Size of common bile duct stones on MRCP predicts likelihood of positive findings at ERCP.

OBJECTIVE To ascertain if direct measurement of the size of common bile duct stones (CBD) on magnetic resonance cholangio-pancreatography (MRCP) can be used to predict the likelihood of a positive endoscopic retrograde cholangio-pancreatography (ERCP) result. Should we be carrying out ERCPs on all patients with CBD stones? METHODS We retrospectively analysed the records of 1812 consecutive patients undergoing MRCP between November 2009 and November 2014 at the Victoria Infirmary. A measurable CBD stone was present in 383 patients, of whom 293 successfully underwent ERCP. RESULTS 221 patients (75%) had stones demonstrated on ERCP. A receiver operating curve (ROC) was plotted correlating stone size with the likelihood of a positive ERCP result, and demonstrates that using a cut off of >4 mm as an indication for ERCP gives the mathematical best-fit correlation with a sensitivity of 83% (95% CI 78-88) and specificity of 66% (95% CI 53-77). CONCLUSIONS In our current practice, all patients with CBD stones at MRCP are considered for ERCP regardless of stone size. Our results would support the hypothesis that as size decreases the likelihood of spontaneous stone passage increases. Although the threshold of mathematical best compromise is >4 mm to minimise both type 1 and type 2 errors, we would favour a lower threshold of 2 mm above which ERCP is performed (sensitivity 98.65, 95% CI 96.1-99.7, specificity 25.71%, CI 16.0-37.6). For patients with stones measuring 2 mm or less, early operative intervention with intraoperative cholangiography to confirm duct clearance could be a suitable alternative.