Improving the Collection and Use of Racial and Ethnic Data in HHS. Appendix C Summary of Racial and Ethnic Data Concerns Related to Healthy People 2010


To strengthen data systems used to provide baseline and monitoring data for Healthy People 2010, the DHHS held five regional meetings throughout the country in fall 1998 to solicit public comment. HHS also engaged in discussions with data users and data system managers both internally and externally to the Department, regarding ways to present the data, set targets for the year 2010, and eliminate gaps in data collection for racial and ethnic minority populations. During the regional meetings, participants recommended ways to improve racial and ethnic data for Healthy People 2010. The statements, questions, and recommendations provided below are culled directly from written and verbal comments delivered at the various meetings.

Need to be responsive to changes that are occurring

  1. Changes to the International Classification of Diseases codes for mortality (ICD-10) and morbidity. Compared to ICD-9, it will see an increase in cardiovascular disease and cancer mortality, and decrease in chronic liver disease and cirrhosis, and pneumonia and influenza mortality. Asked whether would re-calculate by cause.
  2. Changes in classification of race and ethnicity by Federal agencies (October 1997 revision of OMB Directive No.15). In addition to changes in terminology, the recommended order of questions on race and ethnicity, encouraging the two question format, creating new Native Hawaiians and Pacific Islander category, are allowing persons to identify with more than one race. How will Federal government and others use race and ethnicity data for those who indicate more than one race? Asked what happened to non-Hispanic whites and blacks. NCHS has been asked to tabulate natality data by race of the child. What about Middle Easterners?
  3. Change in standard for age adjustment. Why not compare age-specific rates by race and ethnicity? Suggest adding a note stating that age-adjusted rates and numbers are based on year 2000 standard population.

Overarching data needs identified

  1. Funding for data and link to resources. Hard to garner funds for data collection at local level with change of program administration down to local level. Need to ask decisionmakers what they want to know, need to get buy-in from policy people.
  2. Need for national and State monitoring and surveillance. How does Federal government plan to provide for State and local estimates? Need information on community basis.
  3. Need for more detailed studies of relative importance of underlying factors. How do we examine all of these underlying factors together? Need discussion of the underlying factors. Need to address underlying factors in Healthy People 2010 objectives; high school education not sufficient (e.g., poverty).
  4. Need for better evaluation of programs funded by Federal government.

Other data issues

  1. How are targets set? In general, support single target, but there was at least one question of using the better than the best approach (unachievable).
  2. Patient satisfaction data needed. There are not many tracking systems available to States and perhaps national level for health services use and access. Concern about Government Performance and Results Act (GPRA). GPRA measures are outcome measures while process is easier to measure at the local level. Problematic to address Healthy People 2000 objectives in grant application.
  3. SES and disability status should be used to breakout objective data.
  4. Rural area data?
  5. How do we train States and others on these data issues? How do we plan to inform people of these changes?
  6. Aggregate vs. detail
  7. Time to make changes (eight years), since can't get people to code. May not feel comfortable using data collected by other agency.
  8. Important to find "victories" in the data and be able to say something good has happened.
  9. Healthy People 2000 contained a chapter on Surveillance and Data Systems. For Healthy People 2010, only a few objectives addressed data within the Public Health Infrastructure chapter. Considering the importance of data for setting baselines and monitoring, especially for low income, racial/ethnic, and disabled populations, it is imperative that surveillance and data be given higher visibility in Healthy People 2010. Health information and data are core public health functions and should be emphasized in Healthy People 2010 (California Conference of Local Health Officers).

