Reply to Letter: “Cost-Effective Decisions in Managing Silent Common Bile Duct Stones Should Include all the Management Options to Help Decision Makers”
We thank Date et al1 for their interest and comments of our analysis on cost-effective strategies in detecting silent common bile duct (CBD) stones during laparoscopic cholecystectomy (LC). The authors raised important points regarding the applicability of our conclusion that under the circumstances of most general surgery practices, intraoperative ultrasound (IOUS) would be the test of choice for most patients with asymptomatic choledocholithiasis undergoing cholecystectomy.
Our study extends prior work in which it included a wide range of study parameters—that is, preoperative probability of CBD stones, cost of IOUS, cost of intraoperative cholangiogram (IOC), and so on, making it applicable to a variable of practice settings. In addition, efficacy was assessed in terms of utilities [Quality adjusted life years (QALYs)]. Our analysis indicates that the most cost-effective strategy may not be the cheapest option but one that yields an incremental cost per QALY below the decision-makers willingness to pay. In the base case, the probability of CBD stones was 9%, based on a meta-analysis of the literature. We believe this base case reflects the circumstances of general surgery practice and in it included a sensitivity analysis that varied the prevalence of CBD stones from 0 to 100% when holding other model parameters the same. This analysis showed that for patients whose risk of CBD stones was < 3% on preoperative evaluation including patient history, physical examination, laboratory values, and radiologic findings, expectant management (EM) is the most cost-effective choice. Date et al cited the study by Collins et al2 who evaluated the prevalence of CBD stones in patients undergoing LC for symptomatic cholelithiasis. The prevalence of CBD stones in their patient population was 3.4% (4.6% of patients had filling defects on IOC, but 26% of those were deemed false positives). Approximately, 38.9% of those patients (2.2% patients of the total study population) had persistent stones that necessitated Endoscopic retrograde cholangiopancreatography (ERCP). The study concluded that in their population, EM was acceptable when compared with IOC, which is consistent with our conclusion that in patients whose probability of CBD stone is <21% EM would be cost-effective compared with IOC, but neither is as cost-effective as IOUS, which was a strategy not discussed in the Collins study.
Intraoperative laparoscopic [Laparoscopic common bile duct exploration (LCBDE)] and endoscopic (ERCP) management of gallstones have both emerged as useful pathways for streamlining management in high-volume centers. We recognize that inclusion of these modalities would increase the comprehensiveness of the analysis, but also its complexity. For both arms of detection, there would be outcomes and uncertainties associated with two principal intraoperative (LCBDE and ERCP) and one principal postoperative management strategies for stones (ERCP). In addition, intraoperative management of clinically silent gallstones is not widely practiced outside of leading institutions and requires considerable investment in resources and team training, and sufficient volumes to maintain proficiency. The small likelihood of detecting clinically silent stones in general practices argues against intraoperative management becoming widely disseminated in lower or even intermediate volume hospitals where most cholecystectomies are performed. For these reasons, we felt inclusion of LCBDE and ERCP in the analysis was outside the scope of our study, which is directed at intraoperative detection in general surgical practices.
We agree with Dr. Date's comment that most general surgeons are not trained in IOUS. However, it has been reported that the learning curve for proficiency in IOUS is relatively brief and that when proficiency is achieved its sensitivity, specificity, and rate of successful CBD visualization is higher than IOC.