The outcome of early versus late tracheostomy in neurosurgery patients: A preliminary study of open RCT

Abstract

Background: Tracheostomy is a commonly performed procedure and necessary in patients predicted to use long-term mechanical ventilation. Many neurosurgery patients need prolonged mechanical ventilation but there is controversial evidence to guide clinicians regarding the optimal time of the procedure.
Objective: The primary objective was to compare length of ICU stay, mechanical ventilation support duration, ventilator-associated pneumonia, and mortality rate between early tracheotomy (within 7 days after intubation) and late tracheostomy (more than 7 days after intubation). Whereas, secondary objective was to identify tracheostomy related complications
Patients and Methods: A randomized controlled trial conducted between June 2014 and September 2015 in Maharaj Nakhon Chiang Mai hospital. After informed consent, fifty-six eligible neurosurgery patients, whom intubated more than 72 hours and assessed for tracheostomy by clinical physicians were randomized into early and late tracheotomy group, 28 patients in each group. Primary outcomes, length of ICU stay, mechanical ventilation support duration, ventilator-associated pneumonia and mortality rate, and related complications were measured. The analysis was by intention-to-treat.
Results: In the early group, 28 patients, tracheostomy was performed after seven days in 6 patients (21.4%). Average age and tracheostomy day after intubation were 54.9 ± 20.9 years and 5.9 ± 2.0 days respectively. In the late tracheostomy group, 28 patients, average age and tracheostomy day after intubation were 63.2 ± 12.9 years and 14 ± 3.9 days respectively. Baseline characteristic was not significantly different between the groups. The average length of ICU stay in early and late tracheostomy groups were 14 and 21 days. The hazard ratio was 0.455 (p-value = 0.007), which showed that ICU discharge rate in early group was significantly 2.2 times earlier than in late tracheostomy group. Accordingly, mechanical ventilation support duration in early and late tracheostomy groups were 9 and 18 days. The hazard ratio was 0.22 (p-value < 0.001), which showed that the mechanical ventilator wean-off rate in early group was significantly 4.55 times faster than in late tracheostomy group. Ventilator-associated pneumonia were identified in 15 patients, 6 (21.4%) and 9 (32.1%) patients in early and late group (p-value = 0.55). The mortality rate in early and late group were 10.7% and 17.8%, respectively (p-value = 0.71). There was no tracheostomy complication that occurred during the study.
Conclusion: In neurosurgery patients, early tracheostomy decreases length of ICU stay and mechanical ventilator support duration significantly but no statistical difference in ventilator-associated pneumonia and mortality rate.