American Indian / Alaska Native (AI/AN) Cancer Data Misclassification and Regional Differences in Incidence and Mortality

1

• American Indian / Alaska Native (AI/AN) Cancer
  Data Misclassification and Regional Differences
  in Incidence and Mortality

Linda Burhansstipanov, DrPH (Cherokee Nation of
Oklahoma) Brenda Seals, PhD (Eastern Band
Cherokee) Delight Satter, MPH (Umpqua/ Klickitat
of the Confederated Tribes of Grand Ronde
) Native American Cancer Research 393 South
Harlan Street, suite 125 Lakewood, CO
80226-3571 303-325-3377 fax
303-495-3040 http//www.NatAmCancer.org
Native American Cancer Research ARS polling
question
1
2
Pre-Training Knowledge2 Items
3
What is the most significant factor contributing
to racial misclassification for AIAN data?

1. AIAN is a response category for too many
   healthcare facilities
2. Tribal enrollment blood quantum changes
3. Lack of tribal identification
4. Use of Spanish surnames
5. Dont know / not sure

d1_misclas
0 / 250
4
What region of Indian Country has the highest
deaths for Native women from breast cancer?

1. Alaska
2. East
3. Northern Plains
4. Southern Plains
5. Southwest
6. Pacific Coast
7. Dont know / Not sure

b1_BC_NDN
0 / 175
5
Workshop Objectives By the end of the session,
the participant will be able to

1. Identify common reasons for data
   misclassification among AIANs.
2. Delineate cancer incidence and mortality
   differences among tribal Nations and AIANs living
   is diverse geographic regions of the US.

6
Identify common reasons for data
misclassification among AIANsNote Results in
under-counting of AIAN health / cancer issues
that subsequently results in under-funding of
AIAN programs
7
Cancer
8
Statistics are merely aggregations of numbers
with the tears wiped away.
Dr. Irving Selikoff
9
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 LB

10
Without looking at your handout, which database
and its respective formulas for identifying race
is the most accurate for AIANs?

1. Indian Health Service
2. CDCs National Center for Health Statistics
3. US Department of Finance
4. Tribal Voters Registration
5. Dont know / Not sure

accurate database
0 / 175
11
Most Common Reasons for Racial Misclassification
In some areas of the country, AIANS are
statistically significantly under-counted because
of Spanish surnames. In Nebraska
misclassification may be based on skin color,
surnames, and identity issues

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

12
Most Common Reasons for Racial Misclassification
1.Imprecise and inconsistent definitions of AI
2.Changing self-identification 3.Tribe formerly
unrecognized
13
Most Common Reasons for Racial Misclassification
4 Tribal enrollment blood ordinances changing
5. Tribal enrollment ordinances re Paternal or
Maternal lineage
14
First documented misclassification of American
Indians
Thank you Chuck Wiggins, PhD, UNM SEER!
15
Every time an elder dies ...
a library burns Patricia Cochran, MS, Inupiaq,
Executive Director Alaska Native Science
Commission
Picture from Tinka Duran
16
Delineate cancer incidence and mortality
differences among tribal Nations and AIANs living
is diverse geographic regions of the US.
17
(No Transcript)
18
AIAN Cancer Incidence Data (finally) THANK YOU!

• CHSDA Counties IHS Contract Health Service
  Deliver Areas
• NHW Non-Hispanic White

19
Incidence Rates (1999-2004)
US US So. Plains So. Plains No Plains No Plains
AIAN NHW AIAN NHW AIAN NHW
All 368.4 475.9 492.6 461.2 538.1 464.8
Breast 85.3 134.4 115.7 129.6 115.9 130.3
Cervix 9.4 7.4 14.1 9.1 12.5 7.4
Colon 46.3 50.8 60.2 51.8 72.5 52.3
Kidney 17.9 11.8 21.3 12.4 23.7 12.1
Lung 48.5 58.6 87.1 82.2 104.3 65.5
Prostate 105.4 154.4 156.7 146.5 174.5 162.2
What is the racial group that has the highest
incidence of breast cancer? How about for
prostate cancer?
Source Cancer registries in Centers for Disease
Control and Prevention's (CDC) National Program
of Cancer Registries (NPCR) and National Cancer
Institute's (NCI) Surveillance, Epidemiology, and
End Results Program (SEER) D. Espey, Monograph
Group.
20
2007 Annual Report to the Nation

