN greater/more i.e., “the health disparities”
             627% alcoholism
             533% tuberculosis
             249% diabetes mellitus

             204% accidents
             72% suicides
             71% pneumonia and influenza
             63% homicides
      The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   70
Some Possible Interventions

   Includes both those that failed
    and those that are successful



  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   71
Less Successful Strategies (or
      those that failed)

Waiting for the government to “fix”
the problems
Relying on one or two key, dedicated
individuals in high level positions (i.e.,
they move on and are replaced by
others who do not share their
enthusiasm to improve the data for a
group that is so small)
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   72
 Less Successful Strategies (or
       those that failed)

SEER and CDC Registries
    NOTE: a few exceptions = NMTR,
    ANMC
    Another Exception: The New
    NCI/CDC Northern Plains
    SEER/Cancer Registry
YEE HAW … finally! thank you thank you thank you!

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   73
Less Successful Strategies (or
      those that failed)

Use of Surveillance Systems
   Not designed for “studies”

   Less than 15% accurate for AIAN




  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   74
What we learned from “Increasing
Mammography for Urban American
 Indians” [NCI R25 CA 77665]
 Surveillance programs are insufficient
 to serve as a sampling source

 1/3 of the women identified as
 “American Indian” were actually “Asian”

 Marked the wrong race on the form?

 Very mobile population (50% changing
 phone numbers and addresses within 18
 months)
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   75
        Alternative data source:
         Ethnographic database
 Tailored to American Indian women for
 marketing purposes
 Database included errors ...


insufficient accuracy = <10 AIAN out of
sample of 113 supposed AIAN living in
Denver

   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   76
Less Successful Strategies (or
      those that failed)

Use of Hospital Records
   Race not identified or racial
   misclassification
   Records too old (2-3 years to update
   selected hospital records in CO)
   Urban AI community too mobile


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   77
What we learned from “Increasing
Mammogram Adherence in Minority
  Women” [NCI R25 CA 96540]
 Hospital programs are insufficient to
 serve as a sampling source

 >85% of the women identified as
 “American Indian” were already in our
 NACR NAWWA database
 Again: Very mobile population (50%
 changing phone numbers and addresses
 within 18 months)
   The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   78
  More Successful Strategies
Have members of the tribal or other
communities collaborate on the
interpretation of the data (i.e., avoid
SAIAN fiasco 1988-1993)

Allow for cultural modification of
standardized instruments (results are
not valid for many medically underserved
communities anyway) [holding all
variables constant may be less valid than
is usually assumed, e.g., “Pathways”]
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   79
  More Successful Strategies
Allow for culturally appropriate
variations of data collection
instruments (e.g., WHI and physical
activity items)

Allow for smaller or culturally specific
organizations to take responsibility for
data collection and management (e.g.,
NAWWA)


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   80
  More Successful Strategies

Dr. Charlie Key‟s training program for
data collectors within hospitals, clinics,
other facilities on how to ask, collect
and manage racially accurate data

Encourage healthcare facilities who
currently do not collect ethnicity or
racial data to do so


  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   81
  More Successful Strategies
CHIS cultural training program for
WESTAT (who conducted the
interviews)

Train tribal programs and Nations on
how to collect, manage, store accurate
data

Make estimates based on local tribal
representation and IHS data

  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   82
  More Successful Strategies
When NCHS refuses to correct their
data after CO improves its, make
certain their less accurate data do not
replace the revised CO database
(NOTE: KS and AI HIV/AIDS data
1994)

Allow for coding of tribal data to avoid
violation of HIPAA and tribal
ordinances

  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   83
    Estimating Cancer Screening
    and Risk Factor Rates among
         American Indians:
     California Health Interview
           Survey (CHIS)
                       Judith Swan, MHS
                       Nancy Breen, PhD
                 Linda Burhansstipanov, DrPH
                     Delight E. Satter, MPH
                     William W. Davis, PhD
                      Timothy McNeel, BA
                    C. Matthew Snipp, PhD

NOTE: paper will be published in AJPH December 2005   84
      Cancer among AIAN
Mortality rates have not declined
significantly.
Gap persists in self-reported health status
compared to non-Hispanic whites.
National health indicators remain among
the worst for all major race groups.
High level of use of preventive services
reported, but also higher levels of risk
factors such as smoking and obesity.
                                              85
               2001 CHIS
Telephone survey conducted Nov 2000 – Oct
2001.
Random digit dial sample of over 54,000 adults.
Over-sample of American Indians and Alaska
Natives.
The largest population-based cancer risk-factor
and screening data source on AIAN (based on
self-identified tribal nation definition, n = over
2,000).
                                                     86
     Race Definition Issues
With small populations, rates can vary
widely for rare events such as cancer.
Race and ethnicity are defined and
measured separately in the U.S.
Federal surveys use self-identification.
OMB Directive 15 allows for identification
of multiple races.

