Cost-analysis of implementing robot-assisted versus open pancreatoduodenectomy

Objective: To perform a cost-analysis during the implementation of robot-assisted pancreatoduodenectomy (RPD) in a high-volume center. Background: Many high-volume centers are implementing RPD as alternative to open pancreatoduodenectomy (OPD) but the persevering concerns about increased costs of RPD versus OPD have not been addressed by large studies. Methods: Post-hoc cost-analysis of consecutive RPD versus OPD in a singly high-volume center (Jan 2015–May 2024). The eligibility criteria for RPD (i.e. no vascular contact, no chronic and necrotizing pancreatitis, and BMI<35 kg/m²) were used to select patients undergoing OPD, to minimize selection bias. Primary outcomes were total hospital costs and total hospital stay. Sensitivity analyses excluded the first half of the RPD experience. Results: Overall, 588 patients undergoing pancreatoduodenectomy were included (214 RPD, 374 OPD). Total hospital stay was shorter after RPD (10 [6–17] vs 12 [8–21] days, p=0.001). Mean total hospital costs were €4,804 higher for RPD (€27,307 vs €22,503, p=0.010). Outcomes improved in the second half of the RPD experience (n=107): total hospital stay (12 [7–23] to 9 [6–15] days, p<0.001), pancreatic fistula (48.6% to 31.8%, p=0.012), delayed gastric emptying (33.6% to 15.0%, p=0.001), and operative time (373 [341–411] to 310 [272–352] minutes, p<0.001). Consequently, costs per RPD procedure decreased with €11,058 (€32,836 to €21,778, p=0.001). Costs in the second half of the OPD experience remained stable (€24,025 to €21,013, p=0.210). Hence, total costs for RPD and OPD became comparable in the second half (€21,778 vs €21,013, p=0.644). Conclusion: Implementing RPD is associated with considerable additional costs compared to OPD. With growing experience, both the outcomes and cost-efficiency of RPD improve, with costs decreasing by approximately a third, leading to similar costs as OPD. Large scale randomized trials will have to confirm these findings.