Emergency laparotomy preoperative risk assessment tool performance: A systematic review.

Background: Preoperative assessment of risk for emergency laparotomy may enhance decision making with regards to urgency or perioperative critical care admission and promote a more informed consent process for patients. Accordingly, we aimed to assess the performance of risk assessment tools in predicting mortality after emergency laparotomy. Methods: PubMed, Embase, the Cochrane Library and CINAHL were searched to 12 February 2022 for observational studies reporting expected mortality based on a preoperative risk assessment and actual mortality after emergency laparotomy. Study screening, data extraction, and risk of bias using the Downs and Black checklist were performed in duplicate. Data on setting, operation undertaken, expected and actual mortality rates were extracted. Meta-analysis was planned but not possible due to heterogeneity. This study is registered with PROSPERO, CRD42022299227. Results: From 10,168 records, 82 observational studies were included. 17 risk assessment tools were described, the most common of which were P-POSSUM (42 studies), POSSUM (13 studies), NELA (12 studies) and MPI (11 studies). Articles were published between 1990 and 2022 with the most common country of origin being the UK (33 studies) followed by India (11 studies). Meta-analysis was not possible. Observed mortality and expected mortality based on risk assessment is reported for each study and generally shows most studies show accurate risk prediction. Conclusions: This review synthesises available literature to characterise the performance of various risk assessment tools in predicting mortality after emergency laparotomy. Findings from this study may benefit those undertaking emergency laparotomy and future research in risk prediction.