To the Editor.
—After examining community-wide data, O'Connor et al1 and Williams et al2 reported that coronary artery bypass graft (CABG) surgery has a perioperative mortality of 4% and 8%, respectively. These mortality estimates approximate those from other reports of community-wide studies3 but are twofold higher than the mortality reported in studies of CABG surgery done in referral centers.4,5 At least two other surgical procedures for which communitybased mortality data are available show a similar pattern. One study of carotid endarterectomy showed a perioperative mortality of 3.4%,6 far exceeding the 0.6% mortality rate reported in a recent multicenter clinical trial of the efficacy of carotid endarterectomy.7 Similarly, in a community-wide study of prostatectomy, Wennberg et al8 observed a short-term mortality rate of 3.7%, in contrast to the accepted short-term mortality rate of 1.0%.9A number of biases or distortions could account for
McDonald CJ, Fitzgerald JF. CABG Surgical Mortality in Different Centers. JAMA. 1992;267(7):932-933. doi:10.1001/jama.1992.03480070048022