Development of a risk scoring system for the prediction of acute pancreatitis after ERCP: results of a prospective multicenter study

Aims Acute pancreatitis represents the most frequently reported adverse event following ERCP. Identifying patients at higher risk may help to prevent the occurrence of this complication. Aim of this study was to investigate the incidence and factors associated to post-ERCP pancreatitis (PEP) and to provide a risk scoring system.

Methods Cross-sectional prospective study was conducted on adult patients undergoing ERCP among 16 Italian hospitals between May 2019 and April 2022. We collected detailed sociodemographic, clinical and endoscope-related information, including pre-, intra- and post-exam variables.The main outcome was PEP. A multivariable logistic regression model was derived based on (i) the results of univariable logistic regression analyses, including as adjustments those variables being significant at p-value<0.05, and (ii) variables with biological plausibility. Therefore, the variables sex, age and study center were included a priori, because of their relevance. To describe the risk of PEP qualitatively and quantitatively in this population, we built a score (0-8 points) assigning a weight (w) to the principal pre-procedural factors found to be associated with this outcome, namely age<70 (w=1), female sex (w=1.5), non-supine position during the exam (w=2) and naïve papilla (w=3.5). We established a 4-level score with low risk being 0-2.9, 3-4.9 low-medium risk, 5-6.9 medium-high risk, and 7-8 high risk.

Results A total of 1961 consecutive patients were enrolled, and 157 cases of PEP (8%; 95% CI 6.8-9.2) were observed. The analysis included 1673 individuals, after excluding missing values. PEP was significantly associated to female sex (OR=1.67, 95% CI=1.16-2.40), age>=70 (OR=0.66, 95% CI=0.45-0.95), supine position (OR=0.46, 95% CI=0.23-0.93), history of acute pancreatitis (OR=2.45, 95% CI=1.42-4.24), naïve papilla (OR=3.18, 95% CI=1.68-6.04), Wirsung cannulation (OR=1.90, 95% CI=1.25-2.88) and rescue method of cannulation (OR=1.84, 95% CI=1.14-2.95). The largest proportion of the participants were at medium-high risk (39.6%), followed by low-medium (24.9%), high (22.4%) and low-risk (13.1%). When considering the risk score, we found a 50% increasing risk of PEP by 1 level of increase in the score (OR=1.51,95% CI=1.19-1.93). Considering low risk level as reference (predicted incidence=4.0%;), each increased risk score level predicted the risk of PEP as follows: low-medium risk OR=1.97 (95% CI=0.83-4.62) and predicted incidence of 6.5%; medium-high risk OR=3.10, (95% CI=1.35-7.10), predicted incidence=10.7%, and, finally, high-risk OR=4.07 (95% CI=1.69-9.83), predicted incidence=12.8%.

Conclusions In this multicenter, large, prospective study, we identified factors associated with the risk of PEP and derived a risk scoring system based on pre-procedural factors.This scoring system may help the clinician to correctly stratify the risk of the procedure and adequately inform the patient.

Conflicts of interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
15 April 2024