The “Big short” of minimally invasive pancreaticoduodenectomy for pancreatic ductal adenocarcinoma. A cost-effectiveness analysis based on randomized trials

Background The cost-effectiveness of a laparoscopic (LPD) and robotic pancreaticoduodenectomy (RPD), compared with the open approach (OPD), is still under debate. Methods A Markov decision model was developed to compare OPD, LPD, and RPD. The healthcare costs and quality-adjusted life years (QALYs) were estimated by calculating the incremental cost-effectiveness ratio (ICER) per QALY gained. A willingness-to-pay (WTP) of $130,049 was assumed as the threshold. A probabilistic sensitivity analysis (PSA) was performed to reflect the uncertainty of various parameters. Results In the base case scenario, LPD and RPD were associated with increased costs of US $30,047 and US $30,822, respectively, leading to an ICER of US $-3,911,669 and US $-1,164,992 per QALY. When comparing LPD with OPD, three main factors influence the model: OPD costs (68.8 %), LPD costs (27.9 %), and the complication rate after LPD (2.3 %). In comparing RPD with OPD, two main factors affect the model: RPD costs (75.2 %) and OPD costs (23.3 %). PSA analysis confirmed that OPD was the most cost-effective choice in most cases (62.6 %), while RPD and LPD were the most cost-effective procedures in 26.2 % and 11.9 %, respectively. Conclusion The RDP and LPD were not cost-effective. OPD remained the best approach.