Economical and clinical outcomes of alternative treatment strategies in the management of common bile duct stones in the elderly: wait and see or surgery?

BACKGROUND: Common bile duct stones (CBDS) are especially prevalent in the elderly population. Although the standard of care for stone removal is endoscopic retrograde cholangiography with sphincterotomy (ERC-S), the clinician's decision to refer a patient for cholecystectomy after ERC-S depends on several factors including potential for future biliary symptoms and complications, morbidity and mortality related to cholecystectomy, and costs associated with referral for cholecystectomy versus conservative approach. Using decision analysis, we explored the economic implications of cholecystectomy versus expectant management following ERC-S in elderly patients with CBDS. MATERIALS AND METHOD: A decision tree was constructed with DATA 3.5 (Williamstown, MA) to estimate the costs and outcomes associated with two treatment strategies following ERC-S for CBDS in patients age 60 yr and older: (1) elective cholecystectomy, and (2) expectant management. Probabilities for potential complications and outcomes were derived from the medical literature and cost reflected Medicare reimbursement rates at our institution. The time horizon of the analysis was 2 yr. RESULTS: Elective cholecystectomy was associated with total costs of $5,259 with 94.3% of the cohort alive (1.886 life-years) at 2 yr, whereas expectant management was associated with total costs of $1,173 with 94.7% of the cohort alive (1.894 life-years). The results were sensitive to the probability of recurrent biliary symptoms in patients treated conservatively. Compared to elective cholecystectomy, expectant management became less effective and more expensive at a yearly probability of recurrent symptoms greater than 40% and 90%, respectively. CONCLUSIONS: In patients aged 60 and older, expectant management after ERC-S for CBDS is a reasonable approach, but the economic attractiveness of this strategy is highly dependent on the probability of recurrent symptoms.