Dr King and colleagues1 reported that
women had a higher risk of death than men immediately following coronary artery
bypass graft (CABG) surgery, but that the risk of death was equal for men
and women after 1 year. The authors used graphic analysis rather than Cox
regression because the results violated the proportional hazard assumptions
of the Cox model. We are concerned that the results of this analysis reflect
residual confounding due to the reported baseline differences in demographic
factors and comorbid conditions, rather than an independent effect of sex.
For instance, Koch et al2 reported that
early mortality after CABG surgery was higher in women but that this difference
disappeared after propensity matching for demographic factors, risk factors,
medications, coronary disease, and cardiopulmonary bypass.
Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ, Shah A. Sex Differences in Mortality After Coronary Artery Bypass Graft Surgery. JAMA. 2004;292(1):40-41. doi:10.1001/jama.292.1.40-c