Search

Objective: The value of optimal timing of tracheostomy in patients with subarachnoid hemorrhage is controversially debated. This study investigates whether early or late tracheostomy is associated with beneficial outcome or reduced rates of adverse events., Design: Retrospective observational multicentric on patients prospectively inserted into a database., Setting: Neurologic ICUs of one academic hospital and two secondary hospitals in Germany., Patients: Data of all patients admitted to the Goethe University Hospital between 2006 and 2011 with poor-grade subarachnoid hemorrhage were prospectively entered into a database. All patients who underwent tracheostomy were included for analysis. Follow-up was maintained in primary and secondary ICUs., Interventions: Patients underwent tracheostomy upon expected long-term ventilation. Early tracheostomy was defined as performed on days 1-7 and late tracheostomy on days 8-20 after admission., Measurement and Main Results: We compared 148 consecutive patients admitted with poor-grade (World Federation of Neurosurgical Societies, 3-5) subarachnoid hemorrhage. Early tracheostomy was performed in 39 patients and late tracheostomy in 109 patients. In early versus late tracheostomy groups, no significant differences were observed with regard to ICU mortality (7.7% vs 7.3%; p=0.93) and median modified Rankin Scale after 6 months (3 vs 3; p=0.94). Of the early group, pneumonia developed in 19 patients, whereas in the late group, pneumonia developed in 75 patients (48.7% vs 68.8%; p=0.03; odds ratio, 2.32; 95% CI, 1.1-4.9). Six patients of the early group (15.4%) and 36 patients of the late group (33%) suffered from respiratory adverse event (p=0.04; odds ratio, 2.71; 95% CI, 1.04-7.06). Mechanical ventilation was shorter (17.4 vs 22.3 d; p<0.05) and decannulation occurred earlier (42 vs 54 d; p=0.039) in the early tracheostomy group., Conclusions: Tracheostomy within 7 days of critical care admission is a feasible and safe procedure for patients with poor-grade subarachnoid hemorrhage. Early tracheostomy was not associated with an improvement in mortality or neurologic outcome but associated with fewer respiratory adverse events.