- Design
- Systematic review and meta-analysis of observational studies
- Population
- 39,918 patients (13,638 robotic-assisted, 26,280 conventional CABG)
- Primary outcome
- Mortality and postoperative complications
- Effect
- Mortality OR 0.55 (95% CI 0.43–0.69); atrial fibrillation OR 0.53
This meta-analysis pooled nine observational studies, 39,918 patients in all, comparing robotic-assisted with conventional coronary artery bypass grafting. No randomised trial was available, and long-term data were sparse.
Robotic surgery was associated with lower mortality (OR 0.55, 95% CI 0.43–0.69), less postoperative atrial fibrillation (OR 0.53), less respiratory failure, fewer transfusions, about two days' shorter hospital stay and about 23 hours less in intensive care. Stroke did not differ significantly.
The authors themselves warn about case selection, and it is the central problem: robotic CABG is offered to fitter patients with simpler anatomy, often single-vessel disease, by high-volume teams. A 45% mortality difference of that size is more likely to reflect who was chosen than what was done. For referral decisions, this supports robotic CABG as a reasonable option in selected patients at experienced centres, not as a better operation.
- Robotic CABG is a reasonable option for selected patients at experienced centres.
- Do not quote a halving of mortality to patients; these were selected, observational comparisons.
- Expect shorter stays and fewer transfusions as the likelier real benefits.
- Multivessel disease and complex anatomy remain the domain of conventional surgery for now.
- Access to robotic cardiac surgery in India is limited to a few centres; refer only where the volume exists.
Why it matters
Large observational numbers can make selection bias look like proof.
Don't overread it
All nine studies were observational, so the mortality difference cannot be attributed to the robotic technique.
The statistics, in plain English
An odds ratio of 0.55 would mean 45% lower odds of death. I² of 0% says the studies agree, but studies that share the same bias will agree with each other. Trial sequential analysis can show the data are statistically settled, not that they are free of confounding.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free