Recommendations and comments submitted during the formal public comment period in fall 1998

  1. For the national health objectives to be used at the State and local levels, data systems need to be identified which allow States and local communities to track progress toward the targets (CSTE). If State-level data are not available, then the objective should be noted as "developmental." Preference should be given to objectives which can be analyzed at State and local levels.
  2. Federal agencies should work together and with States, to assure comparability in data collection methods, operational definitions, and standardized indicators. AAPCHO notes that standardizing data sets to make them compatible, comparable, and useful for review and planning for local, State or regional issues and concerns are critical.
  3. National health objectives on data system development should be included in every chapter.
  4. Extend to all 50 States the capacity for public health surveillance of injuries and other acute health problems using data generated during emergency department visits and reported through electronic networks.
  5. Identify and create, where necessary, national data sources for each objective related to blindness and visual impairment.
  6. Use national Federal sources of cancer statistics (e.g., SEER, NCHS).
  7. State efforts are fragmented by competing needs for data from many of the funding agencies (e.g., NIH, CDC), and State needs to track objectives (Maine Department of Health and Human Services). Achieving consensus about the diseases, conditions, risk factors, services and outcomes that should be under surveillance or subject to assessment throughout the Nation would be helpful to State and local organizations. Basic data collection efforts such as vital statistics should be recognized and stated as essential and integral to the public health infrastructure. Federal agencies need to recognize and support the infrastructure to collect and disseminate data, as well as the process of applying data for program planning, implementation and evaluation by allowing financial support for data infrastructure and application procedures in their block grants and categorical grants.
  8. Mental health and substance abuse communities, the social service system, and the criminal justice system need to be included in integrating data systems.
  9. Managed care systems and insurance organizations should be explored as possible partners in collecting and disseminating data.
  10. Healthy People 2000 objective 22.4 was never addressed, and still should be considered for Healthy People 2010: "Develop and implement a national process to identify significant gaps in the Nation's disease prevention and health promotion data, including data for racial and ethnic minorities, people with low incomes, and people with disabilities of all ages, and establish mechanisms to meet these needs." The Latino Coalition for Healthy California noted that a plan to obtain data for Latinos for each objective needs to be developed, including gender and birthplace.
  11. Standard definitions of children with special health needs, children with disabilities, family support services/community-based care, resiliency, and other data elements/outcome measures should be developed.
  12. Disability, along with race and ethnicity, should be incorporated into demographic sections of national surveys and State-level data collection tools.
  13. Develop a set of health status indicators appropriate for Federal, State and local health agencies, and establish use of the set in at least 40 States. National and State surveys using the health status indicators should be required to include disability identifiers.
  14. An objective should be added to measure satisfaction with health information obtained during their most recent search for information.
  15. An objective should be added to increase the proportion of Healthy People 2010 objectives that can be tracked for select populations.
  16. An objective should be added to improve the availability and consistency of data on public health expenditures by essential public health services.
  17. An objective should be added to increase the number of public health agencies that use summary measures of population health and the public health infrastructure.
  18. NIH and other research vehicles must examine and restructure their funding priorities. Researchers and research institutions that examine health status indicators that can be used for Healthy People 2010 should be prioritized and targeted for additional funding to conduct this type of research. Furthermore, funding streams for research vehicles collecting baseline data on racial/ethnic health status should receive adequate funding proportional to the racial/ethnic populations being examined.
  19. Alternative databases/data sets (e.g., State, regional, local data) should be used to augment current data collection, especially when national data are limited due to sample size (AAPCHO, LCHC).
  20. When not available, AAPCHO would like this further explained: (1) data were not collected; (2) data have been collected but not analyzed; (3) data have been analyzed but have not been reported due to sample size or other methodological issues.
