Prediction and timing of tracheostomy in patients with infratentorial lesions requiring mechanical ventilatory support

Objectives: To determine the frequency and predictors of successful extubat ions and tracheostomy in patients with infratentorial lesions requiring mec hanical ventilation and to determine the optimal time for tracheostomy base d on probability of successful extubation and in-hospital survival accordin g to the duration of translaryngeal intubation. Design: Retrospective chart review. Settings. A neurocritical care unit at a university hospital. Patients: A total of 69 patients with infratentorial lesions who were mecha nically ventilated during their intensive care unit stay. Measurements and Main Results: Of the 69 patients who were mechanically ven tilated, 23 (33%) were successfully extubated. In logistic regression analy sis, both the presence of a Glasgow Coma Scale score >7 at time of intubati on (odds ratio, 4.8; 95% confidence interval, 1.2-21.7) and the absence of brainstem deficits (odds ratio, 4.3; 95% confidence interval, 1.3-16.7), we re independently associated with successful extubation. After extubation, 1 1 patients were reintubated; seven were reintubated within the same day bec ause of poor control over secretions, airway spasm, or hypoventilation. Tra cheostomy was performed in 23 (33%) patients, of whom 19 were successfully weaned off mechanical ventilatory support over a mean period of 3.7 +/- 4.0 days after tracheostomy. Patients undergoing tracheostomy had a significan tly longer intensive care unit stay (19.1 +/- 9.0 vs. 8.7 +/- 6,6 days, p < .01) and total hospital stay (34.8 +/- 18.7 vs. 20.1 +/- 9.9 days, p < .01 ) compared with patients who were successfully extubated. The probability o f successful extubation or death before extubation or tracheostomy was 67% an the day of intubation, which decreased to 5.8% after translaryngeal intu bation for >8 days. Conclusions: An aggressive policy toward tracheostomy is justified based on the low frequency of successful extubations and high frequency of extubati on failures and tracheostomies in patients with infratentorial lesions. The decision regarding tracheostomy should be made on day 8 of mechanical vent ilatory support because of the low probability of subsequent extubation or in-hospital death.