Ros Jacklin1, Nick Sevdalis, Clare Harries, Ara Darzi, Charles Vincent. 1. Clinical Safety Research Unit, Department of Bio-Surgery and Surgical Technology, Imperial College London, 10th Floor, QEQM, St. Mary's Hospital, South Wharf Rd., London, W2 1NY, UK.
Abstract
BACKGROUND: Surgical judgment and decision making require valid methods of assessment. The aim of this study was to evaluate the use of judgment analysis as a technique for quantitative evaluation of surgeons' risk estimates. METHODS: Thirty trainee surgeons' estimates of conversion risk in laparoscopic cholecystectomy were investigated using judgment analysis. Hypothetical cases were created, differing in relevant risk factors. Twenty repeat cases were incorporated to test for reliability. Surgeons' estimates were compared with an outcome-derived gold standard from the published literature. RESULTS: The mean reliability was .77 (range, .47-.98), and regression models indicating the weighting of variables had a mean adjusted R(2) value of .53 (range, .12-.76). Variables were subject to wide variation in weighting. The mean correlation to the gold standard model was .48 (range, .08-.72). CONCLUSIONS: Judgment analysis allows detailed quantitative evaluation of the consistency of surgeons' risk estimates and the influence of different variables on them. Comparison with a gold standard model enables accuracy to be measured.
BACKGROUND: Surgical judgment and decision making require valid methods of assessment. The aim of this study was to evaluate the use of judgment analysis as a technique for quantitative evaluation of surgeons' risk estimates. METHODS: Thirty trainee surgeons' estimates of conversion risk in laparoscopic cholecystectomy were investigated using judgment analysis. Hypothetical cases were created, differing in relevant risk factors. Twenty repeat cases were incorporated to test for reliability. Surgeons' estimates were compared with an outcome-derived gold standard from the published literature. RESULTS: The mean reliability was .77 (range, .47-.98), and regression models indicating the weighting of variables had a mean adjusted R(2) value of .53 (range, .12-.76). Variables were subject to wide variation in weighting. The mean correlation to the gold standard model was .48 (range, .08-.72). CONCLUSIONS: Judgment analysis allows detailed quantitative evaluation of the consistency of surgeons' risk estimates and the influence of different variables on them. Comparison with a gold standard model enables accuracy to be measured.