Ventilator Liberation for Hospitalized Patients With Tracheostomy

Background: Among patients who are invasively ventilated in the hospital, 5–9% undergo tracheostomy for prolonged support. The preferred method of in-hospital mechanical ventilation weaning and liberation strategy for patients with tracheostomy is not well-established. Methods: A team of respiratory therapists and physicians at the Cleveland Clinic Main Campus Hospital implemented a ventilator liberation protocol for patients with tracheostomy that incorporated daily tracheostomy collar trials. We compared clinical outcomes during the first year of protocol implementation with the year prior. All patients with tracheostomy, including those who had an existing tracheostomy on admission and those who had a tracheostomy performed in the ICU, were included. Results: There were 308 unique subjects accounting for 386 ICU admissions. The primary analysis included the first mechanical ventilation episode of each hospital admission (185 in control, and 201 in intervention). Baseline characteristics were similar between the groups. At admission, 251 subjects (65%) had an existing tracheostomy; the rest (135 subjects, 35%) underwent tracheostomy after admission. Median APACHE score (IQR) was 74 (59–95) and was not significantly different between groups. Weaning duration (IQR) was 5.6 (2.2–13.0) days in the intervention group and 7.0 (2.6–11.7) days in the control group ( P = .50). The hospital stay and mortality were similar. More subjects were readmitted to the hospital in the intervention group (30.3%) compared with control (20%) ( P = .02). Considering the first mechanical ventilation episode and after adjusting for age, gender, APACHE III diagnosis and APACHE III score, the intervention group showed a trend of higher likelihood of ventilator liberation on hospital discharge compared with control group (HR = 1.29, 95% CI 0.98–1.71, P = .07). For recurrent events (up to 3), tracheostomy collar trials were associated with a 43% higher likelihood of successful liberation (HR 1.43, 95% CI 1.05–1.94, P = .02). Among subjects who underwent tracheostomy during the ICU admission, the daily tracheostomy trials were associated with an increased hazard of ventilator liberation on discharge (HR = 1.58, 95% CI 0.97–2.57, P = .08), after adjustment for confounders. Conclusions: Tracheostomy collar trials were associated with an increase in-hospital ventilator liberation for subjects with tracheostomy with the balancing measure of increase in hospital readmissions. The benefits were pronounced for subjects who had tracheostomy during the ICU admission.