Is frailty associated with prolonged mechanical ventilation in critically ill older adults? A retrospective cohort study
Year: 2026
Type: article
Study design: Observational Study
Abstract: Objective To evaluate whether frailty, assessed using the Clinical Frailty Scale (CFS), is associated with prolonged mechanical ventilation (PMV) in critically ill patients aged 65 and older, and how the competing risk of early death shapes it. Methods Retrospective single-centre cohort of consecutive patients aged ≥65 receiving invasive mechanical ventilation between November 2024 and July 2025 ( n = 138). The CFS was recorded at admission in routine care; a blinded second physician rescored 40 patients from the records. PMV followed NAMDRC. The adjustment set came from a directed acyclic graph; the primary model was Firth penalised logistic regression with a minimal adjustment set that excluded illness severity as a mediator. Because death before day 21 precludes the outcome, cumulative incidence functions, Fine and Gray and cause-specific hazard models were fitted post hoc with death on the ventilator as a competing event. Results Median age was 76.0 years (IQR 70.0–84.8); 116 patients (84.1%) were frail (CFS ≥5). PMV occurred in 38 patients (27.5%) and 83 (60.1%) died within 28 days. Inter-rater agreement was substantial (weighted κ 0.77, 0.65–0.85). No association between frailty and PMV was detected (adjusted OR 0.97, 0.74–1.27; p = 0.819), and no trend across strata ( p = 0.117). Patients developing PMV had lower APACHE II scores and lower 28-day mortality (15.8% vs. 77.0%; both p ≤ 0.007). Treating death on the ventilator as a competing event, greater illness severity was associated with a lower subdistribution hazard of liberation alive (APACHE II sHR 0.84–0.90 per point, all p ≤ 0.0003; secondary model). At the threshold, assuming unverified events were deaths on the ventilator, death on the ventilator rose across frailty strata (32%, 49%, 66%) while the proportion still ventilated fell (36%, 33%, 22%). Conclusion No frailty–PMV association was detected; the wide interval excludes neither direction. The null persisted in competing-risk analysis, so it is not explained by competing mortality alone. In this population the apparent determinants of PMV are shaped by who remains at risk of the outcome.
Source: Frontiers in Medicine
Language: English
Topic: Frailty in Older Adults
Subfield: Geriatrics and Gerontology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,801
OpenAlex ID: W7236326032
