© Georg Thieme Verlag KG Stuttgart · New York
04 December 2007 (eFirst)
Changing utilization of ERCP
In a cohort study of > 400 000 patients who underwent endoscopic retrograde cholangiopancreatography (ERCP) in the USA between 1988 and 2002 , Mazen Jamal et al. showed that the overall number of ERCPs decreased by 20 % between 1996 and 1998. This sudden reduction occurred after a period of linear increase (+ 200 % in 7 years), and was followed by a stabilization. Interestingly, a steady increase in the absolute numbers of procedures was recorded over the period 1988 - 2002 for two categories of ERCP, namely ERCP with a therapeutic goal (+ 200 %) and those carried out in patients older than 40 years (+ 60 %). The most evident explanations for these variations include a wider access to technologies that compete with diagnostic ERCP, and the introduction of laparoscopic cholecystectomy with a subsequent increase in the number of cholecystectomies (extraction of common bile duct [CBD] stones was the main indication for therapeutic ERCP) .
This study is echoed by another nationwide US study which covered > 2500 hospitals and about 200 000 ERCPs performed between 1998 and 2001. This showed that the median number of inpatient ERCPs/year in US hospitals was 49, with only 5 % of hospitals performing > 200 ERCPs/year. Compared with low-volume centers, high-volume centers disclosed shorter hospital stays (6.9 vs. 7.8 days, respectively; P < 0.0001), and lower failure rates (4.7 vs. 6.0 % with a “dose-response” effect); however, actual rates were underestimated because failures were not specifically recorded in the database analyzed .
The utilization of ERCP is low with currently less than one procedure performed per week in the majority of centers, and it is expected to further decrease because, among other reasons, diagnostic procedures still accounted for 40 % of all ERCPs performed during the most recent period studied by Mazen Jamal et al. Over the same period, the complexity of ERCPs has increased. This combination of factors is potentially detrimental, and it will further intensify the challenge of training in this procedure, as has already been illustrated by a prospective audit of > 5000 ERCPs performed in the UK . This showed that: (i) only 49 % of the units engaged in training had the minimum workload recommended for this activity (250 ERCPs/year); (ii) all trainees succeeded in < 80 % procedures; and (iii) even among trained endoscopists, only 77 % were successful in ≥ 80 % procedures (i. e. the minimum level expected from an adequately trained endoscopist). This study also suggested that the threshold which had been recommended for a trainee to achieve competence in ERCP (i. e. 200 supervised procedures) was no longer applicable; trainees with an experience of > 200 ERCPs before audit initiation were successful in only 66 % of procedures. Using another approach (recording consecutive procedures performed by a single operator), another study also suggested that the threshold number of procedures required to achieve competence in ERCP was higher than previously recommended . The criterion of success used in this study was defined as “deep cannulation of the desired duct in patients with a native papillary anatomy.” In our area of therapeutic ERCP, this criterion is certainly better suited to define competence than those used in previous studies, which included for example deep cannulation in patients with a previous papillary sphincterotomy or simple pancreatography if this was the sole aim of the procedure. This study showed that 350 supervised procedures were required to achieve deep duct cannulation in > 80 % of cases.
It is concluded that there are too many trainees in too many low-volume centers, and that this situation is likely to worsen. Endoscopists will hopefully initiate the necessary changes to improve this situation before third-party payers or policy makers attempt to impose regulations. The range of potential changes is wide (e. g. organization of training at a regional rather than an institutional level, concentration of ERCPs in fewer centers with only a few dedicated endoscopists performing this procedure, technical improvement, and increased availability of simulators), but none of these is easy to undertake.