                                             87
NOTE: this strategy of how ethnicity and
CHIS 2001 Race/Ethnicity Data
race questions were asked can be
implemented in CO
  Respondents were first asked about
  Hispanic/Latino origin, then race group(s).
  Respondents reporting more than one
  race or ethnicity were asked if they
  particularly identified with one.
  Resulted in a rich data set with several
  options for tabulating AIAN.

                                             88
CHIS Race Definitions for AIAN

1. Any mention – respondents counted in
   each group mentioned.
  •   Results in replicate counts of persons
      across groups.
  •   Results in largest sample size.
  •   Results in most diverse grouping.



                                               89
CHIS Race Definitions for AIAN

2. AIAN only
  •   Includes persons who report only one race.
  •   Used by the 2000 Census.




                                                   90
CHIS Race Definitions for AIAN
3. Non-Hispanic AIAN
     California Department of Finance (DOF)
      uses for state population projections.
     Persons responding “yes” to the
      Hispanic/Latino origin question are tabulated
      as Latino and not categorized according to
      race.
     Results in the smallest sample size.


                                                  91
CHIS Race Definitions for AIAN
4. Most identified with AIAN
  •   UCLA Center for Health Policy Research.
  •   Persons tabulated in group with which they
      most identify (every respondent in a single
      category).
  •   Includes persons enrolled in recognized
      tribe, whether or not they most identified
      with AIAN.
  •   Used for AskCHIS online data query system.

                                                92

                   CHIS 2001
          AIAN Sample Sizes
Method                     AIAN Pop.(est)    CHIS
1. Any mention of AIAN      413,493         3,186

2. AIAN only                 210,937        1,155
      (Census/OMB)
3. Non-Latino AIAN            64,349         612
      (California DOF)
4. Most identified with AIAN 114,362        1,132
      (UCLA Center)

                                                93
             Methods
Compared cancer screening and risk
factors, sociodemographics, and access to
care variables.
Compared AIAN men and women to other
racial groups using definition #4.
Compared AIAN using the four tabulation
methods.

                                          94
              Weighting
Census counts from the 2000 Summary
File were used to derive statistical weights.
The RDD and list samples were combined
and weighted as a single sample.
The AIAN file weights matched the
California 2000 Census estimates for
“AIAN only” and “any mention” categories.
Estimates for other definitions may be
biased.
                                            95
           Conclusions
Racial misclassification of AIAN can result
from both under-counting and over-
counting.
The California DOF tabulation method is
an example of under-counting because it
excludes Latinos/Hispanics.




                                          96
           Conclusions

The “any mention” method is illustrative of
over-counting because it includes many
people who do not identify closely with
AIAN culture and society.




                                              97
        Conclusions cont’d.
Which definition to use?
   Not the CA DOF, which excludes Latinos.
   Not “any mention,” which includes those who
    do not identify strongly with AIAN.
   Not “most identified with,” which cannot be
    compared to AIAN reports based on Census,
    IHS, or tribal databases.
   Use the Census method of “AIAN alone,”
    which can be compared to most other
    studies.

                                                  98
     What’s Happening at Cherokee
        Nation Cancer Registry?
   Case Submission
   Follow-Up
   Data Analysis
   Updating Policies and Procedures
   Developing a Quality Control Plan
   Steps to expand coverage of Data Exchange
   Comprehensive Cancer Plan



                                                99
      What’s Happening at Cherokee
        Nation Cancer Registry?

   Submitting approximately 15 cases per week to New
    Mexico Cancer Registry.
   Follow-Up Rate ~90%
   Working with Chuck Wiggins and NMTR staff to
    examine CNCR data.
   Updating Policies and Procedures



                                                        100
        What’s Happening at Cherokee
          Nation Cancer Registry?
   Developing a Quality Control Plan
       Including Hastings Indian Medical Center and Claremore
        Indian Hospitals
       Cherokee Nation Quality Improvement Program


   Steps to expand coverage of Data Exchange
       Initiated new scopes of work for new Memoranda of
        Understanding
       Finalizing CNCR “Road Show” to present to area facilities


                                                                    101
        What’s Happening at Cherokee
          Nation Cancer Registry?
   Comprehensive Cancer Plan
       Draft completed February 2005

       Currently updating plan for implementation

       Great benefit from CNCR

       Final plan anticipated by August 2005




                                                     102
                    Partnerships
   New Mexico Tumor Registry
       Technical Assistance to correct mapping issue

       Simultaneous training for new CN Information
        Systems

       Technical Assistance for analysis



                                                        103
                   Partnerships
   Oklahoma Central Cancer Registry
       Regular Data Exchange

       Regular Meeting Participation

       Other Assistance as necessary




                                        104
 Data corrections [LB edit]
   Based on 4 sequential years of CN data:
     James W. Hampton, MD (Choctaw, Chickasaw) has
      been right all of these years!!
     correcting racial misclassification increased age-
      adjusted cancer mortality rates for CN by 48%, 47%,
      48% and 49%!!




