OBJECTIVES: In recent years, there has been an increasing interest in monitoring the quality of cardiac surgical performance. The aim of the present study was to apply control charts (CUSUM curves) to monitor the performance of minimally invasive mitral valve procedures to enhance quality control for that operation. METHODS: A total of 936 minimally invasive mitral valve procedures were performed from September 2003 to March 2011 by seven surgeons (range 26-401 procedures) at a single institution. Institutional and individual surgeons' performances were monitored using descriptive statistics and control charts, with a predetermined acceptable failure rate of 10% and calculated 80% alert and 95% alarm lines. Perioperative death or one or more of seven adverse events constituted failure. RESULTS: The incidence of in-hospital mortality was 1.8% (17/936) and compared favourably with the predicted mortality (logistic EuroSCORE 7.3%). Institutional CUSUM analysis revealed an initial learning curve and then the surgical process remained in control for all the study period. There were differences between surgeons with regard to the learning curves and perioperative complications (7.3-11.3%, P = 0.9). Five surgeons crossed the 95% reassurance boundary between operations 23 and 48. One surgeon crossed the 95% reassurance boundary after 116 operations. No surgeon crossed the 95% alarm line, which indicates unacceptably high failure rates. CONCLUSIONS: Minimally invasive mitral surgery can be safely performed with low morbidity and mortality. CUSUM curve analysis is a simple statistical method to implement continuous individual and departmental performance monitoring.
OBJECTIVES: In recent years, there has been an increasing interest in monitoring the quality of cardiac surgical performance. The aim of the present study was to apply control charts (CUSUM curves) to monitor the performance of minimally invasive mitral valve procedures to enhance quality control for that operation. METHODS: A total of 936 minimally invasive mitral valve procedures were performed from September 2003 to March 2011 by seven surgeons (range 26-401 procedures) at a single institution. Institutional and individual surgeons' performances were monitored using descriptive statistics and control charts, with a predetermined acceptable failure rate of 10% and calculated 80% alert and 95% alarm lines. Perioperative death or one or more of seven adverse events constituted failure. RESULTS: The incidence of in-hospital mortality was 1.8% (17/936) and compared favourably with the predicted mortality (logistic EuroSCORE 7.3%). Institutional CUSUM analysis revealed an initial learning curve and then the surgical process remained in control for all the study period. There were differences between surgeons with regard to the learning curves and perioperative complications (7.3-11.3%, P = 0.9). Five surgeons crossed the 95% reassurance boundary between operations 23 and 48. One surgeon crossed the 95% reassurance boundary after 116 operations. No surgeon crossed the 95% alarm line, which indicates unacceptably high failure rates. CONCLUSIONS: Minimally invasive mitral surgery can be safely performed with low morbidity and mortality. CUSUM curve analysis is a simple statistical method to implement continuous individual and departmental performance monitoring.