A Comparison of POSSUM and GPS Models in the Prediction of Post-operative Outcome in Patients Undergoing Oesophago-gastric Cancer Resection

Abstract

There is some evidence that a patient's pre-operative condition influences short-term and long-term post-operative outcomes. The aim of the present study is to compare the physiological and operative severity score for the enumeration of mortality and morbidity (POSSUM) and modified Glasgow prognostic score (mGPS) models in the prediction of post-operative outcome, both short term and long term, in patients undergoing resection of oesophago-gastric cancer. Patients who underwent curative resection for oesophago-gastric cancer from January 2005 to May 2009 and who had data to score the POSSUM, P-POSSUM, O-POSSUM and mGPS models were included in the study. Observed morbidity and mortality rates were compared with predicted outcome in different risk groups. Both short-term outcome and long-term survival were recorded. Observed morbidity was 49%, whereas POSSUM predicted post-operative morbidity in 60%, giving an overall standardised morbidity ratio of 0.82. Only male sex [hazard ratio (HR) 3.61, 95% confidence interval (CI) 1.38-9.46, P = 0.009] and POSSUM physiology score (HR 2.13, 95% CI 1.11-4.08, P = 0.023) were independently associated with post-operative morbidity. The post-operative mortality rates predicted by POSSUM, P-POSSUM and O-POSSUM were 16.5, 5.8 and 9.9%, respectively, giving a standardised mortality ratio of 0.25, 0.71 and 0.42. Only mGPS (HR 1.96, 95% CI 1.09-3.54, P = 0.025) and tumour-node-metastasis (TNM) stage (HR 2.21, 95% CI 1.44-3.38, P < 0.001) were independently associated with cancer-specific survival. The POSSUM physiology score was useful in predicting post-operative morbidity, and the mGPS was useful in predicting cancer-specific survival, in patients undergoing surgery for oesophago-gastric cancer.