Transcript Racial and Ethnic Data Issues for Epidemiologic Studies

Racial and Ethnic Data Issues
for Epidemiologic Studies
Olivia Carter-Pokras, Ph.D.
University of Maryland School of
Medicine
What is Epidemiology?



It is the science of public health or the study
of the distribution and determinants of
disease risk in human populations.
Epidemiologists study a diverse range of
health conditions as well as the impact that
various exposures have on the manifestation
of disease.
Epidemiologists come from a variety of
backgrounds and influence health from
academic, governmental, and private industry
positions.
How and Why do Epidemiologists
Use Race/Ethnicity?



Study disease and mortality patterns
and how they affect racial and ethnic
groups differently.
Examine how health risks and behaviors
are different between different
communities.
Use data to target resources where
they’ll have the most impact.
What is Race?
“Race is a social construct (in other words, a
social invention that changes as political,
economic and historical contexts change),
it has real consequences across a wide
range of social and economic institutions.”
SOURCE: American Sociological Association. Statement on the Importance
Of Collecting Data and Doing Social Scientific Research on Race. 2003.
What is Race?


a group, especially of people, with
particular similar physical
characteristics, who are considered as
belonging to the same type, or the fact
of belonging to such a group
a group of people who share the same
language, history, characteristics
SOURCE: Cambridge Advanced Learner's Dictionary
Medical Dictionary Definitions


The On-line Medical Dictionary: The descendants
of a common ancestor; a family, tribe, people, or
nation, believed or presumed to belong to the same
stock; a lineage; a breed
Dorland’s Medical Dictionary: 1. an ethnic stock,
or division of humankind; in a narrower sense, a
national or tribal stock; in a still narrower sense, a
genealogic line of descent; a class of persons of a
common lineage. In genetics, races are considered as
populations having different distributions of gene
frequencies. 2. a class or breed of animals; a group
of individuals having certain characteristics in
common, owing to a common inheritance; a
subspecies.
“…present-day inequalities between socalled "racial" groups are not
consequences of their biological
inheritance but products of historical and
contemporary social, economic,
educational, and political circumstances.”
American Anthropological Association. Statement on “Race”. 1998.
Question: How does the Census Bureau
define race and ethnicity?
Answer: Census Bureau complies with the Office of
Management and Budget's standards for
maintaining, collecting, and presenting data on race,
which were revised in October 1997. They generally
reflect a social definition of race recognized in this
country. They do not conform to any biological,
anthropological or genetic criteria.
SOURCE: Questions and Answers for Census 2000 Data on Race
March 14, 2001. Available at:
http://www.census.gov/Press-Release/www/2001/raceqandas.html
Debunking the Myths: the
Standards



There is a standard for racial and ethnic data
which is to be used by Federal agencies
“NIH Policy On Reporting Race And Ethnicity
Data: Subjects In Clinical Research” requires
the Federal standard for extramural research
The categories used by the Census are NOT
the Federal standards, but do comply with the
Federal standards
1997 Federal Standards (SelfIdentification)

Race






American Indian or Alaska Native
Asian
Black or African American
Native Hawaiian or Other Pacific Islander
White
Ethnicity


Hispanic or Latino
Not Hispanic or Latino
Census 2000 Question on
Ethnicity

Are you Spanish/Hispanic/Latino?





No, not Spanish/Hispanic/Latino
Yes, Mexican, Mexican Am, Chicano
Yes, Puerto Rican
Yes, Cuban
Yes, other Spanish/Hispanic/Latino—Print
group
Census 2000 Race Question

What is your race? Mark one or more races to indicate what
you consider yourself to be.
 White
 Black, African Am., or Negro
 American Indian or Alaska Native—Print name of enrolled or
principal tribe.
 Asian Indian
--Native Hawaiian
 Chinese
--Guamanian or Chamorro
 Filipino
--Samoan
 Japanese
--Other Pacific Islander—print race
 Korean
 Vietnamese
 Other Asian—Print race
 Some other race—Print race
1997 Federal Standards (May be
Used for Observer Identification)






American Indian or Alaska Native
Asian
Black or African American
Hispanic or Latino
Native Hawaiian or Other Pacific
Islander
White
Debunking the Myths: Legal
Issues



No general requirement to collect
racial/ethnic data
There is no Federal law which prohibits the
collection of racial/ethnic data
The vast majority of states do not have laws
which prohibit health insurers or health plans
from collecting racial/ethnic data at the time
of enrollment or eligibility determination
“Hispanics are asked to indicate their
origin in the question on Hispanic
origin, not in the question on race,
because in the federal statistical
system ethnic origin is considered to
be a separate concept from race.”
SOURCE: Bureau of the Census. Racial and Ethnic Classifications
Used in Census 2000 and Beyond. Available at:
http://www.census.gov/population/www/socdemo/race/racefactcb.html
Multiple Options for Data
Presentation




