Evaluating the Ability of the Bedside Index for Severity of Acute Pancreatitis Score to Predict Severe Acute Pancreatitis: A Meta-Analysis

Objective: To evaluate the diagnostic performance of the bedside index for severity in acute pancreatitis (BISAP) score in predicting severe acute pancreatitis (SAP). Materials and Methods: A systematic search was conducted using PubMed, Cochrane library and EMBASE databases up to May 2014, and 9 related studies, including 1,972 subjects, were reviewed. Pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnosis of odds ratio (DOR) and hierarchic summary receiver-operating characteristic (HSROC) curves, as well as the area under the HSROC curve (AUC), were assessed using the HSROC and bivariate mixed effects models. Moreover, a subgroup analysis stratified by cutoff value was performed to measure the effect of the diagnostic threshold on the performance of the BISAP score. Finally, publication bias was assessed using Deeks’ funnel plot asymmetry test. Statistical analyses were performed using the STATA 12.0 software. Results: The pooled sensitivity, specificity, PLR, NLR and DOR of the BISAP for predicting SAP were 64.82% (95% CI: 54.47– 73.74%), 83.62% (95% CI: 70.03–91.77%), 3.96 (95% CI: 2.27– 6.89), 0.42 (95% CI: 0.34–0.52) and 9.41 (95% CI: 5.38–16.45), respectively. The AUC was 0.77 (95% CI: 0.73–0.80). Moreover, the subgroup analysis results demonstrated that the BISAP cutoff point at 3 had a higher specificity and greater accuracy than at 2 to predict SAP. No significant publication bias was detected across the studies (p = 0.359). Conclusion: The BISAP score showed low sensitivity but high specificity for assessing the severity of acute pancreatitis.