Robotic vs. bedside staplers in robotic lung resection: a real-world cohort study.

Comparative studies of stapler strategy within robotic lung resection remain limited, particularly beyond lobectomy and with contemporary staplers. This study evaluated clinical and economic outcomes associated with robotic versus bedside staplers in robotic lung resections. We conducted a retrospective cohort study using the Premier Healthcare Database (2019-2023), including adults who underwent robotic lobectomy or sublobar resection. Patients were classified into robotic stapler and bedside stapler groups. Overlap weighting (OW), a propensity score weighting method targeting the average treatment effect in the overlap population (ATO), was applied to balance measured baseline characteristics. The prespecified primary outcome was postoperative air leak. Secondary outcomes included other postoperative complications, conversion to thoracotomy, operative time, length of stay (LOS), and hospital costs during the index hospitalization and 30-day perioperative period. Among 15,140 patients, robotic staplers were associated with lower air leak (10.9% vs. 15.2%), pneumothorax (8.5% vs. 11.4%), and pneumonia (1.1% vs. 2.8%), as well as lower conversion (0.4% vs. 4.7%), transfusion (1.0% vs. 1.8%), shorter operative time (209 vs. 225 min), and length of stay (3.8 vs. 5.4 days) (all p < 0.05). Index hospitalization cost was lower with robotic staplers (USD 25,887 vs. 27,432; Δ - 1,545), with similar findings for 30-day perioperative cost. In the sensitivity analysis including mixed-use cases, associations remained directionally consistent, including lower air leak (11.3% vs. 15.3%), conversion (0.8% vs. 4.6%), and total hospital cost (USD 26,370 vs. 27,365). In robotic lung resections, robotic staplers were associated with fewer complications and conversions, shorter operative time and LOS, and lower index and 30-day hospital costs. These findings were consistent in a sensitivity analysis including mixed-use cases but should be interpreted cautiously because residual confounding cannot be excluded.