Answers are grounded in this paper and cite their sources.
Systematic review and meta-analysis reveals lower restenosis rates with tracheal resection than endoscopy in benign tracheal stenosis, highlighting the necessity of stratified clinical evaluations.
Objective. To estimate the rate of restenosis after surgical and endoscopic treatment of acquired benign tracheal stenosis in adults with separate analyses by approach, etiology, and follow-up. Material and methods. Systematic searching was performed in PubMed/MEDLINE, EuropePMC, Cochrane CENTRAL, eLibrary.ru, BASE, LILACS databases and trial registries (2005–2025). Cohorts ≥100 patients (primary set) and 50–99 (sensitivity set) were eligible. Two reviewers performed blinded screening and extraction (Cohen’s κ=0.808); the risk of bias was assessed with ROBINS-I. Proportions were pooled under random effects with stratification by approach, etiology and follow-up. Results. Primary set comprised 14 cohorts (3.213 patients; 690 events); sensitivity set — 25 cohorts. Heterogeneity was high (I²=98.0%). The pooled restenosis rate was 16.8% (95% CI 13.5–20.6; 95% PI 1.1–78.2). There were the following values depending on approach: circumferential tracheal resection 7.8%, mixed 10.0%, isolated endoscopy 47.6% (p<0.0001). Idiopathic subglottic cicatricial stenosis of the trachea comprised 58.9%, PITS+PTTS — 5.7%, mixed/other — 15.3% (p<0.0001). Sensitivity set yielded 22.3%. The estimate was robust to sequential cohort exclusion (16.5–25.2%). Egger’s test showed funnel asymmetry (p=0.019) opposite to classical small-study pattern; trim-and-fill imputed no missing studies. Conclusion. The pooled 16.8% does not predict individual outcome; only the stratified estimate is clinically usable. Direct surgical-versus-endoscopic comparison on retrospective data is unsound because patient selection differs by approach. In idiopathic subglottic stenosis, the true restenosis rate under adequate prospective follow-up exceeds the primary-set estimate.
No takes yet. Share an insight, caveat, or question.
Ursov et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: