August 23, 2004

The Honorable Tommy G. Thompson
Secretary
U.S. Department of Health and Human Services
200 Independence Avenue, S.W
Washington, D.C 20201

Dear Secretary Thompson:

As part of its charge, the National Committee on Vital and Health Statistics concerns itself with HHS population-based health data issues. The Committee is particularly interested in ensuring that adequate, timely and relevant data are collected and available for use by health planners, researchers, policy makers and communities in order to monitor and assess health status, health behaviors, care and access to treatment in racial and ethnic minorities and other vulnerable populations.

The present commentary from the National Committee on Vital and Health Statistics examines the limitations of health data on racial and ethnic groups, as currently collected and used. It draws on hearing testimony, previous recommendations, and other sources. On the basis of this analysis, the commentary recommends a set of actions to remedy problems and fill information gaps. The Committee makes these recommendations to assist the Department in achieving its mission of eliminating health disparities by enhancing the collection, use and dissemination of health and health care data on racial and ethnic populations.

We appreciate the opportunity to submit this commentary and would welcome the opportunity to discuss it further with the HHS Data Council.

Sincerely,

/S/

John Lumpkin, M.D., M.P.H.
Chairman, National Committee on Vital and Health Statistics

cc: HHS Data Council Co-Chairs

Enclosure


Recommendations on the Nation’s Data
For Measuring and Eliminating Health Disparities
Associated with Race, Ethnicity, and Socioeconomic Position

A Commentary by the National Committee on Vital and Health Statistics

Introduction

Federal efforts to eliminate health disparities and improve the health of all Americans depend on the ability to target those at high risk and with great need. This requires detailed information on the diverse array of population groups that comprise the United States. Recently, landmark reports have drawn the nations attention to the range of disparities in the health status and health care of racial and ethnic minority populations, and also to the need for better data on minority groups.

The National Committee on Vital and Health Statistics[1], in its advisory role to the Secretary of Health and Human Services on health statistics, has a long history of proposing initiatives to assist the Department in its efforts to eliminate health disparities. The Department of Health and Human Services (HHS) also has a long history of developing and implementing important initiatives to achieve the goal of a healthier America. However, the Institute of Medicines Report on the Collection of Data on Race and Ethnicity[2] and the Departments National Healthcare Disparities Report[3] have created a new sense of urgency about the need for guidance, coordination, monitoring, funding, research and partnership with the private healthcare sector to achieve the goal of healthier Americans, especially with regard to specific racial and ethnic minority populations. These two reports highlight some of the overwhelming disparities in the health status, health care and health information needs of particular groups.

The present commentary from the National Committee on Vital and Health Statistics examines the limitations of health data on racial and ethnic groups, as currently collected and used. It draws on hearing testimony[4], previous recommendations[5], and other sources. Some challenges stem from the Office of Management and Budgets 1997 revisions to the standards on race/ethnicity classification On the basis of this analysis, the commentary recommends a set of actions to remedy problems and fill information gaps. The Committee makes these recommendations to assist the Department in achieving its mission of eliminating health disparities by enhancing the collection, use and dissemination of health and health care data on racial and ethnic populations.

COMMENTARY AND RECOMMENDATIONS

Measuring and eliminating health disparities

Compelling evidence exists that differences in health status, access to care, and the provision of physical and mental health services are significantly related to race, ethnicity and socioeconomic position[6]. Despite advances in medicine and increased emphasis on disease prevention and health promotion in the United States , the country s racial and ethnic minority populations generally have higher rates of illness and disease and poorer quality of care than Whites. For example, life expectancy for Whites in the United States in 2000 was 77 years, while for Blacks it was six years less, at 71 years[7].

The national goal of eliminating health disparities among racial and ethnic groups has come increasingly to the fore in recent years, along with growing recognition that the nation s health must be improved population group by population group, with greatest attention to those at particular risk. The Institute of Medicine (IOM) has made a strong statement about disparities in its landmark report, Unequal Treatment[8], and the recent National Research Council Panel on HHS Collection of Race and Ethnicity Data[9] furthers public attention to these issues with its new report. The issues and goals are at the heart of Healthy People 2010, reinforced by the recent National Healthcare Disparities Report, and addressed in multiple other reports by IOM and others[10]. In addition, as the first of five steps to effectively reduce a communitys cancer burden, the National Cancer Institute and the Centers for Disease Control and Prevention recommend reviewing data to identify populations at greatest risk and the extent of health disparities between populations.[11]

To achieve these goals, HHS must be able to target interventions and monitor progress for those at high risk and with special need. The targeting and monitoring, in turn, depend on having estimates of the health status and knowledge of the health needs of the many population groups and subgroups that comprise the U.S. For each racial and ethnic group and subgroup, data are needed not only on health factors including health status, health care, health care expenditures, access to care and experiences in the health care system but also on contextual factors such as socioeconomic position and the conditions in which people live. The sources of these data include national population-based surveys (the major focus of this commentary) as well as state data, targeted surveys, and administrative data.

The problem is that the very groups at particular risk for ill health and poor health care are those about whom health statistics and contextual data are most deficient. National population-based surveys are essential sources of estimates on the health of the United States population. They are used for general Black-White comparisons and (to a lesser degree) comparisons among Blacks, Latinos, and Whites. However, such surveys, at least as currently designed, are not an effective source of information about the health of American Indians or Alaska Natives, Asian and Native Hawaiian or other Pacific Islanders, or other population groups that are concentrated in a specific geographic area. It is significant that Healthy People 2010, which has an overarching goal of eliminating health disparities, lacks baseline data to quantify specific improvements for some subpopulations in many areas.

A Changing Landscape

Gaps and deficiencies such as these are not easily remedied, and those trying to collect and use data on American racial and ethnic groups face a host of challenges. For example, the U.S. is an increasingly multiracial society; the number of subgroups is daunting; and racial and ethnic self-reported identity is affected by who is collecting the data, the purpose for which the data are being collected, and other characteristics of the respondents that can change over time. Moreover, the confidentiality of personal information can be compromised when group size is small, thus limiting data use as well as raising privacy concerns among those supplying the data; and some institutions cite concerns about legality or acceptability to clients as a reason for not collecting data on clients race and ethnicity. In addition, race is a social construct that is not well defined or understood.

In 1977, the Office and Management and Budget (OMB) required a common language in Federal collection and use of data on race and ethnicity, with Statistical Policy Directive #15[12]. In October 1997, taking effect in January 2003[13], OMB issued a new method for accommodating changing population makeup and identities and providing categories for data on additional population subgroups. Specifically, the new Guidance provides five categories for data on race: American Indian or Alaska Native; Asian; Black/African American; Native Hawaiian or Other Pacific Islander; and White. It also provides two categories for data on ethnicity: Hispanic or Latino and Not Hispanic or Latino. In addition, for the first time, significantly, the revision offers respondents the option of selecting more than one racial designation. While the new OMB classification opens the way to greater detail in racial and ethnic data, it also has created complex challenges in data collection, tabulation and analysis for those charged with monitoring and improving population health. These are discussed below.

NCVHS Contributions

The National Committee on Vital and Health Statistics, in its advisory role, has completed many important initiatives to improve population health data, including a number of activities on behalf of racial and ethnic minorities. The NCVHS Subcommittee on Populations:PersonName> focuses on health data concerning the U.S. population generally, as well as data about specific vulnerable groups within the population that are disadvantaged by virtue of their special health needs, economic status, race and ethnicity, disability, age, or area of residence. The present commentary summarizes recent work on race and ethnicity data by the Subcommittee.

While the new OMB classification and the changes and questions it has engendered are one important focus of concern, the Subcommittee finds it useful to look at classification-related issues in the context of the broader topic of the information needed to monitor and eliminate population-level health disparities. This commentary recommends, and underscores the urgency for, a range of actions by the Federal government to create the conditions for the collection and use of reliable and complete data on racial and ethnic minorities. It draws on past NCVHS recommendations, the testimony at a February 2002 hearing convened by the Subcommittee, and the recent National Research Council report. It also builds on data quality improvement efforts already underway at NCHS, NIH, the Department s Data Council and elsewhere. The ultimate goal is ensuring that the Department can continue to play its key role in helping to eliminate health disparities within the population.