  21. HHS should disseminate guidance on sentinel or other appropriate data that can be used in the interim while national database/systems are being developed or upgraded to meet the new OMB Directive No.15 Standards.
  22. AAPCHO recommends that HHS increase the data collection and surveillance of critical health issues (e.g., tobacco, environmental health, diabetes) among the six U.S. associated Pacific Island jurisdictions. HHS must work through the proper channels (i.e., Ministries of Health in Republic of Palau, Marshall Islands and Federated States of Micronesia; Departments of Health in Guam, CNMI and American Samoa) to ensure that appropriate objectives are developed.
  23. It was noted that, of the 521 numbered objectives, only 100 (19.19 percent) have collected baseline Hispanic data. Therefore, a specific goal for Hispanics should be included under each objective for, but not limited to, the following sections: arthritis, osteoporosis and chronic back conditions, cancer, health communication, public health infrastructure, heart disease and stroke, diabetes, HIV, substance abuse, tobacco use, environmental health, immunization, and infectious disease. In addition, it was recommended that baseline and tracking data systems be identified, and data collected, analyzed and reported, for the aforementioned sections.
  24. Data for females aware of heart disease as the leading cause of death should include linguistically competent data collection in place of the telephone survey (Objective five, Heart Disease and Stroke Chapter).
  25. Behavioral Risk Factor Surveillance System should be used as a data source to get racial/ethnic group data for the breast examination and mammograms objective and Pap tests objective (Objectives 13 and 10, respectively, Cancer Chapter).
  26. The cancer registries should add a sub-objective to capture data broken down by race and ethnicity (Objective 15, Cancer Chapter).
  27. Include persons with disabilities as a select population as well as incorporate questions on disability into all existing and future data collection efforts conducted by the Federal government.
  28. Nineteen developmental objectives for vision were suggested, including "Identify and create, where necessary, national data sources for each of the blindness and visual impairment-related objectives".
  29. A reexamination of the decision to drop the data and surveillance goals from Healthy People 2000 and the re-inclusion of Chapter 22 in Healthy People 2000 on data and surveillance issues was suggested, given the importance of reliable data to monitor the Healthy People 2010 objectives. For example, with the addition of new objectives and revision of others in Healthy People 2010, it would seem that carrying the intent of several of the objectives from Chapter 22 would be crucial. The omission of 22.2 implies moving away from taking the lead at the national level for identifying and creating data sources to monitor the Healthy People 2010 objectives. HHS is passing the lead to the State and local agencies to develop and implement the processes and mechanisms with the omission of objective 22.4, which was part of Healthy People 2000. There does not appear to be a Healthy People 2010 counterpart addressing procedures for collecting comparable data and incorporating these procedures into the national health data system, or a comparable Healthy People 2010 objective for several of the objectives from Chapter 22.
  30. "Increase quality and years of healthy life" are great goals, but there is not a commitment to data collection to see if the goals are met. Availability of data, collection strategies, utilization of other existing data, and payment of the costs of data collection are some of the issues that need to be addressed to "measure and assure" that the goals in Healthy People 2010 are achieved.
  31. It was suggested that the final set of objectives use the updated Healthy People 2000 standard population for age adjustment.
  32. There should be a requirement that data be collected at the national, State and local levels, if subgroups are to be represented. The private sector (HMOs, MCOs, etc.) should be providing data.
  33. Certain "special populations"--prisons, military, reservations, and farm workers--are not addressed, and should be included in every section.
  34. All American Indian and Alaska Native tribes should be tracked in terms of their health status.
  35. The Background does not mention the BRFSS, but it is often mentioned as a data source for specific objectives. This system was not designed to make national estimates, and its appropriateness for making such estimates has not been established. Many of the objectives cite the BRFSS as the major data source. It was recommended to require a review by an independent group of the issues related to such use.
  36. It was noted that comparable questions do not necessarily mean comparable data. A systematic review of the comparability of results from BRFSS and NHIS was recommended.
  37. Because the BRFSS only provides a State-level estimate of adult diagnosed diabetes (>80 years old) and State prevalence rates vary from national prevalence rates, direct substitution may not be appropriate.
