Mortality rate prediction by Physiological and Operative Severity Score for the Enumeration of Mortality and Morbidity (POSSUM), Portsmouth POSSUM and Colorectal POSSUM and the development of new scoring systems in Chinese colorectal cancer patients.

BACKGROUND: The aim of this study was to compare the Physiological and Operative Severity Score for the enumeration of Mortality and Morbidity (POSSUM), Portsmouth POSSUM (P-POSSUM), and Colorectal POSSUM (Cr-POSSUM) for predicting surgical mortality in Chinese colorectal cancer patients and to create new scoring systems to achieve better prediction.
METHODS: Data from 903 patients undergoing surgery for colon and rectal cancers from 1992 to 2005 at Peking University Third Hospital were included in this study. POSSUM, P-POSSUM, and Cr-POSSUM were used to predict mortality. Stepwise logistic regression was used to develop the modified P-POSSUM and Cr-POSSUM. Their performances were tested by receiver operating characteristic curve, Hosmer-Lemeshow statistic, and observed:expected ratio.
RESULTS: The actual inpatient mortality was 1.0% (9 of 903). The predicted mortality of POSSUM, P-POSSUM, and Cr-POSSUM were 5.6%, 2.8%, and 4.8%, respectively, which were significantly higher than the actual mortality in our cohort. The predicted mortality of the modified P-POSSUM and Cr-POSSUM was very close to the observed mortality. Both the modified models offered better accuracy than P-POSSUM.
CONCLUSIONS: The predicted mortality of POSSUM, P-POSSUM, and Cr-POSSUM were significantly higher than the observed mortality in our patients. The modified P-POSSUM and Cr-POSSUM models provided an accurate prediction of inpatient mortality rate in colorectal cancer patients in China.