Perioperative outcomes of robot-assisted, laparoscopic or thoracoscopic, and open surgery across 13 cancer operations: a single-institution cohort of 36,208 cases

Why clinicians should know about it

• Picked for Surgery (paper of the day, 19 August 2026): Robotic/video-assisted lower complications than open across cancers

Abstract

To compare perioperative outcomes of robot-assisted surgery, video-assisted (laparoscopic or thoracoscopic) surgery and open surgery across 13 oncologic procedure groups, because aggregate technology-level comparison is confounded by procedure mix and by the calendar time at which each specialty adopted robotics. In this single-institution retrospective cohort study, 36,208 index oncologic procedures in 34,785 patients (January 2016 to December 2025) were identified from the institutional clinical data warehouse. The whole cohort and 22 procedure-level pairwise comparisons were adjusted by stabilized inverse probability of treatment weighting, with indication-specific cancer-site adjustment where tumor stage was not comparable across arms. The primary outcome was major complications (Clavien-Dindo grade ≥ 3); secondary outcomes were perioperative transfusion, length of stay, 30-day readmission and conversion to open surgery. The robot-assisted share rose from 16.8% in 2016 to 48.2% in 2025, while the open share fell to 19.4%. After whole-cohort weighting, all covariates reached a standardized mean difference of 0.10 or less. Major complications were lower for robot-assisted than open partial nephrectomy (odds ratio 0.33, 95% confidence interval 0.19 to 0.55) and radical prostatectomy (0.24, 0.13 to 0.42). Transfusion was lower in 14 of 22 comparisons and length of stay in 13, the largest reduction being robot-assisted radical cystectomy (geometric mean ratio 0.64). Robot-assisted and video-assisted approaches were associated with fewer complications and transfusions and shorter stay than open surgery, most robustly for partial nephrectomy and radical prostatectomy. Procedure-level rather than aggregate evidence should guide interpretation and robotic investment.

Abstract as published, via PubMed.