One of the continuing controversies in the management of acute respiratory failure is whether patients who require prolonged ventilatory support should undergo tracheotomy early (within the first 3-5 days after intubation) or late (after 2-4 weeks, or even later) in order to minimize the likelihood of laryngeal damage, subglottic stenosis, and other complications. This multicenter, controlled, clinical trial attempted to resolve this controversy by randomizing surgical ICU patients to early (at 3-5 days), later (at 10-14 days), or no tracheotomy, stratifying for type of illness (trauma with and without severe head injury; non-trauma). Prospectively defined outcomes were mortality, ICU length-of-stay (LOS), development of pneumonia, and airway complications; these last sought via flexible laryngoscopy at the time of extubation and at follow-up after 3-5 months.
At least that was the plan. Numerous problems prevented the study from being carried out as planned. Potential investigators at several major trauma centers declined to participate because of their strong opinions about when tracheotomy should be performed. One center started the study, but the center later dropped out and failed to submit any data. Two other centers withheld the forms reporting data on some of their patients.
Ultimately, 155 patients were entered into the study at five centers, but data were available on only 126. Because of numerous protocol violations and stratification by diagnostic category, relatively small numbers of patients were compared in each group with respect to early tracheotomy, later tracheotomy, or no tracheotomy, and few of the patients actually underwent follow-up laryngoscopy. As a result, there were no significant differences in any of the outcome variables with respect to the timing of tracheotomy. Sugarman and colleagues conclude that they were unable to demonstrate any benefit to early tracheotomy with respect to mortality, ICU LOS, pneumonia, or airway complications, which was the main hypothesis being tested. (Sugarman HJ, et al. J Trauma 1997;43:741-747.)
COMMENT BY DAVID J. PIERSON, MD
Sugarman et al acknowledge that their inability to detect differences in outcomes among patients undergoing tracheotomy at different times may have been due to the small sample sizes rather than the actual absence of differences (type II statistical error). In addition, APACHE III scores at entry were about 20% higher in patients who underwent early tracheotomy (P < 0.05), and Glascow Coma Scale (GCS) scores were lower among trauma patients without severe head injury who underwent early tracheotomy (mean, GCS 10 vs 13; P < 0.05). This suggests that randomization may not have been carried out as described, and that the option of early tracheotomy may have been withheld from some less severely ill patients because of investigator bias.
This study does not settle the controversy about when to perform a tracheotomy in a patient with prolonged acute respiratory failure. What it does clearly show is how difficult it is to perform a good study on this subject. There may be no hard evidence that performing tracheotomy early or delaying it until later in the course, or deferring it indefinitely, is better for patients, but most clinicians have strong opinions on the matter, albeit differing opinions. If one "knows" that early tracheotomy reduces complications and gets patients out of the ICU faster, then one may well consider it unethical to randomize some patients not to get it. On the other hand, if one "knows" that early tracheotomy is often unnecessary and predisposes to pneumonia and other complications, it would be hard to conscience participation in a study in which half of one’s patients would have this done.
Although a number of studies have recently appeared extolling the ease, efficiency, and cost effectiveness of percutaneous tracheotomy (Crit Care Alert 1997;4:90-91; 1996;4:65-66), few good data have appeared in the last 25 years to clarify the issue of when tracheotomy (by any technique) should be performed. Reasons for continued uncertainty in this matter include the heterogeneity of the patients for whom the issue is important and the variety of physicians who perform tracheotomy in the ICU (Kasper CL, et al. Respir Care 1996;41:37-42). At the institution in which I practice, for example, tracheotomy is performed on ICU patients by general surgeons, otolaryngologists, neurosurgeons, oral-maxillofacial surgeons, and burn surgeons.
Considering the variety of clinical settings in which the issue arises, the persistence of widely divergent approaches is not surprising. It is probably safe to say that either early or late tracheotomy remains acceptable for patients with prolonged acute respiratory failure requiring mechanical ventilation, in the absence of more clear-cut factors (such as high cervical spinal cord injury or bulbar paralysis) that would favor an earlier procedure.
Randomized, controlled, clinical trials of early vs. late tracheotomy in acute respiratory failure demonstrate that :
a. early tracheotomy reduces mortality.
b. early tracheotomy reduces the incidence of pneumonia.
c. early tracheotomy reduces the incidence of laryngeal stenosis.
d. all of the above.
e. none of the above.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content