Development of a risk score for choledocholithiasis in pediatric patients

Purpose The prevalence of gallstones in children has increased over the last years. Choledocholithiasis (CD) is present in up to 30% of the cases. There is a scarcity of studies on the management of choledocholithiasis in children. The aim of this study was to develop a score that would allow predicting accurately the risk of CD in children with gallstones and reduce the number of non-therapeutic ERCP. Materials and methods We conducted a retrospective study in children with gallstones and suspected CD seen between January 2010 and December 2019. The main outcome was the presence of CD confirmed by at least one of the following diagnostic tests: magnetic resonance cholangiopancreatography (MRCP), endoscopic retrograde cholangiopancreatography (ERCP), and/or intraoperative cholangiography (IOC). We developed a risk score based on the presence or absence of the following risk factors: acute biliary pancreatitis, ascending cholangitis, elevated liver function tests (AST, ALT, total bilirubin [TB, >= 2 mg/dl], conjugated bilirubin, gamma-glutamyl transpeptidase, and alkaline phosphatase), CD on ultrasound (US; this was considered predictive but not confirmatory of CD), and dilation of the common bile duct (> 6 mm) by US. The score was divided into three different categories: low risk (no risk factors), intermediate risk (one risk factor present), high risk (>= 2 risk factors present or ascending cholangitis). Given the main goal of reducing the number of diagnostic ERCPs, a very-high-risk subgroup (3 risk factors present or ascending cholangitis) was identified. Results We reviewed 133 patients with gallstones and suspected CD. In 56 (42.1%) patients, the presence of CD was confirmed by one or more of the definitive diagnostic tests (MRCP, ERCP, and IOC). The following variables were found to be the strongest predictors of CD: ascending cholangitis, TB >= 2 mg/dl, common bile duct > 6 mm, and the presence of CD by US. The positive predictive value for CD was 7.5% in the low-risk group (OR 0.06, P = < 0.001); 22.9% in the intermediate-risk group (OR 0.31, P = 0.007); 77.6% in the high-risk group (OR 20.14, P = < 0.001); and 95.7% in very-high-risk subgroup (OR 49.18, P = < 0.001). Conclusion The risk score proposed in this study predicts accurately the presence of CD in children with gallstones. It can serve as a helpful tool to triage the need for costly and complex studies in the workup of CD, particularly in centers with limited resources. Finally, due to its high specificity and positive predictive value (PPV), the use of the very-high-risk criteria would allow for an important decrease in the number of non-therapeutic ERCP.