Correspondence

Race, Sex, and Physicians' Referrals for Cardiac Catheterization

N Engl J Med 1999; 341:285-287July 22, 1999

Article

To the Editor:

We were surprised by the conclusion of Schulman et al. (Feb. 25 issue)1 that “the race and sex of a patient independently influence how physicians manage chest pain.” In fact, their data do not support this statement. The results of the multivariable logistic-regression analysis presented in Table 5 of their article clearly show that with the use of white men as the reference group, only black women were significantly less likely to be referred for cardiac catheterization. In fact, the odds ratio for catheterization was 1.0 for both black men (P=0.99) and white women (P>0.99), as compared with white men, suggesting an equivalence rather than a difference in the rates of referral for cardiac catheterization for these groups. It therefore seems probable that the significant differences in referral patterns with race and sex used as separate factors are dependent solely on the striking results for black women.

There is an obvious bias in the study that may explain why black women were referred for cardiac catheterization less often than the other groups. A significantly higher number of female physicians were assigned to black female patients, as shown in Table 1 of the article. In addition, we do not know the racial composition of the group of female physicians and the group of male physicians. We are not supplied with the information to determine whether female physicians were less likely than male physicians to refer patients for cardiac catheterization. If female physicians were less likely to make such referrals, there may be no difference whatsoever in referral practices according to the race or sex of the patient.

Gérard Helft, M.D., Ph.D.
Stephen G. Worthley, M.B., B.S.
Sylvie Chokron, Ph.D.
Mount Sinai Medical Center, New York, NY 10029

1 References

1. 1

   Schulman KA, Berlin JA, Harless W, et al. The effect of race and sex on physicians' recommendations for cardiac catheterization. N Engl J Med 1999;340:618-626
   Full Text | Web of Science | Medline

To the Editor:

The study by Schulman et al. suggesting that physicians' recommendations for cardiac catheterization are influenced by patients' race and sex is disturbing, although perhaps not surprising, given the pervasiveness of racism and sexism in American life. The study considers only minimally the influence of other factors as determinants of the differences in rates of referral for catheterization. This narrow perspective has unfortunately led to the unwarranted interpretation in the press that the physician's own bias is the predominant reason for the known differences in the use of catheterization.1 In actual practice, however, the situation is considerably more complex. One recent study has shown that substantial preexisting differences in cardiac-procedure rates for a defined group of black and white patients disappeared when all patients were equally and adequately covered by insurance and were provided with a source of comprehensive, clinically appropriate care.2 A separate study found that women and members of minority groups actually underwent appropriate cardiac procedures at the same rates as all other patients, as long as they were adequately covered by insurance and received their care at hospitals that provided these procedures.3 In the real world, it seems, “access, quality, and equity are related. It is a mistake to consider them separately.”4

I would also like to note that although the relative “risk” of referral for cardiac catheterization for black patients and women as compared with white patients and men was 0.93 (an absolute difference in referral rates of about 5 percentage points), the press picked up on the odds ratio of 0.6 and reported that “blacks and women with chest pain are 40% less likely than whites or men to be referred by physicians for cardiac catheterization.”1 It is unfortunate that the authors did not emphasize the simpler, more direct measure. Of equal concern, patients thought to have nonanginal pain were referred for catheterization almost as frequently as those believed to have definite angina (relative risk, 0.94), which casts considerable doubt on the clinical validity of referral rates as determined under the study conditions.

Frank Davidoff, M.D.
, Philadelphia, PA 19106-1572

4 References

1. 1

   Rubin R. Heart care reflects race and sex, not symptoms. USA Today. February 25, 1999:A1.
   

2. 2

   Daumit GL, Hermann JA, Coresh J, Powe NR. Use of cardiovascular procedures among black persons and white persons: a 7-year nationwide study in patients with renal disease. Ann Intern Med 1999;130:173-182
   Web of Science | Medline

3. 3

   Leape LL, Hilborne LH, Bell R, Kamberg C, Brook RH. Underuse of cardiac procedures: do women, ethnic minorities, and the uninsured fail to receive needed revascularization? Ann Intern Med 1999;130:183-192
   Web of Science | Medline

4. 4

   Kravitz RL. Ethnic differences in use of cardiovascular procedures: new insights and new challenges. Ann Intern Med 1999;130:231-233
   Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: We agree with the statement by Schwartz et al., in their Sounding Board article in this issue of the Journal, 1 that the concept of odds ratios is difficult to understand and that odds ratios have the potential to mislead readers who are unfamiliar with statistical methods. Although our analyses used appropriate statistical methods, in retrospect we should have converted odds ratios to risk ratios and underscored absolute event rates in our univariate analysis. Our study hypotheses, as stated in the original grant application, were that blacks would be less likely to be referred for cardiac catheterization than whites and that women would be less likely to be referred than men. Our reporting of the sizes of the main effects of race and sex is therefore consistent with fundamental statistical principles.

