Citation: International Journal of Surgery. 2013, 11(8), 724-725
Author: Farrell A.; Stephenson R.
Abstract: Introduction/Aim: NHS requirements for transparency and accountability in surgical outcomes require accurate and fair means of measuring performance. Crude mortality/morbidity rates may discriminate against surgeons operating on more complex cases. Risk prediction models such as POSSUM may enable post-operative outcomes to be viewed objectively. This audit aimed to determine whether POSSUM/P-POSSUM accurately predicts mortality/morbidity in patients undergoing major urological procedures. Methods: We reviewed POSSUM/P-POSSUM scores of 110 patients undergoing major urological surgery under a single surgeon. Observed and predicted post-operative complications and mortality were compared, measuring significance of differences using Chi-square analysis. Results: The mean P-POSSUM 30 day mortality risk was 4.86%, predicting 5 deaths throughout the sample. Observed mortality was 0(p = 0.0177). The mean POSSUM predicted morbidity risk was 50.96%, meaning we would have expected to observe 56 cases in which complications occurred. Observed morbidity was 43 (39%)(p = 0.0127). Conclusion: Although predictions of both mortality and morbidity were significantly over-estimated, the observed morbidity was proportional to stratified P-POSSUM predictions. Whilst further studies will be required to confirm or refute these findings more generally, the POSSUM and P-POSSUM predicted values
could have value as benchmarks to allow a fairer comparison of performance/safety between urological surgeons/ units.
