Bronchodilating effect of intravenous magnesium sulfate.

We have read the report by Milam et al about the usefulness of an immediately postbronchoscopy roientgenogram as a control at 525 to) benefit fromn tracheostomy. It would be difficult to draw any conclusioin on the tmsefulmmess of the clinical features for selecting patients for early intubation without reviewing the data from these 40 patients. Our study examined patients with ARDS for clinical features apparent on day 7 of respiratory failure that would predict the need for long-term mechanical ventilation. Forty of 85 ARDS patients were excluded because their early deaths prevented determination of the study end point, which was intubation for moire than 14 days. In examining day 7 data from these 40 patients, we found that 21 were hypotensive while receiving vasopressor medication, with multiorgan failure and sepsis. These patients died betweemm day 7 and day 10 and would not have undergone tracheotomy before death at our institution because ofclinical instability. An additional 11 patients were not candidates for tracheotomy because of " do not resuscitate " status ordered in the setting of irreversible coma, advanced age, or malignancy. Four patients had total skin surface burns and appeared terminal according to the clinicians' notes by day 7. Two additional patients died suddenly of unknown causes during the second week of respiratory failure; on the basis of the day 7 clinical features, they appeared to be improving and would have received continued translaryngeal intubation. The remaining two patients might have undergoine tracheotoimy oo n the basis of the day 7 criteria because of severity of respiratory failure. These patients, however, died of intractable sepsis on days 12 and 13, respectively; it is not clear froim the records that tracheotoimy would have been advantageous or deleterioius to) their coiurses. In considering these data, one should recognize that mamiy ARDS patients with courses sufficiently protracted to warrant consideration foir tracheotomy will subsequently die regardless oif the timing of the procedure. Indeed, in our study only four of the 1 1 patients who underwent tracheotomy survived their hospitalizatiomm even though surgery was performed 7 to 51 days (mean, 20.6 ± 3.4 days) after intubation. Delaying tracheotomy, therefore, is no guarammtee that patients with a high prol)ahihty of survival will be selected. We continue to conclude that patients can benefit from trache-otomy perfoirmed in a moire timely fashion rather than waiting for 14 to 21 days of respiratoiry failure. Our study demonstrated that …