The Bedside Index for Severity in Acute Pancreatitis: a systematic review of prospective studies to determine predictive performance

Background: predicting the development of severe disease has remained a major challenge in management of acute pancreatitis. The Bedside Index for Severity in Acute Pancreatitis (BISAP) is easy to calculate from the data available in the first 24 hours. Here, we performed a systematic review to determine the prognostic accuracy of the BISAP for severe acute pancreatitis (SAP). Methods: major databases of biomedical publications were searched during the first week of October 2015. Two independent reviewers searched records in two phases. Studies that reported prognostic accuracy of the BISAP for SAP from prospective cohorts were included. The pooled area under the receiver operating curve (AUC) was calculated. Results: Twelve studies were included for data-synthesis and methodology quality assessment was performed for 10. All the studies had enrolled consecutive patients, had a broad spectrum of the disease severity, reported explicit interpretation of the predictor, outcome of interest was well defined and had adequate follow-up. Blinded outcome assessment was reported in only one study. The pooled AUC was 0.85 (95% CI 0.80-0.90). There was significant heterogeneity, I² 86.6%. Studies using revised Atlanta classification in defining SAP had a pooled AUC of 0.92 (95% CI, 0.90-0.95), but heterogeneity persisted, I² 67%. Subgroup analysis based on rate of SAP (>20% vs <20%) did not eliminate the heterogeneity. Conclusion: the BISAP has very good predictive performance for SAP across different patient population and etiologies. Studies to evaluate the impact of incorporating the BISAP into clinical practice to improve outcome in acute pancreatitis are needed before adoption could be advocated with confidence.