Prediction of postoperative mortality and morbidity in octogenarians with gastric cancer - Comparison of P-POSSUM, O-POSSUM, and E-POSSUM: A retrospective single-center cohort study.

BACKGROUND: The purpose of this study was to evaluate various POSSUM scoring systems in predicting postoperative morbidity and mortality in elderly patients with gastric cancer. METHODS: A total of 1,262 patients with gastric cancer who underwent curative gastrectomy between January 2006 and December 2013 were retrospectively reviewed. The subjects were stratified by age into /= 80 years old. To assess the predictability and efficacy of various POSSUM scores (POSSUM, P-POSSUM, O-POSSUM, and E-POSSUM), the observed-to-expected (O:E) ratio and area under the receiver operating characteristic curve (AUC) were calculated and compared with actual postoperative morbidity and mortality. RESULTS: Among the 1,262 patients, 75 were elderly (>/=80 years old). The observed mortality rates were 0.5% (n=6) in the whole cohort, and 4.0% (n=3) in elderly patients. The predicted mortalities of POSSUM, P-POSSUM, E-POSSUM, and O-POSSUM for elderly patients were 13.2%, 5.3%, 5.7%, and 21.8%, respectively (O:E ratio = 0.3, 0.75, 0.7, and 0.18, respectively). P-POSSUM and E-POSSUM showed superior discriminatory power compared to POSSUM and O-POSSUM. In terms of morbidity, E-POSSUM showed better predictive capabilities than POSSUM in elderly patients (O:E ratio = 0.56 and 0.74, respectively). CONCLUSIONS: All POSSUM scoring systems tend to overestimate postoperative mortality and morbidity in gastric cancer patients. E-POSSUM and P-POSSUM provided a better prediction of mortality and morbidity after curative gastrectomy in elderly patients compared to other POSSUM scores.