Does Early Tracheostomy Improve Outcomes?

Abstract & Commentary

Synopsis: This prospective, randomized trial evaluated the role of early percutaneous tracheostomy in critically ill adults projected to require more than 14 days of mechanical ventilation. They found significant reductions in ICU length of stay, the incidence of ventilator-associated pneumonia, and mortality.

Source: Rumbak MJ, et al. Crit Care Med. 2004; 32(8):1689-1694.

Percutaneous dilatational tracheostomy (PDT) is now a common procedure in the intensive care unit (ICU). Typically, it is performed when a patient fails weaning from mechanical ventilation or has a prolonged need for an artificial airway because of a neurological disorder. However, the optimal timing for performing PDT remains uncertain. Previous studies have not been conclusive as to whether early tracheostomy offers any advantage. Therefore, Rumbak and colleagues studied the effects of early (within 48 hours) or late (14-16 days) PDT on the outcome of mechanical ventilation, ICU length of stay. and mortality in a select group of mechanically ventilated critically ill adults.

Rumbak and colleagues performed a prospective randomized trial involving patients requiring intubation for acute respiratory failure in medical ICUs in United States. The inclusion criteria called for patients to be older than 18 years with a projected need for mechanical ventilation > 14 days, APACHE II score > 25 and the availability of informed consent. Patients were excluded from the study if a) they had had a previous tracheostomy; b) they were not a candidate for PDT due to neck deformity; c) the platelet count was < 50,000/mm3 or another bleeding tendency was noted; and d) positive end-expiratory pressure (PEEP) was > 12 cm H2O, or e) they had already been on mechanical ventilation for > 48 hours. The patients were cared for under a protocol with low tidal volume, standard precautions for prevention of ventilator-associated pneumonia (VAP) were observed, and invasive methods were used to diagnose VAP. A standard protocol was used for weaning.

A total of 120 patients were enrolled in the study. All 60 patients in the early tracheostomy group, and 50 out of 60 assigned to delayed tracheostomy, received the procedure. The overall mortality was significantly lower in the early tracheostomy group (19 out of 60, 31.6%) as compared to the delayed group (37 out of 60, 61.7%; P < 0.005). VAP developed in 3 patients (5%) in the early group as compared to 15 (25%; P < 0.005) in the delayed group. Similarly, days of mechanical ventilation, days in ICU, and days sedated were significantly lower (P < 0.001 for all 3 outcomes) in the early tracheostomy group (4.8±1.4 days, 7.6±4.0, and 3.2±0.4 days respectively) compared to the delayed group (16.2±3.8 days, 17.4±5.4 days, and 14.1±2.9 days). Rumbak et al conclude that PDT is not only safe but also a very effective procedure in improving outcomes of critically ill patients.

Comment by Uday B. Nanavaty, MD

There is ample evidence in the literature that PDT can be safely performed at the bedside in critically ill patients. PDT requires much less time compared to traditional surgical tracheotomy performed in the operating room. The procedure is relatively easy to learn, and since intensivists can perform it at the bedside, the procedure is often accomplished much faster than when it must be scheduled in an operating room, even when elective. In both published studies and practice, complication rates are very low. It has been shown before that tracheostomy is well tolerated by patients compared to translaryngeal endotracheal intubation, resulting in more comfort, less sedative use and perhaps less incidence of VAP.

One thing that remains unclear is the appropriate timing for doing the procedure. After initial enthusiasm for early tracheostomy in the 1980s, benefits of performing this procedure early have been questioned. The present study is the first to show such a dramatic improvement in outcome, including mortality, by performing early PDT. An earlier, larger, randomized clinical trial in mainly trauma and other surgical ICU patients compared early vs late surgical tracheostomy and failed to show any significant improvement in outcome.1

Which variables did Rumbak et al use to predict the need for mechanical ventilation for greater than 14 days? It was evidently a very accurate prediction model since more than 80% of patients in their "delayed" group indeed needed a tracheostomy. One previous study2 suggests that physicians and nurses can correctly predict the need for < 3 or < 7 days of mechanical ventilation about 60% of the time. If an accurate predictor for prolonged mechanical ventilation need can be established, then early tracheostomy is clearly a favorable procedure. At this point, I find it hard to convince myself, and the relatives of critically ill patients, that PDT should be performed within first 48 hours. More studies are needed to confirm the successful prediction of the need for prolonged mechanical ventilation, as well as the benefits of early (within 48 hours of intubation) vs later PDT.


1. Sugerman HJ, et al. J Trauma. 1997;43(5):741-747.

2. Afessa B, et al. Chest. 1999;116(2):456-461.

Comment by David J. Pierson, MD

No hypothesis relating to respiratory care in the ICU has proved more difficult to study in an objective fashion than the commonly held tenet that tracheostomy facilitates ventilator weaning. The study by Sugerman et al, referred as reference #1 above, is testimony to that statement. A project of the Western Trauma Association Multi-Institutional Study Group, it took considerably longer to complete, and wound up rather differently than originally planned. Some member centers declined to participate because of the strongly held opinion that it was inappropriate to perform a tracheotomy in the first few days—and others refused to join because they felt it inappropriate not to do the procedure early. Investigators at other centers enrolled patients but failed to complete data collection, and there were numerous protocol violations. The study failed to demonstrate any advantage to early tracheostomy. More than anything else, though, it showed how deeply rooted—and divergent—clinicians’ biases are with respect to the timing of this procedure.

That definitive studies of early vs late tracheostomy have been hard to come by has not been for want of trying. However, several major obstacles stand in the way of any attempt at a large-scale, randomized clinical trial. The inability to blind the investigators to the patient groupings is a big problem. So is the lack of uniformity among criteria for successful weaning, extubation, and reintubation, and also the increasing use of noninvasive ventilation in recent years to avoid reintubation. In addition, today, the performance or withholding of a tracheostomy may be a function of reimbursement and disposition considerations rather than clinical factors: some patients cannot be transferred to long-term weaning facilities unless the receiving institution rather than the referring hospital performs the procedure.

Into this context comes the clinical trial of Rumbak et al, involving somewhat fewer patients than the study of Sugerman and colleagues, with the striking findings that early PDT markedly shortens ventilator time and ICU LOS, reduces the incidence of pneumonia by 75%, and cuts the overall mortality rate in half. These are pretty dramatic outcomes, considering the less dramatic findings of previous studies and the fact that it was essentially just the timing of an elective procedure that was studied.

Rumbak et al make the following statement in their conclusion: "Early tracheotomy in critically ill medical patients who undergo ³14 days of ventilation may have significant benefits over delayed tracheotomy." A primary selection criterion for including patients in the study was the clinicians’ judgment that ventilatory support would be required for at least 14 days. The finding that patients in the early tracheostomy group were in fact ventilated for only 7.6±4.0 days has 3 possible explanations: first, that early PDT somehow makes respiratory failure resolve more quickly; second, that even experienced clinicians cannot tell in the first 2 or 3 days of critical illness which patients will require prolonged mechanical ventilation; or, third, that some aspect of our management tends to keep patients on the ventilator when they no longer need to be. Which of these possibilities is most likely to occur cannot be answered from this study, and definitely needs further investigation.

Uday B. Nanavaty, MD, Pulmonary and Critical Care Medicine Rockville, Maryland, is Associate Editor for Critical Care Alert.

David J. Pierson, MD, Pulmonary and Critical Care Medicine Harborview Medical Center University of Washington, is Editor for Critical Care Alert.