SP7.7 Evaluation of predictive models of common bile duct stones and the use of biliary investigations

Abstract

Abstract Aims This study identifies clinical, biochemical and radiological factors that may predict CBD stones prior to cholecystectomy to reduce costs and delays of non-indicated biliary imaging. Method Retrospective data of patients underwent cholecystectomy over 6 months was collected, including first presentations, liver function tests (LFTs), and imaging. Data was tested against guidelines from European Society of Gastrointestinal Endoscopy (ESGE, 2019), American Society of Gastrointestinal Endoscopy (ASGE, 2019) and literature models (Gurusamy-2015, Isherwood-2014). Results 322 included patients had LFTs and US scans in first presentations with symptomatic gall stones. 52 of them had repeated biliary imaging (US, MRCP or CT) without showing any new findings in 92% of cases. 118 patients had MRCP prior to cholecystectomy, with only 22 cases of confirmed CBD stones. After applying the above-mentioned guidelines and predictive values, new criteria, combining cut-off values of 30 µmol/L for bilirubin and 180 IU/L for Alkaline phosphatase, showed the highest sensitivity (72.73%) and specificity (81.25%), while maintaining high positive predictive value (PPV) (46.99%) and accuracy (79.66%). (Table 1) Conclusion Using clinical and radiological findings to guide the need for definitive biliary imaging, commonly MRCP, can reduce unnecessary radiological investigations and costs. Our new suggested combined cut-off values of Bilirubin and ALP can complement the clinical judgement.