                                                       105
               What’s Next?
   Focus on completing coverage area
   Evolving Capacity
   Continue Partnerships
   Sharing Experience
   Providing Assistance
   Graduate



                                        106
                     What’s Next?
   Focus on completing coverage area
       Ensure completeness of data

       Finalizing “Road Show” plans

       Scheduling presentations to area facilities to facilitate
        MOU



                                                               107
                  What’s Next?
   Evolving Capacity
     Certified Tumor Registrars (2)
     LinkPlus



   Continue Partnerships
     NMTR
     OCCR




                                       108
                    What’s Next?
   Sharing Experiences
     Presentations
     Manuscript Development



   Providing Assistance
       Emerging tribal interest in surveillance




                                                   109
                 What’s Next?
                  (Long Term)
   Graduate
       SEER Jr       SEER




                                 110
           Acknowledgements
   Kymberly Cravatt, MPH, Director
   Ann Wheeler, Registrar
   Donna Bush, Admin Asst
   Ruth Hummingbird, CCC Coordinator
   Gloria Grim, MD, Medical Director
   Chuck Wiggins, PhD, NMTR
   Janis Campbell, PhD, OCCR

                                        111
                     IHS / Tribal Data Linkage Project
      Special thanks to David Espey, MD, IHS National Epidemiology
                                  Program

        Activities:

      Link IHS data to cancer registries
      Link IHS data to death records
      CRC Screening in AK
      Canada US Indigenous Cancer
        Collaborative


112
      Registry linkages
 Link IHS patient registration data to
  cancer registries in 49 states/10
  metropolitan areas (NPCR and SEER)
 Ca Registry data sent to IHS
  (Albuquerque) where linked to IHS pt
  registration data using LinkPlus



113
      Registry linkages
 Linkage status captured in “IHS Link”
  variable
 Monograph of site specific manuscripts
  with state participation (? Cancer)




116
Regions used for AI/AN cancer incidence analysis
                 Cherokee Nation Cancer Registry
   Case Enumeration by Cancer Type: Males

          Type of Cancer                Percent
          Prostate                        22.6
          Lung                            23.2
          Colorectal                      12.4
          Bladder                         5.1
          Kidney                          4.8
          Leukemia                        4.1
          Pancreas                        2.5
          Esophagus                       2.4

           Cherokee Nation Data Linkage Project
special thanks to Chuck Wiggins, PhD, Director, New Mexico
                      Tumor Registry                         118
            Cherokee Nation Cancer Registry
Case Enumeration by Cancer Type: Females

     Type of Cancer                Percent
     Breast                          28.9
     Lung                            13.8
     Colorectal                      9.9
     Uterus                          4.9
     Cervix                          4.3
     Ovary                           4.4
     Kidney                          3.9




                                              119
                                             Cherokee Nation Cancer Registry
                             Incidence Rates: All Primary Types/Sites (Combined)

                             700

                             600

                             500
Incidence rate per 100,000




                             400

                             300

                             200

                             100

                               0
                                   Adair/Cherokee   NM Non Hispanic   NM Hispanic   NM American Indian
                                                        White


                                                                                                    120
                               Conclusions

Medically underserved communities
(e.g., AI) need accurate data to help
drive local programs

The efforts to improve data are
tedious and have multiple cultural and
technological challenges

Efforts will require personnel, time,
money and diligence
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   121
                               Conclusions

We need to be less concerned with
“standardization” and more concerned
with valid outcomes (i.e., are MUP
findings from „standardized‟
instruments or strategies really valid?)

We need long-term strategies for
maintaining the quality of corrected
databases

  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   122
                               Conclusions

We need collaborative partnerships
with medically underserved
organizations to address the data
issues in culturally acceptable manners

We need to “think outside the box” to
address the issues



  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   123
                               Conclusions

We need to collaborate with others
(e.g., Wiggins, Espey) who are
succeeding in improving the accuracy of
their databases
We need to used Community-Based
Participatory Research methodology
whenever feasible

We need to protect the privacy of
sensitive data (concepts of HIPAA)
  The Culture of Cancer Data - Burhansstipanov- NACR 303-838-9359; http://www.NatAmCancer.org   124

				
DOCUMENT INFO
Description: Correcting a Texas Birth Certificates document sample