Linking processes and outcomes to improve surgical performance: a new approach to morbidity and mortality peer review

Peer review of surgical cases resulting in death or potentially avoidable complications is a longstanding tradition, but intensive reviews of individual cases rarely produces tangible improvements in clinical outcomes. The systematic comparison of care received by patients who experienced adverse outcomes to care received by patients who had uneventful surgical courses is a promising alternative to intensive case review. However, the results of these studies may be misleading because physicians often choose interventions based on their perceptions of patients' preoperative risks, and higher adverse outcome rates among patients with higher preoperative risk may distort comparisons of alternative interventions. The creation of a control sample by carefully matching each patient who experienced an adverse outcome to a patient who had a similar preoperative risk but did not experience an adverse outcome can overcome this problem and provide excellent insight into how to improve clinical performance most effectively and efficiently. By using currently available electronic data to compute each patient's risk of an adverse outcome, a series of cases with adverse outcomes can be matched to an equal number of controls. Peer review committees can then direct focused data collection and analyses of potentially critical processes of care to determine which, if any, are associated with significantly poorer clinical results. A simulated scenario is analyzed in detail to illustrate the ability of this technique to correctly determine best practices when other approaches fail to do so.