Airway Anatomy as a Predictor of Tracheostomy in Critically Ill Patients: A Multicampus Analysis of Airway Grade, Admission Diagnosis, and Clinical Trajectory

Vollständiger Abstract

Worum geht es in dieser Arbeit?

ABSTRACT Objective Determine whether the Cormack–Lehane airway grade at intubation predicts subsequent tracheostomy in critically ill adults and whether this association persists after controlling for admission diagnosis. Methods Retrospective multicenter cohort study of 10,234 ICU patients requiring ventilation across three campuses from April 2018 to June 2024. The primary predictor was Cormack–Lehane grade (1–4) during intubation; the primary outcome was tracheostomy. Principal problem documentation classified patients into 12 diagnostic categories. Multivariable logistic regression adjusted for age, sex, BMI, Charlson Comorbidity Index, intubation method, and admission diagnosis. Sensitivity analysis compared the initial versus the worst grade. Results Of 6189 patients with documented airway grades, 922 (14.9%) underwent tracheostomy. Rates generally increased with grade severity through Grade 3B (24.5%), with Grade 4 (19.4%, n = 31) showing wide confidence intervals consistent with small sample size and high mortality in this subgroup. Otherwise, each ordinal increase in grade conferred 14% higher unadjusted tracheostomy odds (OR: 1.14, 95% CI: 1.05–1.23, p = 0.001). Rates varied by diagnosis, from 28.4% for COVID‐19 respiratory failure to 6.5% for gastrointestinal/hepatic conditions. In the fully adjusted model ( N = 4770), airway grade remained independently significant (aOR: 1.13, 95% CI: 1.03–1.24, p = 0.014). The strongest predictors were transplant (aOR: 2.95), COVID‐19 respiratory failure (aOR: 2.60), and cardiac disease (aOR: 1.84). Patients with Grade ≥ 3 airways had 70% higher odds compared to Grade 1–2B (OR: 1.70, p = 0.002). Patients with worsening grades across re‐intubation had higher tracheostomy rates (38.8% vs. 14.2%; OR: 3.8, p < 0.001). The model achieved an AUC of 0.638. Conclusion Difficult airway anatomy predicts tracheostomy risk after controlling for admission diagnosis, comorbidities, and demographics. Diagnosis provides additional discrimination, with COVID‐19, transplant, and cardiac patients at highest risk. Grade worsening across re‐intubation identifies an ultrahigh‐risk subgroup. These findings support systematic airway grade documentation integrated with clinical context for ICU risk assessment. Level of Evidence III.