Analysis
Univariate analysis using chi-square and t-tests were performed that compared sex with mortality, with each procedure, and with each comorbidity. The relationship between the patient’s sex and each of the 8 outcomes of interest (adjusted for age, race, insurance smoking, hypertension, diabetes, and hypercholesterolemia) was investigated by logistic regression analysis for dichotomous variables and survival analysis for time to event variables. The significance of each analysis was set at P = .01, based on the Bonferonni adjustment9 for multiple comparisons and an overall P = .05. Analysis was performed using STATA (STATA Corporation, College Station, Tex.) and SAS (SAS Institute, Cary, N.C.) statistical software. We estimated that a sample of 1600 patients was needed to detect an absolute difference of 6% in the presence or absence of an intervention between men and women (two-tailed alpha = .05, beta = 0.20).
Results
A total of 1669 patients (631 women, 1038 men) were available for our analysis. Baseline characteristics by sex are displayed in Table 1. Men were significantly younger, less likely to be African American, less likely to be Medicaid insured, more likely to smoke, and less likely to have diabetes mellitus (P < .05) than women. In the univariate analysis (Table 2), women had significantly higher rates of hospital mortality (P < .01) and diabetes (P < .01) and a longer mean length of stay in the hospital (P = .01). Men had significantly higher rates of stent placement (P < .01) and CABG (P < .01).
We found no significant difference between men and women for hospital mortality, time in the ICU, total time in the hospital, stent placement, angiogram, angioplasty, or the 3 catheterization procedures combined in the multivariate analysis (Table 3). Men had significantly more CABG (relative risk [RR] 1.96, 95% confidence interval [CI] 1.41-2.76) than women.
In a separate analysis of patients who underwent CABG (n = 211), men had significantly more 3-vessel coronary disease and advanced left anterior descending artery disease (LAD >50%) than women (Table 4). There was no difference between men and women undergoing CABG for either single-vessel or double-vessel coronary artery disease. The extent of coronary artery disease was only known for patients who were catheterized (N = 1204). Again comparing sex regarding the risk of CABG, but additionally controlling for the extent of coronary artery disease (LAD >50% and 3-vessel CAD), now reveals no significant increase associated with male sex (RR 1.30, 95% CI 0.82-2.08).
TABLE 1
CHARACTERISTICS OF THE STUDY POPULATION
| Men (N = 1038) | Women (N = 631) | P Value | |
|---|---|---|---|
| Age | 62.2 + 13.3 | 68.7 + 13.8 | < .05 |
| Race | < .05 | ||
| White | 789 (76%) | 482 (76%) | |
| Black | 62 (6%) | 67 (11%) | |
| Asian | 3 (0%) | 1 (0%) | |
| Other | 184 (18%) | 82 (13%) | |
| Insurance | < .05 | ||
| Medicaid | 14 (1%) | 25 4%) | |
| Comorbidities | |||
| Hypercholesterolemia | 224 (22%) | 108 (17%) | NS |
| Hypertension | 444 (43%) | 293 (46%) | NS |
| Smoking | 411 (40%) | 184 (29%) | < .01 |
| Diabetes | 267 (26%) | 209 (33%) | < .01 |
TABLE 2
UNADJUSTED OUTCOMES BY SEX
| Men (n = 1038) | Women (n = 631) | P Value | |
|---|---|---|---|
| Hospital mortality | 76 (7%) | 70 (11%) | < .01 |
| Mean time in ICU | 2.1 days | 1.9 days | NS |
| Mean length of stay | 5.9 days | 6.6 days | < .01 |
| Angiogram | 241 (23%) | 139 (22%) | NS |
| Angioplasty | 67 (6%) | 38 (6%) | NS |
| Stent placement | 346 (33%) | 162 (26%) | < .01 |
| CABG | 157 (15%) | 54 (9%) | < .01 |
| 3 catheterization procedures (angiogram, angioplasty, stent) | 654 (63%) | 339 (54%) | < .01 |
| CABG denotes coronary artery bypass grafting; ICU, intensive-care unit. | |||
TABLE 3
ADJUSTED RELATIVE RISK FOR CARDIOVASCULAR OUTCOME
| Outcome | Relative Risk* |
|---|---|
| Hospital mortality | 0.86 (0.6-1.23) |
| Time in ICU | 0.95 (0.85-1.05) |
| Angiogram | 1.02 (0.80-1.31) |
| Angioplasty | 0.90 (0.59-1.39) |
| Stent placement | 1.04 (0.82-1.32) |
| Coronary artery bypass graft | 1.96 (1.41-2.76)† |
| Angiogram, angioplasty, or stent | 1.04 (0.82-1.28) |
| * Men compared with women (ie, men who underwent coronary artery bypass grafting were more likely to have more extensive disease). | |
| † P < 0.05. | |
| ICU denotes intensive-care unit. | |
TABLE 4
EXTENT OF CORONARY ARTERY DISEASE IN PATIENTS UNDERGOING CABG, BY PATIENT’S SEX (N = 211)
| Extent of Disease | Relative Risk(95% CI)* |
|---|---|
| Single vessel | 0.79 (0.29-1.19) |
| Two vessel | 0.86 (0.45-1.21) |
| More than 2 vessels | 1.44 (1.10-1.88) † |
| Left anterior descending artery > 50% | 1.58 (1.14-2.04) † |
| * Men compared with women (ie, men who underwent CABG were more likely to have extensive disease than were women who underwent CABG). | |
| † P < .05. | |
| CABG denotes coronary artery bypass grafting. | |
Discussion
In our study, men had significantly higher rates of bypass surgery and all procedures combined, as has been found in previous studies.10-12 Age was the greatest confounder for the mortality outcome. Mortality rates were significantly higher in women with all confounding variables in the logistic model except age. The close similarity between the mortality outcome in this study and the findings of Vaccarino et al1 may be explained by the considerably smaller sample size of the current investigation. Alternatively, this similarity may reflect greater recognition of sex disparities and changing practice patterns since those studies were published.
The increased adjusted risk of bypass surgery and of all procedures is explained in part by the anatomic differences in coronary artery disease as found in our study and in others.3,14 Men undergoing CABG had significantly more 3-vessel and advanced left main disease than women. In our data, controlling for the extent of coronary artery disease eliminated any sex bias. We are limited, however, by not having data on all men and women who had an acute MI and by knowing only the coronary anatomy of those undergoing coronary catheterization. Future research should address this question. Because the prevalence of diabetes is higher in women, they may have more generalized coronary artery disease that is less amenable to bypass surgery and angioplasty, as was the case in the GUSTO IIb trial.3