Combine all who report multiple race
Use a follow-up question on “best” or
“preferred“ to recode those who report
more than one race
Present data for most commonly
reported multiple race combinations
Ignore those who report more than one
race
Guidance on Aggregation and Allocation of Data on Race for Use in
Civil Rights Monitoring and Enforcement
1
2
3
4
5
6
7
8
9
10
11
12
13
American Indian or Alaska Native
Asian
Black or African American
Native Hawaiian or Other Pacific Islander
White
American Indian or Alaska Native and White
Asian and White
Black or African American and White
American Indian or Alaska Native and Black or African
American
> 1 percent: Fill in if applicable_____________________
> 1 percent: Fill in if applicable______________________
Balance of individuals reporting more than one race
Total
SOURCE: OMB BULLETIN NO. 00-02; March 9, 2000
Minimum Template for
Healthy People 2010
Race:
American Indian or Alaska Native
Asian or Pacific Islander
Asian
Native Hawaiian or Other Pacific Islander
Black or African American
White
Hispanic origin and race:
Hispanic or Latino
Not Hispanic or Latino
Black or African American
White
Minimum Presentation of Data on
Race: Washington State DOH








Total
AIAN
Asian
Black
NHOPI
Other White
More than one race
Race Not Reported
Percent Uninsured, 0-64 years,
California, 2000
30
White
25
Latino
20
Asian American
15
Native Hawaiian or
Other Pacific Islander
African American
10
5
0
% Uninsured
American Indian or
Alaska Native
Other and Multiple
Race
Lifetime Alcohol Drinking Status,
Adults >=18 years: 1997-98
Current
NonHispanic Asian or
Pacific Islander
NonHispanic Black
Former
Regular
NonHispanic White
Former
Infrequent
Hispanic
Lifetime
Abstainer
0
20
40
60
80
SOURCE: Advance Data No. 324, Revised April 18, 2002.
Perspective from Project RACE


“Will the federal agencies be able to "play with
the numbers" because of this?
A. Of course they will. Resources will be directed
whichever way they want by manipulating the racial
numbers. In a March 14 story in The New York
Times, officials at the Health and Human Services
Department said they don't know whether they
should plan to care for the 2.6 million people who
identified themselves exclusively as American Indians
or the 4.1 million people who said they were at least
part Native American. Government agencies will
misuse the multiracial numbers to their benefit. “
Use of Stand-Alone Multiracial
Category



Ohio, Illinois, Georgia, Indiana, Michigan,
Maryland have enacted legislation requiring
use of stand-alone multiracial category
North Carolina and Florida Boards of
Education have adopted the Multiracial
Classification by Administrative Mandate
Head Start's new Program Information Report
(PIR) software will include an "ethnicity"
category of "Biracial/Multiracial" and staff will
be instructed to "report the category of
ethnicity designated by the family."
Using Multiple Race Responses


If don’t need to bridge data, leave multiple racial
responses as separate categories; avoid reallocating
back to single racial categories.
If multiple responses could make an important
difference in measuring such disparities, it would be
important to use methods that more closely assign
people who report more than one race to the single
group they would otherwise report before the
multiple-response option became available. (Joint
Center for Political and Economic Studies)
% No Health Insurance for Bridge
Tabulation: NHIS 1993-5




White 13.4% (13.4%-13.5%)
Black 18.1% (18%)
AIAN 32.2% (26.7%-32.3%)
API 18.9% (18.2%-18.9%)
SOURCE: Office of Management and Budget. The Bridge Report: Tabulation
options for trend analysis. In: Provisional guidance on the implementation of the
1997 standards for Federal data on race and ethnicity. December 15, 2000.
Strategies for Users Who Need
to Bridge


Plurality method or one of the Fractional Allocation
methods provided the closest approximations to a
past distribution.
If interested in numerically small population and want
to maximize numbers for analysis, Smallest Group
method and Largest Group Other than White method
would yield larger counts for the category--this could
raise problems of misclassification of race for a
certain proportion of responses.
SOURCE: Sharon M. Lee, Using the New Racial Categories in the 2000
Census, Prepared for the Anne E. Casey Foundation. March 2001.
Jones CP, LaVeiwt TA, Lillie-Blanton M. Race in the
Epidemiologic Literature: An Examination of the
American Journal of Epidemiology, 1921-1990. Am J
Epidemiol. 1991 Nov 15;134(10):1079-84.



