The use of the Thoracic Morbidity and Mortality system for the internal analysis of performance: a case-matched temporal audit.

OBJECTIVES The aim of the present study was to evaluate the usefulness of the Thoracic Morbidity and Mortality (TMM) scoring system in auditing the quality of care of our unit. METHODS We analysed the performance of our unit comparing the incidence of complications and mortality occurring after anatomic lung resections during two different periods: early period (January 2000 to December 2009: 830 lobectomy, 134 pneumonectomy and 78 segmentectomy) and recent period (January 2010 to August 2012: 191 lobectomy, 8 pneumonectomy and 19 segmentectomy). The cardiopulmonary complications as traditionally defined in the European Society of Thoracic Surgeons (ESTS) database were also classified according to the TMM system: this method grades the postoperative adverse events from 1 to 5 reflecting an increasing severity of management regardless the type of complication. Complications graded higher than 2 are regarded as major complications. To account for confounders, several baseline and surgical factors were used to build a propensity score that was applied to match the patients of the most recent group with their early-group counterparts. These two matched groups were compared in terms of cardiopulmonary morbidity (codified by ESTS definitions) and mortality rates and incidence of major complications according to the TMM system. RESULTS The propensity score analysis yielded 209 well-matched pairs of patients operated on in the two periods. The two groups had similar rates of ESTS-defined cardiopulmonary complications (recent: 38 patients vs early: 37 patients, P = 0.9). The use of the TMM system revealed a higher incidence of major (grade > 2) complications rate in the recent period (recent: 29 patients vs early: 14 patients, P = 0.02). CONCLUSIONS The TMM scoring system for classifying the postoperative complications revealed a decline of quality of care of our unit otherwise undetected by applying traditional outcome measures. This tool can be used as an additional graded outcome endpoint to refine internal audit of performance.