Individual and practice differences among physicians who perform ERCP at varying frequency: A national survey

Background: ERCP practice patterns in the United States are largely unknown. Objective: To characterize the ERCP practice of U.S. gastroenterologists, stratified by their annual case volume: high volume (HV, >200), moderate volume (MV, 50-200), and low volume (LV, <50). Design: Anonymous electronic survey. Subjects: American Society for Gastrointestinal Endoscopy members who are practicing U.S. gastroenterologists. Results: Among all responders (N = 1006), 63% were derived from community practices. Physicians who performed ERCPs and provided data on annual volume (n = 669) were classified as LV (n = 254), MV (n = 284), and HV (n = 131). During training, 77% of LV physicians did not complete 180 ERCPs compared with 58% of MV and 34% of HV physicians (P < .0001). Only 58% of LV physicians enjoy performing ERCP compared with 88% of MV and 98% of HV physicians (P < .0001); 60% reported being "very comfortable" with ERCP compared with more than 90% of MV and HV physicians (P < .0001). LV physicians are less comfortable with pancreatic duct stenting (PDS) (57% vs 92% [MV] and 98% [HV], P ≤ .02) and using prophylactic PDS. Although HV physicians (42%) were least likely to use short-wire devices (P < .02), use of wire-guided cannulation was similar (74% LV, 72% MV, 66% HV, P = .13). Thirty-seven percent of LV physicians reported comfort with needle-knife sphincterotomy compared with 75% (MV) and 99% (HV) (P < .0001). Limitations: Survey completion rate of 18.5%. Conclusions: Self-reported comfort and/or enjoyment with ERCP is lower among LV physicians. Wire-guided cannulation is used by the majority of all ERCP practitioners, but prophylactic PDS is less frequently used by LV physicians. Because many LV physicians perform ERCP for higher-grade indications and use advanced techniques (eg, needle-knife sphincterotomy), further LV physician ERCP outcomes data are needed.