CLINICAL RESEARCH: HEART FAILURE: EDITORIAL COMMENT
Racing away from bias*
Jonathan D. Sackner-Bernstein, MD, FACC ,* and
Hal A. Skopicki, MD, PhD
Clinical Scholars Program, Heart Failure and Cardiomyopathy Center, North Shore University Hospital, Manhasset, New York, USA
* Reprint requests and correspondence: Dr. Jonathan D. Sackner-Bernstein, Division of Cardiology, 1st Floor Cohen, North Shore University Hospital, 300 Community Drive, Manhasset, New York 11030, USA.
jonathansb{at}yahoo.com
Retrospective and subgroup analyses have created a means to dissect clinically relevant subgroups that may otherwise be difficult to study. Although often provocative and hypothesis-generating, these analyses have been used by many to dictate clinical practice. This is especially true of subgroup analyses from large, prospective multicenter studies in regards to race. Such an approach has led to several perceptions that shape current medical practice. These include: 1) that African Americans are at higher risk of developing cardiovascular disease than whites (1); 2) that African-American hypertensives are best treated with diuretics or calcium blockers instead of angiotensin-converting enzyme inhibitors or beta-blockers (2); 3) that African Americans have a worse outcome once developing ventricular dysfunction (3); 4) that the effect of beta-blocker therapy for African Americans with heart failure relative to whites is not consistent across the beta-blocker therapeutic class, with bucindolol being associated with a higher risk of death (4) and carvedilol associated with having equal efficacy (5); and 5) that African Americans with heart failure might not respond as well as whites to long-term angiotensin-converting enzyme inhibitor therapy (6,7). The literature is less clear about the risk of death in heart failure based upon race, with reports both of increased (1,8) and reduced risk (9,10). More consistent reports of increased risk of hospitalization are present (11,12).
In this issue of the Journal, Deswal et al. (12) investigated the risks of death and hospitalization for African Americans with heart failure relative to whites, and by using the Veterans Affairs (VA) national database, minimized the confounding factor of differential access to care between races. Although the database relies on administrative rather than clinical data sources, it does provide a large population source to be queried. Using rigorous statistical methods, the data show that African Americans and whites in the VA system are at equal risk of hospital readmission, and because the total outpatient visits were similar between African Americans and whites, the authors conclude that providing equal access to care eliminated racial differences in care. The authors also reported a lower risk of death for African Americans with heart failure at both 30 days and one year after the index hospitalization. Both findings are statistically robust, unaffected by the introduction of appropriate covariates. However, a major limitation of any retrospective analyses, particularly those relying on an administrative database, is the inability to ensure that the two groups studied were at equal risk (13).
It is possible that the African-American cohort would have a higher rate of rehospitalization than a cohort of whites with identical baseline risk but that the risks instead appeared equal because the African Americans in the VA system had less advanced disease, a possibility supported by the lower mortality rates at 30 days and one year (12). This would not be surprising if there were inequalities in disease severity or access to ancillary services, information and counseling (i.e., diet, exercise), but an administrative database cannot address these possibilities. Hypothetically, consider a group of African Americans who were modestly sicker than this cohort, sufficient enough to increase their mortality rate higher but still significantly less than the mortality risk of the cohort of white patients. In this hypothetical model, the risk of hospitalization should increase as well. The equivalent rate of readmission then argues that African Americans are at higher risk than is suggested by their disease severity, where disease severity is measured by mortality rate. Because the findings of their study indicate that readmission rates and mortality may be affected differentially in the population studied, as has been previously reported (11), these disparities warrant further investigation (12). Therefore, a reader should not draw the conclusion that universal access to care equalizes the risk for African Americans and whites. Instead, these data further strengthen this perspective as a hypothesis.
The authors provide several explanations for the equal risk of hospitalization with a lower risk of death, but three others should be considered. First, by using the emergency department instead of the clinics, African Americans may be entering an environment where they are more likely to be admitted. Second, even though a previous heart failure hospitalization portends an increased risk of death (14), the perception that the risk of heart failure hospitalization is related to disease severity, particularly as reflected by functional limitation, has not been prospectively tested and may be wrong. Third, treatment differences could be due in part to biases that influence decision-making (15). Although Deswal et al. (12) do not directly address this possibility, the potential for racism in clinical investigation and clinical practice is difficult to control.
As has been reviewed elsewhere, the use of race to categorize people extends from anthropologists attempting to understand the geographic distribution of homo sapiens (16). In the mid 1700s, people were categorized by both physical and psychological characteristics, and by the end of that century, the terms race and Caucasian were introduced, the latter to describe the residents of the Caucasus region of Asia (16). (The most prominent group in the region today is the Chechens.)
In clinical investigation, categorizing people based on skin color is not an appropriate scientific approach because skin color is not a biologic characteristic. As Goldstein and Chikhi report (17), the genetic differences between groups account for only 15% of the genetic variation seen across our species. As an example, Wilson et al. (18) surveyed the genetic homology among eight ethnic groups, focusing on the CYP2D6 hepatic enzyme system. Three of the groups were of African origin; yet, 62% of the Ethiopians and 21% of the Afro-Caribbeans were similar to Western European groups of Norwegians and Ashkenazi Jews. In contrast 24% of Ethiopians and 73% of the Afro-Caribbeans were like the Bantu. Should Ethiopians be considered Norwegian and Afro-Caribbeans be considered African? Additional studies point out the disparities that can exist for women and men because in many geographic regions there was unidirectional mating, as male explorers inhabited new lands (19).
