The Antidote

Counterspin for Health Care and Health News

Sunday, November 05, 2006

Thinking about health disparities

A number of new research articles and news stories recently have spurred me to think about how we think about the disparities in health, and what this vague term means. According to Wikipedia, the U.S. Health and Resources Services Administration (HRSA) defines it health disparities as "population-specific differences in the presence of disease, health outcomes, or access to health care." These populations are usually defined by race, ethnicity, age, sex, insurance status, rural vs. urban residence, and/or socioeconomic group.

I think of health disparities as an element of quality, or the lack thereof. Sure, everyone may be getting the same (poor-quality) health care. It's more likely that poor quality health care is seen more often in some populations than others, via reduced access to quality care (which could take many forms, from underinsurance to language barriers) or via discrimination (which can also affect health in ways unrelated to health care per se). Often, when you look at health outcomes, it's hard to tease apart the relative roles of access, discrimination, and other factors. Because access to quality care for everyone is (or should be) a priority, however, much good research is emerging on differences in care between different populations. But it isn't easy, as this post by my blogging colleague Cervantes points out, starting with the difficulties of defining ethnic populations.

Here are a few recent articles, which I chose because they illustrate different levels at which disparities are manifested:

Mays and colleagues review research on the psychological/physical effects of discrimination on health outcomes. A press release on this article explains the general mechanism of effect of discrimination thus:
When a person experiences discrimination, the body develops a cognitive response in which it recognizes the discrimination as something that is bad and should be defended against, Mays said. She said this response occurs for the most part even if the person merely perceives that discrimination is a possibility.

Starting with the brain's recognition of discrimination, the body sets into motion a series of physiological responses to protect itself from these stressful negative experiences, Mays said. These physiological responses include biochemical reactions, hyper-vigilance and elevated blood pressure and heart rate. With many African Americans, these responses may occur so frequently that they eventually result in the physiological system not working correctly.
A second paper, by Trivedi et al. documents lower-quality care received by older African-Americans compared to whites. Specifically, they are less likely than whites to have their blood pressure, cholesterol, and blood sugar under control. Each of these measures is a reliable indicator of health-care quality. The results were not explained by blacks being in lower-quality health plans; the differences were seen within all 115 of the Medicare plans studied. The paper did not settle the question of why such differences exist, but it did find that demographic factors like income and education explained only some of the gap observed and, of course, lifestyle factors like diet that do not relate to quality likely explain some of the results.

I would suggest that a next step would be to tease apart the contributions of lifestyle and health care quality to the health differences - one way to do that would be to compare process measures, which measure actual delivery of care as opposed to health outcomes. I checked the National Healthcare Disparities Report; the 2005 report gives a similar result to the Trivedi paper - control of hemoglobin A1C (a measure of blood glucose) is better in whites than blacks. However, 2004 report presents a related process measure: adults with diabetes who had a hemoglobin A1C measurement at least once in the past year. Interestingly, for this measure, blacks and whites appear to be approximately equivalent. This is not to say that A1C measurement is not related to good diabetes outcomes, or in other words unrelated to quality, but it demonstrates the role of other factors - possibly even care-related factors - in determining health outcomes. (By the way, I highly recommend the above-cited Disparities Report, and its companion the National Healthcare Quality Report, as useful overviews of U.S. data on healthcare quality; I did play an advisory role in both of these documents.)

In an accompanying press release to the Trivedi study, the first author noted that many plans don't even collect information on race and ethnicity of their patients, so they may not even know they have a problem. Perhaps this can be explained by a naive and unfortunate assumption that their care is race-blind, but it certainly means that any existing disparities will go unaddressed.

A third paper on disparities relates somewhat, but less directly, to quality. A little background: in the past few years, studies have emerged that address the hypothesis of whether the number of surgeries of a particular type (i.e., surgical volume) done by individual surgeons or within a facility is related to health outcomes. The consensus seems to be that volume at the facility level, but not at the surgeon level, is related to outcomes. In other words, facility-level surgical volume is an indicator of quality. The paper by Liu et al. found that, minorities c minorities (Blacks, Asians, Hispanics) compared to white patients, the uninsured, and Medicaid patients were more likely to receive surgical care at lower-volume surgical centers. That's not a direct measure of either discrimination or access to care related to race or ethnicity, but could represent geographic differences (e.g., proximity to quality hospitals), and the fact that it's related to economic differences (Medicaid and uninsured populations) suggests that access to care could play a role. An editorialist on the study pointed out that referral of patients to higher-volume hospitals does not solve the problem of quality differences, but is an "end run" around it, by shifting patients away.



0 Comments:

Post a Comment

<< Home

Listed on BlogShares