Abstract
Background
Patients requiring prolonged mechanical ventilation (PMV) after tracheostomy represent a severely deconditioned population with high mortality and substantial healthcare burden. Weaning outcomes vary widely across centers, and evidence supporting standardized protocol in dedicated rehabilitation settings remains limited.Methods
We conducted a prospective, single-arm interventional trial at a specialized weaning unit (SWU) of a tertiary rehabilitation hospital. From January 2025 to October 2025, consecutive tracheostomized PMV patients were enrolled and managed with a standardized stepwise weaning protocol integrating spontaneous breathing trials, high-flow oxygen therapy, and noninvasive ventilator. The primary outcome was the 60-day weaning success rate, which was tested against a pre-specified historical benchmark (p₀ = 65%) using exact binomial, score-based, and Katz log-method analyses. Secondary outcomes included 6-month mortality, analyzed through modified Poisson regression with propensity score matching and overlap weighting. An individualized calculated weaning score was developed using LASSO-Poisson regression with log(p₀) as an offset to elucidate the heterogeneity in treatment response.Results
Weaning success was achieved in 189 of 250 patients (75.6%; 95% CI 69.9-80.5%), significantly exceeding the historical benchmark (RR, 1.16; 95% CI 1.08-1.25; p Conclusions
A standardized weaning protocol delivered in a dedicated rehabilitation unit was associated with a 60-day weaning success rate that exceeded a pre-specified historical benchmark in PMV patients. The individualized calculated weaning score identifies clinical phenotypes for individualized decision-making; the accompanying web-based calculator is provided as an exploratory research tool and warrants external validation before any routine clinical application.Trial registration
ClinicalTrials.gov, NCT06642714.