  38. While data may not be available, the objectives should still be stated in a measurable way. Goals cannot be achieved if there is no way to measure the extent to which related objectives have been accomplished. The Summit Health Coalition recommended that the U.S. Surgeon General convene a national working conference before the year 2000 to develop solutions to technical and implementation issues related to race and ethnicity data collection, to be attended by representatives of Federal, State, and local government agencies and the private sector, with substantial participation from minority communities.
  39. Because indicators should identify and prioritize the most important aspect of each focus area, it was recommended to reduce the number of indicators to five to ten within each focus area, and include a similar number of measurable, practical indicators per focus area, as well as an established time frame for accomplishment. In addition, these indicators should be directly tied to existing data measurement tools or data collection systems currently in place or under development.
  40. The American Dietetic Association suggested that HHS use BRFSS as a data source and consider leading health indicators that address healthy eating and obesity.
  41. Emphasis on epidemiology services is limited, and it was suggested to broaden this objective to include behavioral research and other evaluation areas that are practice-based. A similar recommendation pointed out that the team described is not a comprehensive epidemiology team, but an interdisciplinary data/epidemiology/evaluation research team of which epidemiology is a part.
  42. Newer evaluation models that include community-level indicators and policy change indicators should be developed and disseminated.
  43. An asthma surveillance system is crucial to successfully addressing respiratory disease. One example suggested establishing a surveillance system in at least 15 States to track asthma mortality, morbidity, access to medical care and asthma management for children 1 to 4 years old, 5 to 14 years old and 15 to 19 years old, in addition to the adult population of adults with asthma (Objectives zero and 13, Respiratory Disease Chapter).
  44. A statistical survey on representative drink servings, which has been proposed to members of the Dietary Guidelines Advisory Committee, should be conducted by HHS and SAMHSA. The survey would gather data from a sufficient number of restaurants, bars, nightclubs and other establishments to determine a scientifically defensible definition of "one drink". This approach would allow alcohol-related health policy to be used on accurate and meaningful data (Substance Abuse Chapter).
  45. A significant issue is the State's capacity to collect and use data; Objective 19 helps prioritize oral health surveillance. Also noted was the need for more systematic, comprehensive and usable data on the oral health status of minorities (Objectives nine and 19, Oral Health Chapter).
  46. Because primary and secondary syphilis is 13 times more prevalent among Hispanics than among whites, relevant data on Hispanic populations is needed (Objective one, STD Chapter).
  47. Few data are available to indicate whether dropping out of school increases the use of tobacco, or whether tobacco use is an indicator of who may or may not drop out of school. Also, there are no data linking drop-out rates, levels of education and tobacco use in one set of goals and objectives, ergo, smoking rates.
  48. It was recommended, as part of the State-wide cancer registry, that case information be stratified by age, including a zero to18 year category, to better effectively track the rate of childhood cancers (Objective 15, Cancer Chapter).
  49. Because non-physician primary care providers, such as nurse practitioners and certified nurse midwives, place a higher priority on addressing health behaviors and preventive services compared to physician PCPs, data should include the non-physician PCPs.
  50. The lack of potential data sources for objectives in the Mental Health and Mental Disorders Chapter was indicated.
  51. A better presentation for early identification of infants and children at high risk for developmental disabilities was recommended. Because the population included in the Metropolitan Atlanta Developmental Disabilities Surveillance Program consists of school-aged children rather than infants and toddlers, it was recommended that State Birth Defects Registries be strengthened to serve as a primary data source. Several States have developed methodological approaches based on the selection of specific ICD-9 codes, extracted from the Birth Defects Registry, and those are used to predict rates of developmental disabilities.
  52. Until a more accurate national surveillance system for violence against women has been established, incidence and prevalence data should be listed as a range (Objective 38, Injury/Violence Prevention Chapter).
  53. For objectives addressing deaths, hospitalization and emergency room visits, the goal is to first collect data, consistently in each State, in order to project a national target. The ICD codes used to measure each objective should be specific. Because interest lies in "intent" for the injury, it would be more useful if E-codes could be used to measure the rates (Objective one, two, nine, 10, 13, 14, 23, 28, 30, Injury/Violence Prevention Chapter).