Darwin and Gender Bias in Cardiac Care
abstract & commentary
Synopsis: Gender comparisons of acute MI mortality rates in women in clinical trials may be artificially inflated in women because of natural selection.
Source: Rieves D, et al. Am J Cardiol 2000;85:147-153.
Gender differences in outcome following myocardial infarction (MI) have been attributed to many factors: a treatment bias against women, the older age of most women with MI, more frequent comorbidities in women with MI, and that women have smaller coronary arteries which are less amenable to acute interventions. Rieves and colleagues postulated that since the sudden death rate in men is higher than in women, using hospitalized patients in clinical studies would include more high-risk women who had not died before hospitalization. To test this hypothesis, they evaluated the GUSTO-1 and INJECT databases for time-to-death using regression modeling. Baseline differences in clinical characteristics between the men and women in these trials were found as expected. Baseline characteristics and in-hospital medication use were used as covariants in the analyses. The unadjusted 30-day mortality rates in women were twice those of men (odds ratio, 2.2-2.3). When adjusted for age, the ratio was 1.4-1.5, but the inclusion of multiple other covariants did not affect the model significantly. In GUSTO, where death was recorded in one-hour increments, time-to-death in men was earlier than in women and a disproportionate number of early deaths occurred in younger men (< 65 years). Almost half of the deaths in these trials occurred in the first 24 hours and men died an average of 1.7 hours earlier than women in GUSTO. In INJECT, where deaths were recorded in one-day increments, no gender differences were noted. These data confirm the hypothesis that the pattern of earlier death in men in thrombolytic trials may be an extension of a gender-specific pattern that began before hospitalization and resulted in proportionately more alive, but high-risk women in hospitalization-based trials. Thus, Rieves et al conclude that gender comparisons of acute MI mortality rates in women in clinical trials may be artificially inflated in women because of natural selection.
Comment by Michael H. Crawford, MD
This analysis of the gender-specific mortality in GUSTO and INJECT confirmed the higher mortality in women even after adjustment for age, comorbidities, and medications. The earlier assertion that women with MI are not as aggressively treated cannot be a factor in these controlled clinical trials, and I believe it has largely been dispelled by the more rigorous analyses of subsequent data. Although this is a retrospective analysis of mainly the GUSTO-1 database involving extrapolation to the prehospital phase, the results are provocative and suggest another potential explanation for the higher observed mortality of women with acute MI: higher risk men may have died before hospitalization, leaving a relatively higher proportion of high-risk women in the study population.
Rieves et al considered whether there was a delay in presentation and treatment in women because some have claimed that women do not present with the same symptoms as men and thus either delay themselves or are misdiagnosed when initially triaged. However, time to presentation and treatment was not different by age or gender in these two trials. This agrees with our own data in men and women in three ethnic groups (Anglo, Hispanic, and Native American), where we observed no difference in presenting symptoms among these groups. Of course, they could not exclude the possibility that some women may not have presented at all, but that would be true for men as well.
Other information tends to support their claim that men may have died suddenly before reaching the hospital in greater numbers than women. Although cause of death was not recorded in GUSTO, more men underwent defibrillation than women (43% vs 32%). Also, other studies of MI survivors have shown a higher incidence of inducible ventricular arrhythmias in men as compared to women. Of interest, the incidence of cardiogenic shock was not different between the two genders in GUSTO. These data support the concept that men are more prone to sudden early death, but not later deaths due to pump failure.
The higher death rate of women with acute myocardial infarction in clinical trials is likely due to:
a. inadequate treatment of women.
b. prehospital death of high-risk men.
c. atypical symptoms in women.
d. more cardiogenic shock in women.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content