Abstract
OBJECTIVES: In the context of comparisons of surgical outcomes, risk adjustment is the retrospective adjustment of a provider's or a surgeon's results for case mix and/or hospital volume. It allows accurate, meaningful inter-provider comparison. It is therefore an essential component of any audit and quality improvement process. The aim of this study was to review the literature to identify those factors known to affect prognosis in hepatobiliary and pancreatic cancer surgery. METHODS: PubMed was used to identify studies assessing risk in patients undergoing resection surgery, rather than bypass surgery, for hepatobiliary and pancreatic cancer. RESULTS: In total, 63 and 68 papers, pertaining to 24 609 and 63 654 patients who underwent hepatic or pancreatic resection for malignancy, respectively, were identified. Overall, 22 generic preoperative factors predicting outcome on multivariate analysis, including demographics, blood results, preoperative biliary drainage and co-morbidities, were identified, with tumour characteristics proving disease-specific factors. Operative duration, transfusion, operative extent, vascular resection and additional intra-abdominal procedures were also found to be predictive of early outcome. CONCLUSIONS: The development of a risk adjustment model will allow for the identification of those factors with most influence on early outcome and will thus identify potential targets for preoperative optimization and allow for the development of a multicentre risk prediction model.