ABSTRACT 52

CUMULATIVE SUM FAILURE LEARNING CURVE ANALYSIS OF A POLICY CHANGE FROM ON PUMP TO OFF PUMP CORONARY ARTERY BYPASS GRAFTING.

Richard Novick, Stephanie Fox, Larry Stitt, Stuart Swinamer, Kris Lehnhardt, Bob Kiaii, Reiza Rayman, Mackenzie Quantz, John Lee, Alan Menkis, Neil McKenzie, Douglas Boyd. London Health Sciences Centre, London, Canada.

PURPOSE: Use of the sequential probability cumulative sum (CUSUM) technique may be more sensitive than standard statistical analyses in detecting surgical failures (J Cardiac Surg 1999;14:312-20). We applied CUSUM methods to evaluate a policy change by a single surgeon from routine on pump (CPB) to off pump CABG (OPCAB).

METHODS: 55 consecutive CABG patients (CPB group) were compared to the next 55 patients undergoing an attempt at routine OPCAB using the same coronary stabilizer. Preoperative risk was calculated via validated multivariable models. The occurrence of mortality and 8 major complications (MI, bleeding, stroke, IABP use, mediastinitis, respiratory failure, life-threatening arrhythmia and sepsis) was compared between the 2 groups using Fisher’s exact and t-tests as well as CUSUM methods.

RESULTS: Results of standard analyses are shown in the table. The CPB and OPCAB groups had similar preoperative risk and number of grafts per patient. 46/55 patients (84%) after the policy change were completely revascularized off pump; 9 were converted electively to CPB due to an intramyocardial LAD (2) or initial hemodynamic instability (7). On CUSUM analysis, the failure curve in CPB patients approached the upper 80% alert line after 8 cases; in OPCAB patients it reached below the lower 80% (reassurance) boundary after 28 cases, indicating superior results in the latter group.

CONCLUSIONS: A policy change from CABG on CPB to routinely attempting OPCAB can be accomplished safely. CUSUM analysis was more sensitive than standard statistical methods in detecting surgical failures.