Also, although we agree that cardiac catheterization and other invasive treatments for heart disease may be overused, we doubt that the lower utilization rates observed consistently among black patients reflect an effort to provide more appropriate care to these patients. We also doubt that our study will undermine trust between physicians and patients. On the contrary, with luck, our study will foster a more honest dialogue between physicians and patients and encourage the medical profession to seek ways to eliminate unconscious bias that may influence physicians' clinical decisions.

Helft et al. raise a question about the potential effects of physicians' characteristics, especially sex, on our findings. Male and female physicians had similar rates of referral for cardiac catheterization (87.3 and 88.5 percent, respectively; P=0.64). Furthermore, as we mentioned in the report, adding physicians' characteristics to our multivariate logistic-regression model did not change the results of the main analysis.

Table 1Table 1Odds Ratios for Referral for Cardiac Catheterization with the Use of a Logistic-Regression Model in Which Race, Sex, and Age Were Specified as Interactions. shows the odds ratios for referral for catheterization with the use of a logistic-regression model in which race, sex, and age were specified as interactions. The original grant proposal did not include hypotheses based on a three-way interaction of race, sex, and age. Furthermore, a joint test of the three-way interaction was not statistically significant.

With regard to Davidoff's comments, we were unable to assess the contribution of certain factors to physicians' referrals for cardiac catheterization, including whether the hospitals where the physicians practiced offered the procedures. However, Hannan et al.2 found racial disparities in the treatment of coronary artery disease, even after accounting for medical appropriateness and necessity, availability of procedures, and patients' refusals. Davidoff also questions the clinical validity of the referral rates reported in our study. We emphasize that the physicians classified patients' chest pain before knowing the results of the thallium exercise test. In contrast, their recommendations regarding referral for catheterization were elicited after they learned that the results of the exercise test were positive.

Kevin A. Schulman, M.D.
Georgetown University Medical Center, Washington, DC 20007

Jesse A. Berlin, Ph.D.
University of Pennsylvania School of Medicine, Philadelphia, PA 19104-6021

José J. Escarce, M.D., Ph.D.
RAND Health Program, Santa Monica, CA 90401

2 References

1. 1

   Schwartz LM, Woloshin S, Welch HG. Misunderstandings about the effects of race and sex on physicians' referrals for cardiac catheterization. N Engl J Med 1999;341:279-283
   Full Text | Web of Science | Medline

2. 2

   Hannan EL, van Ryn M, Burke J, et al. Access to coronary artery bypass surgery by race/ethnicity and gender among patients who are appropriate for surgery. Med Care 1999;37:68-77
   CrossRef | Web of Science | Medline

Author/Editor Response

The editors reply:

We take responsibility for the media's overinterpretation of the article by Schulman and colleagues. We should not have allowed the use of odds ratios in the Abstract. As Schwartz et al. point out, risk ratios would have been clearer to readers. Also, it is unfortunate that the data in Table 1 of the reply by Schulman et al. were removed from the original manuscript in response to reviewers' concerns about subgroup analysis. These data show that the conclusions depended largely on the response to the 70-year-old black actress and, to a lesser extent, on the response to the 55-year-old black actress. This important point should have been made more explicit in the article. As Davidoff points out, although racism and sexism are prevalent in American life, the evidence of racism and sexism in this study was overstated.

Gregory D. Curfman, M.D.
Jerome P. Kassirer, M.D.

Citing Articles (4)

Citing Articles

1. 1

   Hillel A Steiner, John M Miller. (2008) Disparity in utilization of implantable cardioverter-defibrillators in treatment of heart failure based on sex and race. Women's Health 4:1, 23-25
   CrossRef

2. 2

   A. Russell Localio, David J. Margolis, Jesse A. Berlin. (2007) Relative risks and confidence intervals were easily computed indirectly from multivariable logistic regression. Journal of Clinical Epidemiology 60:9, 874-882
   CrossRef

3. 3

   Lavera M. Crawley. (2002) Palliative Care in African American Communities. Journal of Palliative Medicine 5:5, 775-779
   CrossRef

4. 4

   Schwartz, Lisa M., Woloshin, Steven, Welch, H. Gilbert, . (1999) Misunderstandings about the Effects of Race and Sex on Physicians' Referrals for Cardiac Catheterization. New England Journal of Medicine 341:4, 279-283
   Full Text