44 Trends in ERCP Utilization for Sphincter of Oddi Dysfunction: The Real World Impact of EPISOD and Rome IV

Abstract

INTRODUCTION: The widespread clinical adaptation of landmark clinical trials and practice guidelines is often unclear. The EPISOD trial, a randomized, sham-controlled trial, demonstrated that endoscopic sphincterotomy in patients with Type III Sphincter of Oddi dysfunction (SOD) did not result in clinical improvement. The results of this study were presented at the 2013 National Meeting of the American College of Gastroenterology and published in May 2014. Subsequently, the updated 2016 Rome IV criteria removed Type III SOD from the functional biliary disorder classification. Though patients with Type III SOD make up a minority of all SOD patients, we aimed to assess the utilization of ERCP for patients with SOD in a large population database from 2012–2017. METHODS: We searched electronic health record (EHR)-based disparate data from a cloud-based dataset which incorporates over 300 individual hospitals in the United States (Explorys Inc, Cleveland, OH). Using Systematized Nomenclature of Medicine (SNOMED) codes we identified patients with a first-ever diagnosis of “disorder of Sphincter of Oddi” or “spasm of Sphincter of Oddi” annually from 2012–2017. Because the dataset uses SNOMED codes, it was not feasible to sub-classify the types of SOD. Stratified by year, we identified the proportion of patients undergoing ERCP with sphincterotomy and those receiving tricyclic antidepressants (TCA). A Mantel-Haenszel test was used. RESULTS: The number of index cases of SOD and rates of ERCP utilization are outlined in Table 1. From 2013 (the presentation of EPISOD data) to 2016, there was a significant trend demonstrating lower rates of ERCP for newly-diagnosed SOD (P = 0.018). A divergent increase in TCA utilization was concomitantly seen from 2013–2015, but the trend was not statistically significant (P = 0.29). Following the publication of the ROME IV criteria, rates of ERCP utilization for newly-diagnosed SOD significantly increased (P = 0.048) (Figure 1). CONCLUSION: Following the presentation of the EPISOD data, there was a significant 3-year linear decrease in ERCP utilization for all-comers newly-diagnosed with SOD. Concomitantly, a divergent yet insignificant increase in TCA usage was observed. In 2016, after the Rome IV criteria narrowed the diagnosis of SOD, an expected increase in ERCP utilization was observed. These findings call for ongoing observation to see if these trends are durable.