Author + information
- Rita F. Redberg, MD, MSc, FACC⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Rita F. Redberg, School of Medicine, Division of Cardiology, University of California, San Francisco, California 94143-0124.
This reality is common to men and women in the U.S. of all races: they all are more likely to die from heart disease than any other cause (1). Beyond this commonality, however, lie substantial variations in cardiac care according to gender and race. In 2002, a comprehensive review of studies about disparities by the Kaiser Family Foundation and the American College of Cardiology Foundation (2) found significant differences in cardiac care for minority groups. According to these studies, African-American patients, for example, are less likely than white patients to undergo diagnostic tests and revascularization, even after controlling for clinical and socioeconomic factors. The Kaiser findings mirrored a 2002 Institute of Medicine report (3) that concluded that racial/ethnic variations in medical care are widespread.
Disparities by gender also are well documented: Women are diagnosed later than men and receive fewer therapies (4). Women receive fewer coronary angiography and revascularization procedures (5), and women have higher complication and mortality rates after revascularization (6,7).
The congressionally mandated National Healthcare Disparities Reports found that such disparities result from complex factors with many contributors and no single cause (8). Factors may include: 1) that certain minority groups have more risk factors and are less healthy in general; 2) that the course of cardiac disease is different in different subgroups; 3) that physicians are biased in their recommendations; and/or 4) that women and minorities tend to decline certain treatment options. Surprisingly, most cardiologists are unaware of race and gender disparities in treatment (9). To help educate health professionals about disparities and identify and develop strategies to reduce gaps in care based on race and culture, the American College of Cardiology, the American Medical Association, and others recently have formed the Commission to End Health Care Disparities (10).
Differences in care do not necessarily mean inferior care if they are based on differences in patient characteristics. In addition, more care is not always better care, as first demonstrated more than 30 years ago by Wennberg and Gittelsohn (11) and more recently by Fisher et al. (12,13). We can agree, however, that the most egregious reason for differences would be bias—conscious or unconscious. Rathore and Krumholz (14) suggest a framework for categorizing variations in care into three tiers: differences, disparities, and biases. Variations in care can be considered “disparities” if: 1) the difference in health care reflects shortfalls in appropriate care that cannot be explained by other patient factors, and 2) there are associated adverse health consequences. “Bias” would apply to disparities that cannot be explained by health system factors (provider characteristics).
Gender disparities in the CURE trial
This issue of the Journalincludes two papers (15,16) with new insights on disparities in care of patients with acute coronary syndrome (ACS). Anand et al. (15) performed a post-hoc analysis of gender differences in the management and outcomes of patients with ACS from the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial. The authors compared the outcomes of the 4,836 women and 7,726 men from 28 countries who suffered ACS during the period of December 1998 and October 2000. The researchers examined the use of invasive procedures, revascularization rates, and incidence of death, myocardial infarction (MI), or stroke after approximately nine months. The primary composite end point—stroke, MI, or cardiovascular death—was similar in women (4.4%) and men (4.9%).
Although the end point for women and men was nearly the same, there were significant disparities in treatment. For example, women with ACS were less likely to receive coronary angiography than men (25.4% vs. 29.5%, respectively). Interestingly, the women in the CURE trial receiving angiography were twice as likely to have normal coronaries (26.7% normal in women vs. 13.2% normal in men) (15). One is tempted to conclude, therefore, that fewer women undergo angiography because women are less likely to have cardiac disease. However, the authors found that the probability of a woman being referred for angiography was unrelated to her Thrombolysis In Myocardial Infarction (TIMI) risk score, that is, high-risk women were not more likely to be referred to angiography than low-risk women.
The findings in the CURE study suggest, therefore, that the differences in referral to angiography are due to bias, not health factors. Thus, the general perception that women are at lower risk for heart disease may negatively impact the care of high-risk women. (Note, however, that a limitation of the CURE study is that the authors did not address whether angiography in women was related to better outcomes.)
The Anand et al. (15) study provides evidence that the disparity in treatment of women with ACS derives not from differences in the disease itself but rather from physician decisions unrelated to risk assessment. Of course, risk stratification, not physician bias, should determine how aggressive cardiac treatment should be in individual women, just as in men, even though women are at lower risk as a group.
Race disparities in Kansas City
Spertus et al. (16) report results of an observational study comparing one-year outcomes, measuring quality of life and physical function, between African-American and white patients in Kansas City. From an ACS registry, the researchers used the Seattle Angina Questionnaire and the Short Form (SF)-12 to interview patients admitted with ACS at two hospitals from March 2001 to October 2002. Follow-up interviews were used to assess the patients’ clinical and health status one year later. The authors also looked at rates of revascularization by race. On average, the African-American patients (196 of 1,159, or 17%), although younger, were more likely to suffer from diabetes, hypertension, obesity, and renal failure and to smoke.
The African-American patients started with essentially the same baseline quality of life as white patients (approximately 50 Seattle Angina Questionnaire [SAQ]), but with slightly lower physical function measures than whites (approximately 37 SF-12 vs. approximately 39 SF-12; Fig. 1B of the Spertus et al. paper ). After one year, both groups enjoyed improved quality of life, although the improvement for African-American patients was significantly less than for the white patients (70.6 SAQ vs. 83.9 SAQ). Distressingly, the African-American patients’ physical function actually decreased after one year (to 36.8 SF-12; Fig. 1C of the Spertus et al. paper ) and lagged substantially behind the white patients, who improved in physical function (to 43.2 SF-12). These differences remained after adjusting for medical conditions and sociodemographic factors.
