Risk Stratification in Post-ERCP Pancreatitis: How Do Procedures, Patient Characteristics and Clinical Indicators Influence Outcomes?

Abstract

Background. Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) remains common, and severe complications are associated with ERCP. There is no previous study detailing the effect of race and gender in a US-based population on risk of PEP. Methods. Data were collected on 269 “first-performed” consecutive ERCPs followed by division by race (White vs. African-American) and sex (Female vs. Male). A total of 53 probable risk factors were evaluated by uni- and multivariate analysis followed by outcomes expressed as an odds ratio (OR) (with a 95% confidence interval, 95% CI). Finally, a principal component analysis was performed to construct a risk prediction model for PEP, which can be used by clinicians at bedside. Results. After analyzing the risk factors based on race and gender-based groups, Caucasian males with PEP are more likely to have prior history of pancreatitis (p = 0.009), lower hemoglobin (p = 0.02)/blood urea nitrogen (BUN) (p = 0.01)/creatinine before ERCP (p = 0.07) and lower BUN (p = 0.01)/creatinine after ERCP (p = 0.07), while Caucasian females with PEP are more likely to have higher white blood cell (WBC) count before ERCP (p = 0.08) and lower amylase (p = 0.10)/bilirubin (p = 0.09)/aspartate aminotransferase (AST) after ERCP (p = 0.08). African-American males with PEP are more likely to have lower weight (p = 0.001)/smaller height (p = 0.0005)/lower alkaline phosphatase (p = 0.002)/AST (p = 0.04)/alanine transaminase (ALT) (p = 0.03) before ERCP and lower alkaline phosphatase (p = 0.002)/AST (p = 0.01)/ALT (p = 0.004) after ERCP, while African-American females with PEP are more likely to have prior history of pancreatitis (p = 0.004)/higher lipase before (p = 0.0001) and after (p = 0.05) ERCP along with increased risk with pancreatic duct cannulation (p = 0.0001) and injection (p = 0.0001)/biliary sphincterotomy (p = 0.0001). Importantly, prior history of ERCP, elevated AST after ERCP, and BUN prior to ERCP were found to be important clinical features predicting post-ERCP pancreatitis. To our knowledge, this is a first known attempt at developing a risk scoring system for PEP in a US population with decision tree learning. Conclusions. It is very evident that both patient and procedure-related risk factors vary by race and gender in the US population, leading to the development of a new risk assessment tool for PEP that can be used in clinical practice. We need to follow up with a larger prospective study to validate this novel race and gender-based risk scoring system for PEP.