Centralizing coronary artery bypass grafting surgery

In its 2001 annual report on Canada's health care system, the Canadian Institute for Health Information discussed inverse relationships between the number of surgical procedures and the incidence of adverse post-operative events, for surgery in general and for coronary artery bypass grafting in particular.1 The Institute's president was quoted as saying “we have too many centres undertaking … heart surgery. The result may be unnecessary complications and … death.”2 Similarly, the chair of the Montreal regional health board stated “it's proven beyond a reasonable doubt that patient outcomes are better” at high-volume centres.2 A Quebec task force on tertiary cardiology recommended that centres each perform a minimum of 400 to 450 operations annually to maintain the quality of the service, without reference to supporting evidence.3 The policy implication of these statements is clear.

Does the evidence warrant such a degree of certainty? The report from the Canadian Institute for Health Information provides 3 references to support its position: 2 original research studies from the 1980s4^,5 and a review article that provides details on only 1 of 10 studies reviewed.6 We are aware of 8 recent cohort studies that were not cited. Six of these reported mortality odds ratios for low- versus high-volume centres (low-volume centres treated fewer than 225 cases), and 5 had ratios very close to 1, which indicates an absence of a demonstrable volume–mortality relationship. For the other 2 cohorts, multiple regression analysis showed no volume effect on mortality after controlling for other factors. The odds ratios from early and recent cohorts together showed a linear progression over time, from –0.45 to 1. The most conservative conclusion at this time is that the available evidence does not provide a basis for a policy decision to centralize coronary artery bypass grafting surgery. It should be noted that these studies are concerned with hospital outcomes, not with those of individual surgeons.

We believe that there are 2 lessons to be learned. First, even highly regarded institutions may be subject to error or bias in presenting information. Sec- ond, bandwagons, buzzwords and self-interest often influence our thinking. Health care policymakers must avoid these effects and must base their decisions on careful review of the evidence, just as physicians are urged to do.

Norman Kalant Department of Medicine Sir Mortimer B. Davis – Jewish General Hospital Montreal, Que. Ian Shrier Centre for Epidemiology and Community Studies Sir Mortimer B. Davis – Jewish General Hospital Montreal, Que.