Endoscopic ultrasound versus endoscopic retrograde cholangiopancreatography for primary palliation of malignant distal biliary obstruction: a cost-effectiveness modeling analysis.

Journal: Endoscopy

Study type

  • Decision-analytic, state-transition Markov model over 1 year
  • US healthcare payer perspective
  • Population modeled: 70-year-old with locally advanced, unresectable pancreatic cancer, malignant distal biliary obstruction, and common bile duct dilation >15 mm

Interventions compared

  • Endoscopic ultrasound–guided choledochoduodenostomy (EUS-CDS) using a lumen‑apposing metal stent (LAMS)
  • Endoscopic retrograde cholangiopancreatography–guided transpapillary biliary drainage (ERCP-BD) using a self‑expandable metal stent (SEMS)

Data sources and outcomes

  • Transition probabilities: from meta-analyses of randomized trials
  • Main outcome: incremental cost-effectiveness ratio (ICER), expressed as cost per quality-adjusted life year (QALY) gained
  • Willingness‑to‑pay threshold: $100,000 per QALY
  • Sensitivity analyses: one-way and probabilistic sensitivity analyses performed

Key results

  • EUS‑CDS with LAMS was cost-effective versus ERCP‑BD with SEMS, with an ICER of $47,711/QALY.
  • One-way sensitivity analyses:
    • EUS‑CDS remained cost-effective if its cost was < $15,502 or if ERCP‑BD cost > $11,174.
    • ERCP‑BD could become cost-effective if:
      • Technical success > 91%, or
      • Reintervention rate < 11%, or
      • Post-procedural pancreatitis rate < 4%.
  • Probabilistic sensitivity analysis: EUS‑CDS was cost-effective in 74.1% of simulations at the chosen threshold.

Conclusions and implications for practice

  • For patients with malignant distal biliary obstruction and significant biliary dilation (>15 mm), EUS‑CDS with LAMS appears not only clinically favorable (based on prior trials) but also economically attractive as a primary palliative drainage strategy.
  • Wider adoption may be supported by further reducing LAMS costs, optimizing stent performance (less dysfunction/reintervention), and refining anatomic selection criteria.

Leave a Reply