Robotic bronchoscopy matched CT-guided biopsy with fewer pneumothoraces

Design

systematic review and meta-analysis restricted to direct comparative studies, ROBINS-I and GRADE assessed

Population

three mutually independent retrospective cohorts from one US health system, adults with peripheral pulmonary lesions

Primary outcome

strict 2024 ATS/CHEST diagnostic yield, with pleural complications

Effect

yield RR 0.99 (95% CI 0.93–1.06, I²=24%); pneumothorax needing drain or admission RR 0.25 (0.14–0.46, I²=0%)

Five retrospective studies comparing robotic-assisted bronchoscopy with CT-guided transthoracic needle biopsy for peripheral pulmonary lesions were identified, all from a single US health system. Four shared patients, so the authors used a cohort-genealogy step to pool only the largest mutually independent set — three cohorts — which is a more careful approach than most reviews of this kind manage.

Diagnostic yield was comparable under the strict 2024 ATS/CHEST definition: risk ratio 0.99 (95% CI 0.93 to 1.06, I² = 24%). Absolute yields fell from 88% to between 74% and 84% when the strict definition was applied rather than a looser one, which is a useful reminder of how much definition choice moves these numbers. Pneumothorax requiring a chest drain or admission was about three-quarters less frequent with the robotic approach (RR 0.25, 0.14 to 0.46, I² = 0%). Procedure time was about 50 minutes longer where same-session staging endobronchial ultrasound counted toward robotic time, and about 8 minutes longer where it did not.

The authors rate certainty as low for pleural complications and very low elsewhere, and say plainly that this is hypothesis-generating and needs a multicentre randomised trial. All the evidence is retrospective from one system, and yield was almost never stratified by lesion size, bronchus sign or lung zone — the three things that actually determine whether a bronchoscopic approach can reach a lesion.

• Choose the sampling route by lesion size, location and bronchus sign, which this review could not stratify by
• Where robotic bronchoscopy is available, the lower pleural complication rate is the substantive argument
• Count same-session staging EBUS separately when comparing procedure times
• Ask which yield definition a reported local figure used — strict criteria drop it by 4 to 14 points

Why it matters

The comparison usually made on yield may turn on complications instead.

Don't overread it

Retrospective, single health system, low to very low certainty — this does not establish that either approach is better.

The statistics, in plain English

A risk ratio of 0.99 with an interval of 0.93 to 1.06 shows no detectable difference, but the authors are explicit that equivalence was not formally established — absence of a detected difference is not proof of sameness. The pneumothorax interval of 0.14 to 0.46 excludes 1.0 comfortably, though the Hartung-Knapp correction widens it to 0.07 to 0.96, which is much less reassuring. Retrospective data from one health system means the patients were selected for each procedure by clinicians who knew both options, which is the central confounder.

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.