Specific Data Issues

The sources noted above highlight a number of data issues that must be resolved. As seen in the recommendations below, in many cases their resolution must begin with accelerated research, Federal guidance, and coordination both within the Department and with the private sector. Some issues pertain to data collection and others to tabulation, analysis, and interpretation. Sample design is one of the first challenges, partly because of the difficulty of collecting and reporting numbers from small racial and ethnic groups sufficient for analysis but without violating confidentiality protections. Other data collection issues include the need for adequate subgroup data as well as questions about primary race, given the options offered by the revised classification system. There also is a need for surveys in languages other than English and Spanish.

The greatest impact of the new OMB guidance is on data usage. In particular, allowing persons to report more than one race presents a host of challenges. These and other changes impose a new burden on states, at a time of reduced or flat budgets, forcing them to retool their data systems and bridge old and new statistics. Data collectors and users alike are calling for Federal guidance on multiple questions including how to present multi-race data, what Census denominators to use to calculate rates, and how and when to collect data on subgroups. To permit trend analyses, greater guidance is needed in methodology to bridge or compare data collected under the new standards with data collected under the old ones. Small analytic sizes are another challenge.

An additional category of data issues concerns the types of information besides race and ethnicity that are needed to understand, monitor and eliminate health disparities. For example, topics cited on which additional information is needed are socioeconomic position, geographic location, age, primary language, other factors that contribute to health status, and health care expenditures.

Cross-Cutting Policy and Capacity Issues

Much of the testimony received by the Subcommittee about data on race and ethnicity suggested that these data collection and use challenges are not insoluble; rather, they require a centralized approach to solving them. The Department has for many years expanded and improved the collection of data on race, ethnicity and socioeconomic position. Major contributions include basic techniques for bridging across classifications and for measuring the extent of disparities. However, work in this area needs to be expanded and accelerated, and at times conducted in partnership with other Federal agencies such as the Bureau of the Census, the Office of Management and Budget, and the Departments of Labor and Education. Enhancing data on healthcare would benefit, in particular, from partnerships with the private sector in the collection of data on racial and ethnic populations. All of this calls for a higher federal investment in research, guidance and coordination, with appropriate levels of funding.

The NCVHS hearings also identified important infrastructure capacity issues. They include the need for training and diversification of the survey and research workforces as well as the need to provide researchers, policy makers and the American public in general with timely data on the health and healthcare of racial and ethnic minorities.

Overarching Recommendations

Based on its findings, the National Committee recommends the following actions to improve and augment the current data available for measuring, tracking and ultimately eliminating health disparities for racial and ethnic minority populations.

  1. HHS should extend and intensify its current leadership role with other Departments, OMB, the Census Bureau, and private and academic organizations to promote and undertake methodological research associated with race, ethnicity, and socioeconomic position measurements in surveys, research and census data, in order to improve the health and health care data collected on racial and ethnic minority populations and subpopulations.This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendations 4-10 and 6-2.
  2. HHS should accelerate implementation of its multiple-strategy approach to data development on racial and ethnic minority populations and subpopulations, consistent with recommendations that appear in Improving the Collection and Use of Racial and Ethnic Data in HHS. This will require additional dedication of appropriate financial and staffing resources.This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendations 4-2 and 4-3.
  3. HHS is urged to develop consistent strategies and mechanisms for the broad dissemination of data on racial and ethnic minorities, including data on socioeconomic position, that result from Departmental data collection efforts, when targeted to specific racial and ethnic minority populations.This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendations 4-9 and 4-10.

