Timing and guidance modalities of tracheostomy for prolonged mechanical ventilation and associated complications: A systematic review and meta-analysis of randomized trials

This is the authors' abstract. We don't add key points for this paper.

This systematic review and meta-analysis synthesized randomized evidence on whether tracheostomy timing, induction method, and procedure type influence complications in adults. Six databases were searched to May 2025, limited to randomized controlled trials (RCTs). Relative risks (RRs) were pooled using random-effects models with Hartung–Knapp correction. Subgroup and meta-regression analyses explored timing, induction, and procedure type. Generalized linear mixed models (GLMMs) and Bayesian sensitivity analyses addressed rare events. The protocol was registered prospectively (OSF DOI: 10.17605/OSF.IO/E9KCF). Twenty RCTs (n = 2500) were included. Early tracheostomy (≤7 days) significantly reduced overall complications versus late tracheostomy (RR 0.70, 95% CI 0.50–0.97). Dose–response meta-regression suggested greater benefit at earlier cutoffs, though not statistically significant. Exploratory analyses indicated heterogeneity among ultrasound localization techniques. For rare events such as pneumothorax, GLMM and Bayesian analyses yielded neutral probabilities (Pr[RR < 1] = 0.50). Baseline risk meta-regression showed larger absolute benefit in high-risk populations, with number needed to treat ranging from 48 (low-risk) to 14 (high-risk). Overall, RCT evidence supports early tracheostomy and bronchoscopic guidance in carefully selected critically ill patients who are expected to require prolonged mechanical ventilation, have limited short-term prospects for extubation, or exhibit worsening clinical trajectories. The benefits of early tracheostomy should therefore be interpreted within these specific clinical conditions, rather than applied to all mechanically ventilated patients. Dose–response trends, ultrasound variability, Bayesian rare-event assessment, and baseline risk modeling highlight the importance of individualized patient selection. High-quality multicenter RCTs remain necessary to confirm and refine optimal practice.