The benefit of indomethacin suppositories for prophylaxis against post-ERCP pancreatitis (PEP) in high-risk patients was established in a landmark trial published in 2012. The aims of this study were to measure the adoption of indomethacin prophylaxis in widespread clinical practice, evaluate concurrent trends in pancreatic duct (PD) stent utilization, and estimate the impact of these changes on PEP in a high-risk population.
Data were extracted from a commercial database (Explorys, IBM Watson Health, Somers, NY) that aggregates electronic health records from 26 US healthcare systems from 2009 to 2018. Using Systematized Nomenclature of Medicine Clinical Terms, we identified a cohort of patients who underwent endoscopic retrograde cholangiopancreatography (ERCP) and were at high risk for PEP based on narrow criteria. PEP was defined as an emergency department or hospital admission 1–5 days after ERCP with an associated diagnosis of pancreatitis.
Twenty six thousand eight hundred twenty ERCPs were performed on this high-risk cohort from 2009 to 2018. The overall PEP rate during the study period was 8.6%. There was no decrease in PEP rates from 2012 to 2018. Beginning in 2012, indomethacin usage increased linearly (P < 0.001), but remained below 50% in 2018. As indomethacin increased, utilization of PD stents declined abruptly from 2013 to 2014 (40.7%–8.5%) and trended to a nadir of 3.0%.
Despite its low cost, widespread availability, and level I evidence of benefit in reducing the risk of PEP in high-risk patients, the adoption of rectal indomethacin during ERCP has been slow and the medication continues to be under-utilized. In parallel, the PD stent usage has declined dramatically. The lack of change in PEP rates during the study period could be attributable to the persistent low usage of rectal indomethacin or the decline in PD stent use. Further educational efforts and quality assurance measures are warranted to ensure that rectal indomethacin and PD stent placement are more appropriately used in clinical practice.