Centralizing coronary artery bypass grafting surgery

Kalant and Shrier discuss findings regarding low-volume surgical procedures in a recent report by the Canadian Institute for Health Information.1 We agree with their position that health care policymakers must base their decisions on careful review of the entire body of evidence.

In evaluating research findings, we believe that rigorous, systematic lit- erature reviews are important. Well- established methods exist for conducting such reviews (e.g., the QUORUM statement2 referenced by major journals). Use of similarly rigorous standards to synthesize the research literature minimizes the possibility of random or systematic error biasing conclusions.

In Health Care in Canada 2001, we cite several early articles in the field plus the only published recent broad systematic review of the literature on the relationship between volume and mortality. This review, published in JAMA in 2000,3 summarized research on the relationship between hospital caseload and outcomes for 40 conditions based on findings from 72 studies that were evaluated for scientific merit and relevance according to explicit, pre-determined criteria. Eleven published studies on coronary artery bypass graft surgery (CABG) met these criteria. All showed better outcomes with higher volumes; the difference was statistically significant in 9 of them.

In Canada, the number of both rare and common surgical procedures currently performed by individual hospitals varies, often significantly. Some types of care are becoming concentrated over time. For example, as of 1998/99, no Canadian hospital performing CABG surgery had less than 200 cases per year, down from 5 hospitals in 1996/97.1

Volume-outcome relationships are clearly an area of current clinical and policy interest. Indeed, the recent Sinclair inquest found that “the limited number of cases [of pediatric cardiac surgery] that can be undertaken in a province like Manitoba with a population of just over 1 million increases the risk of morbidity and mortality.”4

Based on current evidence, we stand by our original conclusion that “deciding how much to centralize care requires us to strike a balance across [a variety of] issues. This balance is likely to vary from procedure to procedure and place to place.” In this context, systematic reviews of the research literature, an understanding of current Canadian volume patterns, and better information about patient outcomes at individual hospitals can all provide evidence to support decisions about how best to organize health services and distribute health care resources.

Richard Alvarez President and CEO Canadian Institute for Health Information Toronto, Ont. Charlyn Black Director, Centre for Health Services and Policy Research University of British Columbia Vancouver, B.C.