Higher revascularization rates in white patients did not explain their higher quality of life and physical function scores. It appears that comorbidities or other unmeasured factors (perhaps relating to care of patients between discharge and one year) are responsible for the disparate scores. The authors suggest, rightly, that future investigation should examine outpatient treatments, compliance with medications, and biologic mechanisms that might explain the disparities. A limitation of this study is that the authors did not adjust for hospital factors; if the white patients were treated primarily at one of the two hospitals in the study and African-American patients mostly at the other, hospital characteristics may also account for the differences found in the study. Nevertheless, the study offers important new data on racial differences in quality of life in patients with ACS.
Process measures versus outcome measures
Medical care in general, and cardiac care in particular, may be assessed by using direct outcome measures (i.e., how well or how long patients actually live), or by using “process” measures (i.e., measures of procedures or treatments as indirect indicators of health and outcomes). The validity of process measures depends on how well they correlate to outcomes. Processes, such as therapies or procedures, are valuable only if they lead to improved quality and/or quantity of life. Most quality indicators in cardiology studies are process measures, such as the administration of drugs (e.g., acetylsalicylic acid, beta-blockers, angiotensin-converting enzyme inhibitors) (17) or measurement of ejection fraction in patients admitted for heart failure (18). These process measures are presumed to lead to improved outcomes, such as a decrease in rates of myocardial infarction and death and, indeed, many studies show these treatments are associated with improved outcomes.
However, data to support the association of process measures and outcomes are more limited for women and minority populations as the result of lower enrollment in clinical trials. These missing data are crucial. For example, although initial trials showed benefit of glycoprotein IIb/IIIa inhibitors in men with ACS, a meta-analysis of six ACS studies found no benefit of glycoprotein IIb/IIIa inhibitors in women and, more alarmingly, a 15% increased risk of MI and death (19).
Do differences in the course of cardiac disease explain disparities?
Different cardiac care for certain subgroups can be appropriate, based on different patient risk profiles and/or demonstrated differences in outcomes. Differences in treatment, by gender, when based on reliable gender-specific data, should be reassuring, not disturbing. Race disparities, on the other hand, present a more troubling case, because they appear to be associated with many factors, such as insurance status, treatment facility, and geographic neighborhood (20). More research is needed.
Do women and minorities choose to have fewer procedures?
Another possible explanation for some differences in treatment is patient choice. It is well documented that ethnic minorities and women are less likely to receive some types of care (21,22). It is unclear, however, whether women and minorities are less likely to be offered certain procedures or simply are less likely to agree to have them. At least one study found that women are more willing than men to undergo invasive cardiac procedures (23), suggesting that fewer cardiac catheterizations in women are due to fewer recommendations for the procedure. Another study of patient refusal in the use of coronary angiography in post-MI Medicare beneficiaries, however, found that elderly female and black patients are slightly more likely to refuse angiography than male and white patients (24). However, patient refusal overall is uncommon and accounts for only a fraction of observed race and gender differences.
How concerned should we be over disparities?
The articles in this issue of the Journalagain confirm that there are racial- and gender-based disparities in cardiovascular care, and their conclusions leave us with several areas for future exploration. Both studies showed lower rates of revascularization in female and African-American patients that were not related to outcomes differences. Women and minorities had poorer quality of life indicators, poorer physical function (16), and higher rates of rehospitalization for angina (15), but no difference in hard outcomes (myocardial infarction or death). It is important to have such outcomes data; using process measures as surrogates for outcomes data is less reliable in women and minorities because data for these groups are so limited. Including more women and minorities in clinical trials and including gender- and race-specific reporting in medical journals would greatly help us to understand the reasons for race and gender disparities in health care.
Dr. Redberg is supported in part by the Robert Wood Johnson Foundation Health Policy Fellowship and by the Flight Attendant Medical Research Institute.
↵* Editorials published in the Journal of the American College of Cardiologyreflect the views of the authors and do not necessarily represent the views of JACCor the American College of Cardiology.
- American College of Cardiology Foundation
- American Heart Association
- ↵(2002) Racial/Ethnic Differences in Cardiac Care: The Weight of the Evidence (Henry J. Kaiser Family Foundation and American College of Cardiology Foundation, Menlo Park, CA).
- Institute of Medicine
- Lansky A.J.,
- Hochman J.S.,
- Ward P.A.,
- et al.
- Vaccarino V.,
- Lin Z.Q.,
- Kasl S.V.,
- et al.
- Jacobs A.K.,
- Johnston J.M.,
- Haviland A.,
- et al.
- Moy E.,
- Dayton E.,
- Clancy C.M.
- Lurie N.,
- Fremont A.,
- Jain A.K.,
- et al.
- ↵American Medical Association Commission to End Health Care Disparities. Available at: http://www.ama-assn.org/ama/pub/category/14629.html. Accessed June 16, 2005
- Wennberg J.,
- Gittelsohn A.
- Anand S.S.,
- Xie C.C.,
- Mehta S.,
- et al.
- Spertus J.,
- Safley D.,
- Gang M.,
- Jones P.,
- Peterson E.D.
- Antman E.M.,
- Anbe D.T.,
- Armstrong P.W.,
- et al.
- Hunt S.A.,
- Baker D.W.,
- Chin M.H.,
- et al.
- Tonne C.,
- Schwartz J.,
- Mittleman M.,
- Melly S.,
- Suh H.,
- Goldberg R.