Meeting of: Secretary's Council on Health Promotion and Disease Prevention Objectives for 2010
April 30, 1998, Proceedings

Agenda Item: Eliminating Health Disparities

DR. SIMPSON: I think the Assistant Secretary and Surgeon General has gone really beyond his segue in getting over halfway through the presentation. This is, indeed, important for us in the Office of Minority Health. We are certainly glad to have an opportunity to present the issues and challenges this morning.

This is, again, a welcome opportunity to focus our collective attention and efforts on improving the health of all Americans, white and non-white, men and women, those who are above and those who are below the poverty level; those at every age, including the very young, adolescents, and the elderly; the highly educated, and those less so; those with functional limitations due to disease and disability.

The President's Initiative on Race, as described by Dr. Satcher, with its focus on racial harmony and unification in a nation that is increasingly more racially and ethnically diverse, the co-leadership of Dr. Satcher and Dr. Hamburg, and the Department's implementation of the race and health component of the President's Initiative on Race, and the strong and continued leadership and support of the Secretary and the Deputy Secretary, in ensuring that the Department develops and implements specific initiatives to address the health of racial and ethnic minorities in all the program activities, all of this, coupled with Healthy People 2010, offers us an unprecedented opportunity to really make a difference in closing the gap in health disparities between racial and ethnic groups across the country.

The opportunity exists, if we take advantage of it, to frame Healthy People 2010, and how differences in health status and health needs between populations are addressed in a way that is fundamentally fair for, and more inclusive, of all groups. I speak primarily of groups that differ by race and ethnicity. Please note that my remarks are generally applicable to populations that differ by gender, age, functional status, income, educational level and other variables where health disparities are known to exist.

A renewed and vigorous commitment to addressing health disparities in Healthy People 2010 is needed if we are to improve on the experiences of this decade, and close the gaps between racial and ethnic minorities. As Dr. McGinnis and Dr. Lee pointed out in their mid-decade review, while the health of the general population has improved and continues to do so, major racial and ethnic disparities remain.

Approximately half of the Year 2000 objectives are either proceeding in the wrong direction or lack sufficient data to determine progress, with such problems greater for minorities than for the rest of the population. Their review also showed that for some groups, including Hispanics and Asians and Pacific Islanders, looking beyond the aggregate data may be needed in order to unmask disparities for specific sub-populations.

Before I share my views regarding new or revised strategies, our approaches for better addressing racial and ethnic health disparities in Healthy People 2010, let me say a few words about data. Our experience with Healthy People to date illustrates how our ability to accurately identify and monitor, much less to effectively address, health problems for racial and ethnic minorities is seriously compromised by serious data issues and challenges.

Often the issue is one of no national data for racial and ethnic minority groups and subgroups. Where national data exists, it is insufficient or of poor quality.

Lack of good national data does not necessarily mean lack of national health disparity. The usual reason for getting this lack of data, with the small numbers of some of the groups or subgroups, and the lack of resources to do the kind of data collection needed, is that we don't have the money.

The problem is -- no data, no objectives, no sub-objectives. Where there are no objectives or sub-objectives, there is little, if any, policy or programmatic attention at the national level to a number of health problems of real concern to these groups or subgroups.

Without such attention, the gap remains or increases. For example, data from the state of Hawaii indicate that more than four out of 10 Native Hawaiians are overweight, much higher than the rate for the total population, which is approximately 28 percent.

Data from Puerto Rico indicate that more than half of all Puerto Ricans -- a vast disparity from the total population -- lack health insurance. Not only do our national estimates exclude U.S. insular areas like Puerto Rico, but Healthy People 2000 national objectives have also excluded these geographic areas.

Related to this issue of no data or poor data to identify and monitor health disparities by race and ethnicity is the issue of explaining away all such disparities on the basis of socioeconomic status, and replacing race and ethnicity variables with those related to socioeconomic status. As was stated in the final report, issued last month, of the President's Commission on Consumer Protection and Health Care Quality, such factors as economic status contribute to the vulnerability of groups such as racial and ethnic minorities, but cannot fully explain their poor health status, or the significant differences in health care and health services that they often experience in comparison with non-minorities.

Clearly, we need to invest more in racial or ethnic data collection at the federal level. The HHS policy issued by Secretary Shalala last October to include data on race and ethnicity in all Department-funded or sponsored data collection and reporting systems is a tremendous step in the right direction. We also need to encourage such data collection at regional and local levels, especially in jurisdictions where minority populations reside, where such data can illuminate health problems and disparities. We need to consider the use of existing regional and local data when national data are poor or unavailable.

Another major data challenge alluded to earlier is that of identifying and addressing health disparities for racial or ethnic subgroups that may be masked when data are aggregated. For example, the age adjusted death rate for Asian and Pacific Islanders in the aggregate is 350 per 100,000, which is lower than the rate of 524 per 100,000 for the total population, reaffirming for many the inaccurate portrayal of Asian Americans and Pacific Islanders (AAPIs) as a healthier model minority. However, when data for AAPIs are disaggregated, a rate of 901 per 100,000 for Native Hawaiians -- a huge disparity -- is unmasked.

