Training Level (July Effect) and the Risk of Post-ERCP Pancreatitis: A Nationwide Analysis of More Than Half a Million ERCP Procedures

Abstract

Introduction: Post ERCP Pancreatitis (PEP) is the most common complication following ERCP procedure. There is conflicting evidence about whether trainee involvement at teaching hospitals during ERCP influences of the risk of PEP. Our aim was to analyze the variation in the rate of PEP by month and compare that across teaching and non-teaching hospitals. Methods: We used the Nationwide Inpatient Sample (NIS) for the year 2000-2014. Adult patients, who were admitted with biliary obstruction without acute pancreatitis based on the record’s primary and secondary diagnosis and had an inpatient ERCP, were identified using ICD9 codes and included in the analysis. PEP was defined as having a subsequent diagnosis of acute pancreatitis after the ERCP. ERCP was classified into interventional or non-interventional based on the presence of specific ICD9 procedure codes. Monthly rates of PEP were calculated and then stratified by hospital location and teaching status. PEP rates were compared using Chi square analysis. Results: A total of 654,394 adult patients were included. Median age was 59 years and 66% were females. 49% of patients were admitted to urban-teaching hospitals, 45% to urban-nonteaching, and 7% to rural hospitals. The overall rate of PEP was 5.4%. PEP rates varied significantly across hospital types, with rural having the lowest rate of PEP at 4.2%, followed by urban-teaching at 4.8% and the highest rate in urban-nonteaching at 6.2% (p<0.001). No statistically significant difference was observed in the monthly rate of PEP in teaching hospitals (Figure 1). ERCP with intervention was performed in 89% of the cases. Analysis of the monthly PEP rates of interventional ERCP also showed no significant difference for the period of July-December as compared to January-June (Figure 2).56_A Figure 1. Monthly rate of PEP across different hospital location and types.56_B Figure 2. Rate of PEP following interventional ERCP by academic year across different hospital types and locations.Conclusion: In this large nationwide study, there was no evidence for the presence of the ‘July effect’. The rate of PEP did not change in teaching hospitals throughout the academic year suggesting that the presence of in-training fellows is not a risk factor for PEP. Contrary to what was previously thought, our study showed that teaching hospitals had in fact a lower rate of PEP than nonteaching ones. This might be due to the use of pre-procedural evidence-based methodology, which in turn may lead to better patient selection for ERCPs. Rural hospitals had a slightly lower PEP rate likely due to less complex interventions being performed in such localities.