Back to Annual Meeting Program

Background:
Few formal cost effectiveness analyses simultaneously evaluate radiographic, endoscopic, and surgical approaches in the management of choledocholithiasis.

Methods:
Using the decision analytic software TreeAge, we modeled the initial clinical management of a patient presenting with symptomatic cholelithiasis without overt signs of choledocholithiasis. In this base case we assumed a 10% probability of concurrent asymptomatic choledocholithiasis. Our model evaluated four diagnostic/therapeutic strategies: universal MRCP, universal ERCP, laparoscopic cholecystectomy (LC), or laparoscopic cholecystectomy with universal intraoperative cholangiogram (LCIOC). All probabilities were estimated from review of published literature. Procedure and intervention costs were equated with Medicare reimbursements. Costs of hospitalizations were derived from median hospitalization reimbursement for New York State using diagnosis-related groups (DRG). Sensitivity analyses were performed on all cost and probability variables.

Results:
The most effective strategy in the diagnosis and management of symptomatic cholelithiasis with a 10% risk of asymptomatic choledocholithiasis was LCIOC. This was followed closely by MRCP, LC alone, and ERCP; expected values of success in each strategy did not differ in a clinically meaningful way. Varying the prevalence of asymptomatic choledocholithaisis or the probability that retained stones would eventually cause symptomatic biliary obstruction did not affect the optimal strategy. When procedure and hospitalization costs were taken into consideration, LCIOC was the most cost effective approach, followed by laparoscopic cholecystectomy. LC was preferred when the prevalence of asymptomatic choledocholithiasis fell below 9%, or when the probability that a retained CBD stone would eventually become symptomatic was less than 60%. Similarly, if the sensitivity, specificity, or technical success of an IOC fell below 78%, 54%, or 80%, LC alone was the preferred strategy. Universal MRCP and ERCP were both more costly and less effective than surgical strategies, even at high probability of asymptomatic choledocholithiasis. Within the tested range for both procedural and hospitalization related costs for any of the surgical or endoscopic interventions, LCIOC and LC were always more cost effective than universal MRCP or ERCP, irrespective of presence or absence of complications. Varying the cost, sensitivity, and specificity of MRCP had no effect on this outcome.

Conclusions:
LC with routine IOC is the preferred strategy in a cost effectiveness analysis of the management of symptomatic cholelithiasis with asymptomatic choledocholithiasis. MRCP was both more costly and less effective under all tested scenarios.