Proportion of papers containing a reference
to race rose steadily from 1975
Proportion of papers reporting inclusion of
“nonwhite” populations did not show a
parallel increase.
Exclusion of “nonwhite” subjects and
description of predominantly “white” study
populations increased.
Williams DR. The concept of race in Health Services
Research: 1966 to 1990. Health Serv Res. 1994
Aug;29(3):261-74.


Race/ethnicity is widely used in the health
services literature to stratify or adjust results
and to describe the sample or population of
the study.
Terms used for race are seldom defined and
race is frequently employed in a routine and
uncritical manner to represent ill-defined
social and cultural factors.
Bennett T, Bhopal R. US health journal editors' opinions
and policies on research in race, ethnicity, and health. J Natl
Med Assoc. 1998 Jul;90(7):401-8.




Twenty-nine editors of health journals with impact
factors of > or = 1 (based on citation ratings) were
sent a questionnaire including four key problems
identified in research literature and recommendations
from federal agencies; 23 (79%) responded.
Seven editors had relevant policies.
Two had read the federal directive on racial and
ethnic classification; one was aware of its current
review. Most perceived the four key problems as
uncommon.
Approximately 20% had discussed issues with coeditors, editorial boards, or reviewers. About 40%
saw further discussion as beneficial; four planned to
draft guidelines.
“…Identify the age, sex, and other important
characteristics of the subjects. Because the relevance
of such variables as age, sex, and ethnicity to the
object of research is not always clear, authors should
explicitly justify them when they are included in a study
report. The guiding principle should be clarity about
how and why a study was done in a particular way. ..
Authors should avoid terms such as "race," which lacks
precise biological meaning, and use alternative
descriptors such as "ethnicity" or "ethnic group"
instead. Authors should specify carefully what the
descriptors mean, and tell exactly how the data were
collected (for example, what terms were used in survey
forms, whether the data were self-reported or assigned
by others, etc.)..”
SOURCE: International Committee of Medical Journal Editors. Uniform
Requirements for Manuscripts Submitted to Biomedical Journals
Updated October 2001
NIH POLICY AND GUIDELINES ON THE
INCLUSION OF WOMEN AND MINORITIES AS
SUBJECTS IN CLINICAL RESEARCH


“…members of minority groups and their
subpopulations must be included in all NIH-funded
clinical research, unless a clear and compelling
rationale and justification establishes to the
satisfaction of the relevant Institute/Center Director
that inclusion is inappropriate with respect to the
health of the subjects or the purpose of the
research…Cost is not an acceptable reason for
exclusion except when the study would duplicate
data from other sources.”
Clinical research includes: Patient-oriented research,
Epidemiologic and behavioral studies, Outcomes
research and health services research
SOURCE: Federal Register, March 28, 1994, 59: 14508-14513; Revised October 2001
NIH POLICY ON REPORTING RACE AND ETHNICITY
DATA: SUBJECTS IN CLINICAL RESEARCH: August 8,
2001



1997 OMB Standards have been adopted for
“maintaining, collecting, and presenting data
on race and ethnicity for all grant
applications, contract and intramural
proposals and for all active research grants,
cooperative agreements, contract and
intramural projects.”
“Collection of this information and use of
these categories is required for research that
meets the NIH definition of clinical research”
Effective January 10, 2002
Public Health Service Grants
(PHS 398/2590 Rev. 5/01)


Targeted/Planned Enrollment Table does
not include a More than One Race or
Unknown or Not Reported categories
but the Inclusion Enrollment Report
Table does
Principal Investigator Personal Data
Form allows individuals to report more
than one race
Acceptable Minority Inclusion:
NIH Review Criteria


Included in scientifically appropriate numbers and
recruitment/retention has been realistically addressed
Some or all minority groups are excluded because:
 Inclusion would be inappropriate with respect to their
health
 The research question addressed is relevant to only one
racial or ethnic group
 Evidence from prior research strongly demonstrates no
differences between racial or ethnic groups on the outcome
variables
 A single minority group study is proposed to fill a research
gap
 Sufficient data already exists with regard to the outcome of
comparable studies in the excluded racial or ethic groups
and duplication is not needed in this study
Acceptable--II

Some minority groups are excluded or poorly
represented because geographical location of
study has only limited numbers of these minority
groups who would be eligible for the study AND
the investigator has satisfactorily addressed this
issue in terms of:
 size of the study
 relevant characteristics of the disease, disorder
or condition
 feasibility of making a collaboration or
consortium or other arrangements to include
representation
Acceptable--III

