Suboptimal Specificity of Asge High-Risk Group in Evaluation of Patients with Suspected Choledocholithiasis
Huiqin He,Chenfei Tan,Jiaguo Wu,Ning Dai,Weiling Hu,James Scheiman,Loren Laine,John J. Kim
DOI: https://doi.org/10.14309/00000434-201510001-00018
2015-01-01
The American Journal of Gastroenterology
Abstract:Introduction: Patients who are considered high-risk for choledocholithiasis after biochemical testing and abdominal ultrasound (AUS) are recommended to undergo endoscopic retrograde cholangiopancreatography (ERCP) with others to receive less invasive tests. The aim of the study is to determine whether high-risk patients by ASGE guidelines or other factors can accurately identify those with choledocholithiasis.Table 1: Diagnostic Performance of ASGE High-Risk Group and Other Test Characteristics in Patients with Suspected Choledocholithiasis ( N =2,726)Methods: Consecutive patients hospitalized (1/2011-12/2013) with suspected choledocholithiasis who received biochemical testing, AUS and definitive testing for choledocholithiasis (magnetic resonance cholangiopancreatography (MRCP), endoscopic ultrasound (EUS), ERCP, interventional radiology procedure and/or intraoperative cholangiogram (IOC)) at Sir Run Run Shaw Hospital were retrospectively identified. Patients with choledocholithiasis on AUS, bilirubin >4mg/dL, bilirubin >1.8mg/dL + dilated common bile duct (CBD) (>8mm and >6mm with or without prior cholecystectomy) and/or suspected cholangitis (presenting fever and/or leukocytosis) were considered high-risk per ASGE guidelines. Final diagnosis was based on choledocholithiasis identified on definitive testing. Results: Mean age of 2,726 patients with suspected choledocholithiasis was 55+15 yrs, 1,179 (43%) were male and 216 (8%) had prior cholecystectomy. On definitive testing (MRCP in 2,441 (90%), EUS in 69 (3%), interventional radiology procedure in 2, ERCP in 643 (23%), IOC in 189 (7%)), 1,076 (39%) patients had choledocholithiasis. A higher proportion of patients meeting ASGE high-risk criteria compared to other groups (770/1,271 (61%) vs. 306/1,455 (21%); OR=5.8, 95%CI 4.9-6.8) had confirmed choledocholithiasis. Diagnostic performance of test characteristics is shown (Table). On multivariate analysis, choledocholithiasis on AUS (AOR=17.4, 95%CI 12.6-23.9), bilirubin >4mg/dL (AOR=3.0, 95%CI 2.2-4.1) and bilirubin 1.8-4mg/dL (AOR=1.8, 95%CI 1.4-2.4) compared to < 1.8mg/dL and CBD dilation (AOR=1.8, 95%CI 1.5-2.2) increased the likelihood of choledocholithiasis. Conclusion: In patients with suspected choledocholithiasis, the specificity of high-risk group categorized by ASGE guidelines was suboptimal. Combination of factors that include choledocholithiasis on AUS and bilirubin >4mg/dL with or without CBD dilation had high specificities of >90% and moderate to high positive likelihood ratios. A revised algorithm with more stringent criteria may improve patient selection and reduce the risk of diagnostic ERCP.