An assessment of existing risk stratification guidelines for the evaluation of patients with suspected choledocholithiasis.

BACKGROUND: Prior studies have demonstrated that existing risk stratification guidelines for the evaluation of suspected choledocholithiasis lack accuracy, leading to the overutilization of endoscopic retrograde cholangiopancreatography (ERCP). The aim of our study was to evaluate the performance characteristics of published guidelines in predicting choledocholithiasis and to determine the impact of laboratory trends on diagnostic accuracy.

METHODS: We identified patients with suspected choledocholithiasis hospitalized over a 5-year period (2009-2014) at a tertiary care academic medical center. Among eligible patients, we assessed the performance characteristics of the American Society for Gastrointestinal Endoscopy (ASGE) guidelines predicting the presence of choledocholithiasis, confirmed by endoscopic ultrasound, magnetic resonance cholangiography, ERCP, or intra-operative cholangiography. We also evaluated whether a second set of liver function tests improved the accuracy of the guidelines.

RESULTS: On presentation, 71 of the 173 eligible patients (41.4 %) met ASGE high-probability criteria for choledocholithiasis. Of these, only 39 (54.9 %) were found to have a choledocholithiasis on confirmatory testing. Conversely, of the 102 patients (58.6 %) who were classified as low or intermediate probability, 32 (31.4 %) had choledocholithiasis. Overall, the accuracy of the guidelines was 63 % (sensitivity 54.9 %; specificity 68.6 %). Incorporating a second set of laboratory tests did not improve accuracy (62.7 %), and a significant decline in liver function tests did not reliably predict spontaneous stone passage.

CONCLUSIONS: Existing guidelines performed suboptimally for predicting choledocholithiasis in our patient population, similar to other validation studies. These findings further underscore the importance of developing alternate risk stratification tools for choledocholithiasis, aiming to minimize unnecessary diagnostic ERCP.