Evaluation of surgical outcome scores according to ISGPS definitions in patients undergoing pancreatic resection

Abstract

Aims: To evaluate the role of the Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity (POSSUM), the Portsmouth variant (p-POSSUM) and the Glasgow Prognostic Score (GPS) in predicting outcome after pancreatic surgery with reference to the International Study Group of Pancreatic Surgery (ISGPS) definitions of post-pancreatectomy complications.

Methods: All consecutive patients undergoing major pancreatic resection over a 32- month period were included. POSSUM, p-POSSUM and GPS score were calculated for each patient and correlated against the observed mortality and morbidity using the ISGPS definitions.

Results: The observed:expected ratios for POSSUM mortality, POSSUM morbidity and p-POSSUM mortality were 0.24 (p < 0.0001), 0.86 (p < 0.0001) and 0.79 (p = 0.09), respectively. POSSUM had a 'poor fit' with respect to predicting morbidity (χ(2) = 16.4, 8 d.f., p = 0.04). Multifactorial regression analysis revealed the GPS as an independent predictor of post-operative outcome (GPS 1, p = 0.03, OR 2.99, 95% CI 1.4-7.9, and GPS 2, p = 0.02, OR 4.3, 95% CI 1.8-15.5).

Conclusion: POSSUM has a limited role as an outcome score in pancreatic resection. The GPS may be a novel alternative to POSSUM as a pre-operative predictor of outcome.