For Asian Americans alone, the percentage of Asian American births that occur among teenage mothers -- 12.9 for Asian American mothers born in the United States -- masks the fact that the rate varies widely: from three percent for Chinese mothers born in the United States to a high of 19 percent for Filipino mothers in the United States.

The total Hispanic infant mortality rate of 6.1 for 1,000 live births hides the fact that Puerto Ricans, whose rate is 8.6 per 1,000 live births, have a much higher rate than non-Hispanic whites.

These examples reinforce the need to correct racial and ethnic data, not only by the OMB Directive 15 standards for racial and ethnic classification, but also by subgroups, and to disaggregate data in subcategories as much as possible.

In addition to these data issues, what are the other critical issues relating to closing the gap between racial and ethnic groups in this country? Given that current strategies and approaches for addressing the Healthy People 2000 goals of reducing health disparities are not working as well as they should, are there other strategies or approaches that should be considered for 2010?

Let me outline several additional issues, strategies and approaches, to better target policy attention and resources to Americans with the greatest negative disparities, in order to achieve improved health for all.

One of the constant concerns expressed by minority voices in the Healthy People 2000 cross-cutting progress reviews, at minority health conferences, and other venues, has been that of having to accept lower targets and lower expectations and standards for health status and health care than those set for white Americans. Repeatedly, their recommendations have been to change the Healthy People 2010 goal from one of reducing to one of eliminating health disparities, and to use the same targets for all Americans, rather than lower targets for some.

The President, as has been stated, has embraced the elimination, rather than the reduction, of health disparities and the use of the same targets for the six areas in the race and health initiative. A clear definitive position is needed regarding the application of such a principle across the board to all health issues and focus areas in Healthy People 2010. The argument against such a goal and the use of the same targets for all races and ethnicities is that, in many instances, the gaps are too wide and their elimination unrealistic for 2010.

The question then is, if not 2010, when? Because it has been suggested in discussions about the 2010 objectives that the second goal of Healthy People be changed from reducing to eliminating health disparities among racial and ethnic groups, and because it has also been suggested that the targets for any particular health issues be the same for all racial and ethnic groups, a new issue has been raised as to whether to set the targets based on national averages as, in many instances, we have done in Healthy People 2000, or on a figure that is better than the best group.

The Office of Minority Health supports continuation of target-setting based on national averages, rather than on figures that are better than the best, that would be even higher and further away from the baselines of those who are among the worst in these categories.

Before I discuss why we favor basing our targets on national averages rather than better than best, let me discuss one other issue that is closely related, namely, the issue of how best to depict or present such disparities between racial and ethnic groups relative to each other and relative to the targets. Currently, sub-objectives exist only for those racial or ethnic groups with negative disparities compared with the total population and for whom the disparity can be validated by national data.

Groups with positive disparities compared with the national averages are not displayed. An example, Healthy People 2000 objective 14.1, to reduce the infant mortality rate to no more than seven per 1,000 live births, is typical of how disparities are currently depicted. In this example, the infant mortality rate in 1987, based on the national average for the total population, was 10.1 per 1,000 live births. This was used as a basis for setting the target.

Using the rate of decline during the 1980s of 2.8 percent per year, the target for the year 2000 was set at seven per 1,000 live births.

Given the way that racial and ethnic groups are portrayed in this objective, it cannot be determined whether any population group had rates that were either below the national average for the baseline year or below the target for the year 2000. This kind of depiction of racial and ethnic groups has resulted in a backlash by some groups -- who do not see themselves relative to the targets -- against those who, by virtue of such selective identification, are inappropriately criticized for being special populations receiving special attention, possibly at the expense of those not being identified. We believe a more inclusive approach would be to identify all racial and ethnic population groups relative to the target. This approach is also consistent with those initiatives such as the President's race initiative that unify, rather than divide, Americans.

For the health issue-related objectives in Healthy People 2010, we suggest that consideration be given to the depiction of all racial and ethnic population groups using, as a minimum, the categories established by OMB Directive 15, and to allowing for additional subgroups when there is a reason to believe that important negative disparities will be revealed by disaggregating the data. The current OMB Directive 15 standard categories are American Indians or Alaskan Natives, Asian or Pacific Islanders, non-Hispanic black, non-Hispanic white, and Hispanic.

These standards have recently been revised and the new standards should be implemented by the beginning of 2003. The new categories are American Indian or Alaskan Native, Asian, Native Hawaiian or other Pacific Islander, black or African American, white, and Hispanic or Latino.

Using Healthy People 2000 objective 14.1 on infant mortality as an example, let me show you what would happen if all the OMB Directive 15 categories were displayed with additional sub-categories for a couple of groups. This example shows that disaggregation can unmask positive as well as negative variances, and that racial and ethnic minorities may be among those who are doing the best for a given health concern.