Race is an important surrogate, but it is only a surrogate. Health-related differences that exist between races are not caused by belonging to a particular race, but rather are likely due to associated differences in social, educational, religious, or economic factors (20), in addition to genetic differences (17,21,22). The clinical perspective must remain focused on the aggressive treatment of cardiovascular diseases, but because race is not a relevant biologic grouping, our perspectives about mechanisms of disease, access to care, and effects of therapeutics must shift from a phenomenological to a genomic paradigm (16,21,22). We need to become scientifically color blind even while our society is not. The Deswal et al. (12) data facilitate the social arguments for prospectively testing the role of universal access as a means to equalize care but should not be used an argument to prove that this has occurred.
We live in a rich, advanced society. Our medical care is arguably the best in the world, and providing equal access to all can reduce risks. But to have maximum impact, we need more than equal access. Identifying what we mean when we subgroup according to race is critical; altering treatments or access to care based on blanket characterizations of race in the absence of a biologic basis cannot be justified. Unless prospectively controlled studies focusing on specific ethnic groups prove the usefulness of different clinical approaches, care must be blind to race and instead focus on treatable risks, such as hypertension, diabetes, and dyslipidemia. Eliminating biases in this context will improve the quality of health care one patient at a time.
 |
Footnotes
|
|---|
* Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology. 
|
References
|
|---|
1. American Heart Association. Heart Disease and Stroke Statistics—2003 Update. Dallas, TX; American Heart Association, 2002.
2. Douglas JG, Bakris GL, Epstein M. Management of high blood pressure in African Americans. Arch Intern Med. 2003;163:525–541[Free Full Text]
3. Dries DL, Exner DV, Gersh BJ, Cooper HA, Carson PE, Domanski MJ. Racial differences in the outcome of left ventricular dysfunction. N Engl J Med. 1999;340:609–616[Abstract/Free Full Text]
4. The Beta-Blocker Evaluation of Survival Trial Investigators. A trial of the beta-blocker bucindolol in patients with advanced chronic heart failure. N Engl J Med. 2001;344:1659–1667[Abstract/Free Full Text]
5. Yancy CW, Fowler MB, Colucci WS, et al. Race and the response to adrenergic blockade with carvedilol in patients with chronic heart failure. N Engl J Med. 2001;344:1358–1365[Abstract/Free Full Text]
6. Exner DV, Dries DL, Domanski MJ, Cohn JN. Lesser response to angiotensin converting enzyme inhibitor therapy in black as compared with white patients with left ventricular dysfunction. N Engl J Med. 2001;344:1351–1357[Abstract/Free Full Text]
7. Carson P, Ziesche S, Johnson G, Cohn J. Racial differences in response to therapy for heart failure: Analysis of the vasodilator-heart failure trials. J Card Fail. 1999;5:178–187[CrossRef][Medline]
8. Yancy CW. Heart failure in African Americans: A cardiovascular enigma. J Card Fail. 2000;6:183–186[CrossRef][Medline]
9. Tan EJ, Lui LY, Eng C, Jha AK, Covinsky KE. Differences in mortality of black and white patients enrolled in the program of all-inclusive care for the elderly. J Am Geriatr Soc. 2003;51:246–251[CrossRef][Medline]
10. Jha AK, Shlipak MG, Hosmer W, Frances CD, Browner WS. Racial differences in mortality among men hospitalized in the Veterans Affairs health care system. JAMA. 2001;285:297–303[Abstract/Free Full Text]
11. Rathore SS, Foody JM, Wang Y, et al. Race, quality of care, and outcomes of elderly patients hospitalized with heart failure. JAMA. 2003;289:2517–2524[Abstract/Free Full Text]
12. Deswal A, Petersen NJ, Souchek J, Ashton CM, Wray NP. Impact of race on health care utilization and outcomes in veterans with congestive heart failure. J Am Coll Cardiol 2004;43:778–84.
13. Mark DH. Race and the limits of administrative data. JAMA. 2001;285:337–338[Free Full Text]
14. Khand AU, Gemmell I, Rankin AC, Cleland JG. Clinical events leading to the progression of heart failure: Insights from a national database of hospital discharges. Eur Heart J. 2001;22:153–164[Abstract/Free Full Text]
15. Peterson ED, Shaw LK, DeLong ER, Pryor DB, Califf RM, Mark DB. Racial variation in the use of coronary-revascularization procedures—are the differences real? Do they matter? N Engl J Med. 1997;336:480–486[Abstract/Free Full Text]
16. Witzig R. The medicalization of race. Scientific legitimization of a flawed social construct. Ann Intern Med. 1996;125:675–679[Abstract/Free Full Text]
17. Goldstein DB, Chikhi L. Human migrations and population structure: what we know and why it matters. Ann Rev Genomics Human Genet 2002:129–52.
18. Wilson JF, Weale ME, Smith AC, et al. Population genetic structure of variable drug response. Nat Genet. 2001;29:265–269[CrossRef][Medline]
19. Carvalho-Silva DR, Snatos FR, Rocha J, Pena SDJ. The phylogeography of Brazilian Y-chromosome lineages. Am J Human Genet. 2001;68:281–286[CrossRef][Medline]
20. Kaplan JB, Bennett T. Use of race and ethnicity in biomedical publication. JAMA. 2003;289:2709–2716[Abstract/Free Full Text]
21. Haga SB, Venter JC. FDA races in the wrong direction. Science. 2003;301:466[Abstract/Free Full Text]
22. Cooper RS. Race, genes and health—new wine in old bottles? Int J Epidemiol. 2003;32:23–25[Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
Race and Survival After a First Heart-Failure Hospitalization
Journal Watch Cardiology,
May 28, 2004;
2004(528):
5 - 5.
[Full Text]
|
|
|
|
Top
References