Racial or ethnic origin of specimens of
existing datasets cannot be accurately
determined AND this does not
compromise the scientific objectives of
the research.
Unacceptable




Minority representation fails to conform to
NIH policy guidance
Insufficient information is provided
Application does not adequately justify limited
representation of minority groups or
subgroups
The application does not adequately address
recruitment/retention of some or all minority
groups or subgroups.
Other Inclusion Policies


CDC and ATSDR Policy on the Inclusion
of Women and Racial and Ethnic
Minorities in Externally Awarded
Research (September 15, 1995)
AHRQ Policy On The Inclusion Of
Priority Populations In Research
(February 27, 2003)
Kaplan and Bennett Guidelines for
Use of Race and Ethnicity in
Biomedical Publication




Give reason for use of race/ethnicity
Describe the way individuals were
assigned racial/ethnic categories and
whether list of fixed categories used
Do not use race/ethnicity as proxy for
genetic variation
Distinguish between race/ethnicity as
risk factor and risk marker
SOURCE: JAMA, May 28, 2003, No. 20, pg 2710-2714
Kaplan and Bennett
Guidelines--II



Consider all conceptually relevant
factors in interpretation of racial/ethnic
differences
Make every effort to adjust for
conceptually relevant measures of SES
or social class when comparing
racial/ethnic groups
Use terminology that is not stigmatizing
Primer on Measuring
Disparities in Health



Being prepared by NCHS. Contact Ken
Keppel: [email protected]
Discusses issues that arise in measuring
differences among rates and in
measuring changes in disparity over
time
Discusses strengths and limitations of
specific statistics based on the purpose
of the analysis and the number of
groups to be compared
Disparity Dictionary Definitions




Most dictionary definitions define disparity as
inequality; difference in age, rank, condition or
excellence; or dissimilitude
Inequality is defined as condition of being unequal,
or lack of equality as of opportunity, treatment or
status
Inequity signifies an ethical judgment: an instance of
unjustness or unfairness.
More recent dictionaries include this ethical judgment
in definitions of disparity: a lack of equality and
similarity, esp. in a way that is not fair.
Use of term Disparity in U.S.



Although the term “disparity” is widely used
in public health in the United States, there is
a difference of opinion about what is meant
by disparity.
These differing opinions are based on
dictionary definitions as well as personal
beliefs of what is avoidable & what is unfair.
Confusion can arise from different operational
definitions adopted by various health
organizations.
Use of term Disparity (Cont.)



What should be agreed upon is that a disparity acts
as a signpost—indicating that something is wrong.
To make progress in reducing and ultimately
eliminating disparities in health, policy makers should
go beyond discussion of inequality and consider what
is inequitable.
Research priorities should focus on what we do not
know regarding how to avoid a given disparity, what
determinants are amenable to intervention, and how
to make changes based on what we do know.
THE CONFLICT: RACIAL PRIVACY (5/2002)
“Ward Connerly, famous for his opposition of
affirmative action, is now championing a ballot initiative
that would strip all references to race and ethnicity
from government forms with a few exceptions. He
argues that the government has no need to collect such
information on race. His opponents believe the measure
would make it impossible to fight discrimination. “
SOURCE: California Connected (a weekly television news program)
Classification by Race, Ethnicity, Color and
National Origin Initiative (CRECNO)




Often referred to as the “Racial Privacy
Initiative”
Will be on California’s March 2004 ballot to be
effective January 1, 2005
Will likely impact health research, data and
statistics—the exception for “Otherwise lawful
classification of medical research subjects and
patients” is open to legal interpretation in
future
Now has support of the California Republican
Party
Concerned about this
Initiative?




Need to hear from providers and advocates regarding
successful programs that use race/ethnicity data to
address needs of particular populations.
Need to hear from researchers regarding the various
types of data that would be impacted by the
initiative.
Donations as well as letters of support are also
needed
Contact the Coalition for an Informed California:
http://www.informedcalifornia.org/index.shtml
1611 Telegraph Ave, Suite 317
Oakland, CA 94612
Phone: 510 452-2728
Fax: 510 452-3552
NAS Study of DHHS Collection of
Race & Ethnicity Data: 2001-3



Examine the adequacy of race and ethnicity
data collected or used by the Department of
Health and Human Services program.
Will review current policies and practices,
examine data requirements and limitations,
and suggest improved methods.
To hear more about their findings:



American Public Health Association presentation
Future reports
Contact: Dr. Michele Ver Ploeg, Study Director,
202-334-3481, [email protected]