Specific Strategies

The following areas represent strategies aimed at increasing the quality and quantity of data on racial and ethnic populations and subgroups and should be regarded as high priorities by the Secretary of Health and Human Services:

  1. Improve coordination of racial and ethnic group data collection and use, through partnership with entities responsible for data collection and through the issuance of consistent data requirements.This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendations 4-6 and 5-2.
  2. In consultation with the Office of Management and Budget, provide guidance on the following critical technical and methodological areas:a) Methodologies for bridging old and new data

    b) Analysis and interpretation of primary and multiple race categories

    This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendation 4-9

  3. Strengthen the capacities of the health statistics enterprise through improvements in access to Departmental data on racial and ethnic minority subpopulations and through the dissemination of research findings on racial and ethnic minority groups.This recommendation is consistent with National Research Council Panel on HHS Collection of Race and Ethnicity Data, Recommendation 5-2
  4. Provide guidance on critical technical and methodological issues, including:a) Improvements to the quality of race and ethnicity data in vital statistics records

    b) Oversampling in national surveys, where feasible

    c) Targeted studies for data development on subpopulations, when oversampling is not feasible or cost effective

    d) Methodology to improve the quantity, quality, and participation rates in racial and ethnic subgroup data collection

    e) Use of Census denominators for rate calculation

    f) Mechanisms to augment existing samples to target specific race or ethnic groups nationally or in subnational areas with adequate privacy protection

    g) The cultural validity of data collection tools

    h) Linking and matching across surveys and administrative data sources


[1] The National Committee on Vital and Health Statistics (NCVHS), a statutory Federal Advisory Committee established in 1949, advises the Secretary of Health and Human Services and Congress on the health information and data issues underlying Federal policy. This commentary was prepared by the NCVHS Subcommittee on Populations:PersonName>.

[2] Eliminating Health Disparities: Measurement and Data Needs (2004), Panel on DHHS Collection of Race and Ethnicity Data, Michele Ver Ploeg and Edward Perrin, Editors, Committee on National Statistics, National Research Council of the National Academies.

[3] National Healthcare Disparities Report, U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality, Rockville, MD., February 2004.

[4] Notably, a February 2002 hearing on the data needed to measure and eliminate health disparities, and on impact of the 1997 OMB Guidance on the Classification of Race and Ethnicity.

[5] Notably, a March 14 2001 letter from NCVHS to Katherine Wallman, Chief Statistician of the Office of Management and Budget. Improving the Collection and Use of Racial and Ethnic Data in HHSJoint report of the HHS Data Council Working Group on Racial and Ethnic Data and The Data Work Group of the HHS Initiative to Eliminate Racial and Ethnic Disparities in Health December, 1999.Shaping a Health Statistics Vision for the 21st Century, Final Report, Department of Health and Human Services Data Council, Centers for Disease Control and Prevention, National Centers for Health Statistics, National Committee on Vital and Health Statistics, November 2002.

[6] Socioeconomic status has traditionally been measured by education and income. However, more recent research has suggested that a broader view encompassing additional characteristics, such as current wealth, and historical contextual variables, such as socioeconomic conditions during childhood, are important for health outcomes. This more expanded view is referred to as socioeconomic position (SEP). Measuring SEP is a challenge as there is no consistent agreement about which factors should be included in SEP or what time period during childhood is most important. Some of the factors may vary depending upon the specific outcomes of interest or the populations being studied.

[7] http://www.cdc.gov/nchs/fastats/pdf/nvsr51_03t11.pdf

[8] Institute of Medicine (2001), Unequal Treatment: Confronting Racial and Ethnic Disparities in Healthcare. National Academies Press.

[9] Eliminating Health Disparities: Measurement and Data Needs, [2004] Panel on HHS Collection of Race and Ethnicity Data, Michele Ver Ploeg and Edward. Perrin, Editor, Committee on National Statistics Division of Behavioral and Social Sciences and Education. National Research Council of the National Academies. Washington DC.

[10] Tracking Health Plan Enrollees’ Race and Ethnicity. June 2004; www.healthypeople.gov; Eliminating Health Disparities: Measurement and Data Needs, Panel on HHS Collection of Race and Ethnicity Data, Michele Ver Ploeg and Edward. Perrin, Editor, Committee on National Statistics Division of Behavioral and Social Sciences and Education. National Research Council of the National Academies. Washington DC.

[11] http://cancercontrolplanet.cancer.gov/index.html

[12] Office of Management and Budget, Statistical Policy Directive No. 15: Race and Ethnic Standards for Federal Statistics and Administrative Reporting. May 12, 1977.

[13] Office of Management and Budget Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity. Federal Register Notice. October 30, 1997;62(210)