Abstract
This study investigated the ability of the Physiological and Operative Severity Score for Enumeration of Mortality and Morbidity (POSSUM) scoring system to predict postoperative morbidity (complication rate) and compared the ability of POSSUM and four other scoring systems (Portsmouth POSSUM [p-POSSUM], colorectal POSSUM [cr-POSSUM], Association of Coloproctology of Great Britain and Ireland [ACPGBI] and Acute Physiology and Chronic Health Evaluation II [APACHE II]) to predict mortality within 30 days in 1695 patients undergoing surgery for colorectal cancer. Receiver operating characteristic (ROC) curve, Student's t-test and the χ2-test were used to estimate the predictive ability of these scoring systems. The observed complication rate of 38.7% was not significantly different to the rate of 36.3% predicted by the POSSUM scoring system (observed: expected [O: E] ratio 1.07). The observed mortality rate was 3.0%. For predicting mortality, POSSUM had an O: E ratio of 0.37, compared with p-POSSUM O: E ratio 1.00, cr-POSSUM O: E ratio 0.91, APACHE II O: E ratio 0.31 and ACPGBI O: E ratio 1.41. It was concluded that the POSSUM scoring system had high value for predicting the risk of morbidity following colorectal cancer resection. For predicting postoperative mortality, p-POSSUM, cr-POSSUM and ACPGBI were superior to POSSUM and APACHE II, however ROC curve analysis showed that cr-POSSUM and ACPGI discriminated best between survivors and non-survivors, so were more accurate predictors of postoperative mortality than the other three scoring systems.