• Cancer incidence among AI/ANs, though generally
  lower than for the Non-Hispanic White (NHW)
  population, is rising rapidly
• In 3 of the twelve regions of the Indian Health
  Service, cancer incidence rates are equal to that
  of the dominant white population
• Alaska
• Northern Plains (IL, IN, IA, MI, MN, MT, NE, ND,
  SD, WI, WY)
• Southern Plains (OK, TX, KS)

21
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Current Smoker
• US
• 22.8 NHW
• 31 AIAN
• Southern Plains Region
• 33.2
• Northern Plains Region
• 40.1

Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
22
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Ever told have diabetes
• US
• 6.0 NHW
• 12.4 AIAN
• Southern Plains
• 12.6
• No. Plains
• 13.0

What is the relationship between diabetes and
cancer?
Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
23
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Obese (CDC defn. BMI gt 30.0 kg/m2)
• US
• 20.8 NHW
• 29.0 AIAN
• Southern Plains
• 30.9
• Northern Plains
• 31.0

What is the relationship between obesity and
cancer?
Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
24
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Health care coverage
• US
• 88.3 NHW
• 75.6 AIAN
• Southern Plains
• 71.8
• Northern Plains
• 77.2

The BRFSS data over-estimates insurance for
AIANs Inquiring Minds want to know, Why?
Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
25
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Pap Screening with past 3 years, All women 18
  (w/o hysterectomy)
• US
• 84 NHW
• 78 AIAN
• S. Plains 74
• N. Plains 80.4

What was the Pap percentage prior to the
screening guidelines changing?
Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
26
IHS Health Risk Factor Data (BRFSS, 2000-2006)

• Mammogram within past 2 years among women aged
  40
• US
• 76 NHW
• 69.4 AIAN
• S. Plains 66.4
• N. Plains 67.9

Steele CB, Cardinez CJ, Richardson LC, Tom-Orme
L, Shaw K. Surveillance for Health Behaviors of
American Indians and Alaska NativesFindings from
the Behavioral Risk Factor Surveillance System,
2000-2006, CANCER 2008, in press
27
IHS Cancer Mortality

• Note cancers other than breast, cervix, colon,
  lung and prostate
• This is why AIAN comprehensive cancer plans
  cannot be limited to only those sites!

28
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
29
IHS Mortality Data (best of any federal database,
but still has limitations)

• Note the of AIANs summarized in report is
  extreme overestimate
• 64 live in urban areas and those data are not
  included in the reports
• limited to AIANs WHO USE IHS CLINICAL FACILITIES
  WITH PREVIOUS 3 YEARS
• Note the horizontal scale fluctuates for each
  cancer site and can be misleading

30
Omissions

• Note both childhood and adult leukemia
  increasing among AIAN
• Childhood under-reported due to children not
  admitted to a Childrens Hospital and symptoms
  misdiagnosed as diabetes
• Brain cancer increasing among all races

31
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
32
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 1999-2003, IHS, Rockville,
MD 2008
33
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
34
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
35
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
36
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
37
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
38
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
39
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
40
Haverkamp et al, Cancer Mortality among AIAN
Regional Differences, 199-2003, IHS, Rockville,
MD 2008
41
What is the most significant factor contributing
to racial misclassification for AIAN data?

1. AIAN is a response category for too many
   healthcare facilities
2. Tribal enrollment blood quantum changes
3. Lack of tribal identification
4. Use of Spanish surnames
5. Dont know / not sure

d2_misclas
0 / 250
42
What region of Indian Country has the highest
deaths for Native women from breast cancer?

1. Alaska
2. East
3. Northern Plains
4. Southern Plains
5. Southwest
6. Pacific Coast
7. Dont know / Not sure

b2_BC_NDN
0 / 175