As you can see in this case, using the minimum categories of race and ethnicities, whites, AAPIs and Hispanics had rates that were better than the national average of 10.1 per 1,000 live births in 1987, the baseline year. With the year 2000 target set at seven per 1,000 live births, all racial and ethnic groups would be above the target but, again, whites, Hispanics and especially AAPIs would not have very far to go to reach the target.

If you disaggregate the data, as has been done here for Hispanics and AAPIs, you will see that infant mortality rates for Hispanic and AAPI subgroups are not only at or below the national average -- 10.1 for the baseline year -- but rates for some subgroups are at or below the year 2000 target of seven, which was based on the national average for the baseline year. Japanese, Chinese and Filipino subgroups are doing better than the target, with Cubans just under the target by a tenth of a percent. In this case, those who were at or below the national average are already at or below the target, but are hidden because the data are aggregated.

In other cases, populations that are doing better than the targets may be unrecognized because the data are simply not presented for groups that are doing very well. I want to emphasize that the situation has existed all along with targets which have been based, for the most part, on national averages.

Why change the targets to better than best simply because all groups may now be displayed? This brings me back to the issue of target-setting. While we need to encourage efforts that maintain or improve the health of those who are among the best, at the same time we also need to more vigorously encourage improvement in the health of those who are among the worst. We cannot close the gap if we keep setting the target further out of reach for those in the poorest health, which is what setting better than best targets will do.

What is needed is clearly to indicate, with a footnote or some other device, those groups whose baselines are better than the target, and to ensure understanding that their current baselines are the targets which must be maintained or improved.

Let me look at an example of how this would work using the infant mortality objective again -- with the target being proposed for Healthy People 2010 -- which replaces objective 14.1 of Healthy People 2000. This overhead shows the proposed overall infant mortality targets. Here, the new target -- 5.3 per 1,000 live births -- has been set to reflect a 30 percent decrease from the baseline, 7.6 per 1,000 live births. The baseline reflects the rate for the total population. As in the previous examples, whites, all Asians, most Hispanics and, in this case, American Indians, have rates that are below the baseline, 7.6, for the total population. They are also very close to, if not already at or better than, the target 5.3, as are the Chinese, Japanese, Asians, Indians, Koreans and Cubans.

Rather than setting better-than-best targets, we recommend that, for those groups who are already at or better than the target, efforts concentrate on maintaining them among the best rates. At the same time, we should aggressively focus attention on those who are among the worst, in this case, blacks, Samoans, Puerto Ricans and other Hispanics. Efforts will also need to be made in this objective to bring the rest of the groups who are nearing the target up to the target, and to improve data collection for groups, such as Guamanians, for whom national data are poor or nonexistent.

As Dr. Satcher has just described, this is not a zero-sum game. Setting targets based on national averages has never meant that policies and programs should discourage continued health improvement for those who are doing relatively well, that is, those who are at or better than the national average or who have already reached the total population target for a Healthy People objective. This was not intended in Healthy People 2000 and it is not intended in Healthy People 2010.

Given the significant health gaps between racial and ethnic groups in this country, greater and more accelerated efforts will be required over the next decade and perhaps beyond to truly improve the health of those who are among the worst, and to close the gap.

This suggests a way to measure progress, and to hold all of us accountable for such progress. That is, progress in the right direction can be measured not only in terms of how much, but how fast, movement toward the target is taking place. The amount and rate of change can be compared between the best and the worst off to determine if, and by how much, the gap is being narrowed. This approach is preferable to that of setting interim targets for racial and ethnic minority populations, as has been proposed in some discussions.

The last issue I want to mention has to do with whether to include or exclude populations for whom no data are available. As previously mentioned, for Healthy People 2000, population sub-objectives are only allowed and presented when national data are presented.

Whether or not non-national data are allowed in the future, it is critically important to identify those population groups for whom national data are not available as well. This could be done by simply using a no data indicator, when it is applicable for a particular population or subgroup. In the infant mortality group just shown, double asterisks were used to indicate that data were poor for a particular subgroup, but the group was not excluded.

In this way, we would draw continually critically needed attention toward the lack of strong data infrastructure and the need for more robust data collection efforts and show particular populations for whom such data collection is needed. By not removing those populations for whom data are not yet available and using no data indicators, we are avoiding mistaken notions about whether or not a health problem exists for those groups, since we don't and won't know until such data are collected. This approach will also reinforce the notion of inclusivity.

I thank you for this opportunity to present this overview of the issues and challenges facing us as we try to address health disparities in the 2010 objectives. I know that I have probably gone longer than I should, but I am glad that I was first with the data. The issues are complex, inter-related and important to the notion of effectively closing the gap. I look forward to today's discussion on these critical health issues for racial and ethic minorities and for other population groups experiencing significant negative health disparities. Thank you.


DR. SATCHER: Thank you very much, Clay. You didn't go as long as you told me you were going to go, so you got a chance to finish. Thank you very much.

The Secretary has joined us. I told you that she would be coming in. So, we are going to take an interlude here to hear from her. I am sure you will have some discussion of Clay's presentation later, and then we are going to follow from there the remainder of the agenda.

Back to Table of Contents