Robotic cholecystectomy earns its place in the difficult gallbladder, not the easy one

Design

Retrospective cohort with multivariable interaction models and inverse probability weighting for cost

Population

1,828 adults undergoing minimally invasive cholecystectomy at 8 hospitals, 2020–2021

Primary outcome

Conversion or Clavien-Dindo grade 3 or worse complication, and total variable cost

Effect

Adverse outcome 3.3% (95% CI 2.2–4.8) robotic vs 8.7% (7.0–10.5) laparoscopic; robotic cost +$2,211.60 (1,521–2,902)

Earlier comparisons of robotic and laparoscopic cholecystectomy were confounded in a way that flattered neither approach: they mixed surgeons of varying robotic experience and adjusted poorly for how hard the gallbladder actually was. This study addressed both, restricting itself to surgeons who routinely perform both operations across eight hospitals, and stratifying difficulty with the Nassar preoperative score.

The overall figures favour the robot — adverse outcome 3.3% versus 8.7% — but the overall figure is the least useful number here. Approach interacted significantly with both Nassar risk level and body mass index. In high-risk gallbladders at a body mass index of 30, the laparoscopic approach carried an odds ratio of 3.69 for conversion or a serious complication. In low and intermediate-risk patients with a low body mass index, the two were comparable; the laparoscopic disadvantage only reappeared at a body mass index around 35.

Cost went the other way, consistently. Robotic surgery added about $2,212 per case in low and intermediate-risk patients and $2,417 in high-risk ones, with confidence intervals nowhere near zero. So the question this study actually answers is not which approach is better but which patients justify the difference — and it offers a risk calculator, with an area under the curve of 0.744 in training and 0.750 in validation, to make that judgement before the operation rather than during it.

• Score the gallbladder before choosing the platform — Nassar, imaging findings, duration of symptoms, previous cholecystitis
• The interaction with body mass index is the practical part: high body mass index plus a difficult gallbladder is where the approach changes outcomes
• An easy gallbladder in a slim patient does not need a robot on this evidence
• The cost difference is per case and does not include capital or maintenance, so a unit's real figure is higher
• Discrimination around 0.75 is useful for triage, not for individual prediction — treat the calculator as a prompt, not a verdict

Why it matters

It reframes the robotic question from a platform argument into a selection problem, which is the only form of it a unit can actually act on.

Don't overread it

Retrospective and single health system, with approach assigned by operative intent rather than randomised. Surgeons choosing the robot for a hard gallbladder is precisely the confounding this design can reduce but not remove.

The statistics, in plain English

An odds ratio of 3.69 with a confidence interval of 2.04 to 6.69 is a large effect measured imprecisely — the true value could be double or triple, but it is unlikely to be nothing. The cost estimates are the opposite: narrow intervals around a modest figure, so the $2,200 is the number to trust most in this paper. Note the interaction terms are doing the work; reading only the headline 3.3% versus 8.7% would suggest the robot helps everyone, which the model explicitly contradicts.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.