The trusted source for
healthcare information and
Steroids to Prevent Extubation Failure?
Abstract & commentary
By David J. Pierson, MD, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle. Dr. Pierson reports no financial relationships relevant to this field of study. This article originally appeared in the March 2009 issue of Critical Care Alert. It was peer reviewed by William Thomspon, MD. Dr. Thompson is Staff Pulmonologist, VA Medical Center; Associate Professor of Medicine, University of Washington; he reports no financial relationships relevant to this field of study.
Synopsis: This meta-analysis of studies examining the efficacy of systemic corticosteroids for preventing laryngeal edema following extubation concludes that this treatment is effective. This result differs from those of several previous meta-analyses, and raises practical issues such as whether extubation should be delayed for at least 12 hours after a patient passes a spontaneous breathing trial and qualifies for extubation so that a course of steroids can be given.
Source: Fan T, et al. Prophylactic administration of parenteral steroids for preventing airway complications after extubation in adults: Meta-analysis of randomised placebo controlled trials. BMJ. 2008;337: a1841; doi: 10.1136/bmj.a1841.
Post-extubation laryngeal edema, although infrequent, can necessitate reintubation and lead to other complications. The administration of a course of parenteral corticosteroids prior to extubation to reduce the likelihood of laryngeal edema has been advocated for many years, although the studies published to-date have been small and several meta-analyses of their aggregate findings have failed to show this treatment to be effective. Fan et al at Sichuan University in Chengdu, China, performed a new meta-analysis of those studies, including an additional, recently published series and different selection criteria for the data to use in the meta-analysis, and concluded that steroid treatment is effective in preventing post-extubation laryngeal edema and the need for reintubation.
Using PubMed, the Cochrane Controlled Trials Register, and several other databases, Fan et al searched for randomized, controlled trials comparing parenterally administered corticosteroids to placebo for the prevention of laryngeal edema following extubation. Their search produced six eligible trials, reporting a total of 1,923 patients. They reasoned that, because laryngeal edema was the condition targeted by corticosteroid treatment and, hence, the outcome variable of interest, only those patients with this cause for respiratory distress and reintubation following extubation needed to be included in the meta-analysis. Using data selected in this manner where possible, and including 80 patients from a recently reported study from Taiwan that was not available for the previous meta-analyses, Fan et al found the following results: Compared with placebo, steroids given in multiple doses over 12-24 hours prior to planned extubation decreased the odds ratio for laryngeal edema (0.38; 95% confidence interval [CI], 0.17-0.85) and subsequent reintubation (0.29; 95% CI, 0.15-0.58). A single dose of steroids immediately prior to extubation had no significant effect. They found no adverse effects of steroids as used in the studies examined.
By the tenets of evidence-based medicine, the only evidence more authoritative than the results of a randomized, controlled trial is a meta-analysis of multiple such trials. Thus, the current study by Fan et al ought to be pretty much the last word on whether corticosteroids are beneficial for preventing post-extubation laryngeal edema, as well as the need for reintubation. But what happens when more than one meta-analysis is available, based on pretty much the same evidence, and they come to different conclusions? Such is the case here. A Cochrane review, also published in 2008, concluded that corticosteroids had not been shown to be effective for preventing either laryngeal edema or reintubation.1 As pointed out in the editorial accompanying the Fan meta-analysis, methodological differences likely explain the disparate results: "The difference in results comes from a combination of the new data, and a careful selection of the 'most appropriate' data from the five other studies. Where possible, Fan and colleagues included only patients who needed reintubation for laryngeal edema and excluded those who were reintubated for other reasons, who would not respond to corticosteroids and who would dilute any effect. This selection allowed them to use a less conservative (fixed effects) model than that used in the previous review."2
Patients fail extubation for a number of reasons, including inability to protect the upper airway because of altered neurological status, the inadequate clearance of lower respiratory tract secretions, and insufficient recovery of ventilatory muscle and airway function after acute respiratory failure to sustain the required work of spontaneous breathing, in addition to laryngeal edema. Only the last of these would be expected to be prevented by a course of systemic corticosteroids. This fact partly justifies the Fan et al strategy of excluding other causes for reintubation in their meta-analysis. However, there are two problems. First, when patients do not do well after extubation and the managing clinician decides that reintubation is necessary, the specific reason is often unclear and, while a few such patients have clear-cut laryngeal edema, most do not. This makes it unlikely that any post-hoc procedure for classifying reintubation into various causes would be completely accurate, and raises doubts about the appropriateness of selectively omitting some patients in the published studies from analysis. And, second, if laryngeal edema is only one of several potential reasons for reintubation for which steroids can help, a large number of patients would have to receive the preventive therapy for those with laryngeal edema to benefit.
There is another important matter that influences my decision whether to give steroids to all my patients to prevent laryngeal edema: I am not used to deciding that a patient is ready to be extubated 12-24 hours in advance. We make rounds in the morning, assess the patient's status, including the results of a spontaneous breathing trial, and decide on extubation right then based on that information. Waiting until that evening or the next morning to carry out the extubation so that several timed doses of steroids could be administered would prolong the period of intubation for a large number of patients who would not benefit from that therapy. Given that a single bolus of steroids immediately prior to extubation does not seem to be effective, the regimen used in the studies included in the meta-analyses (which is either effective or ineffective, depending on which of the latter you prefer) seems ill-suited to current ICU practice.
The use of corticosteroids prior to a second extubation attempt in a patient who had stridor during an earlier failed extubation makes sense. So does their administration to patients who had difficult intubations, who have unusually large endotracheal tubes for their size (particularly women), who have sustained airway injuries or trauma to the head and neck, or who have no cuff leak on repeated measurements although the required delay while several doses of steroids are administered needs to be factored into the clinical decision. However, I remain unconvinced that more liberal administration of steroids to intubated patients to diminish the likelihood of post-extubation stridor is currently justified by the available evidence.
1. Markovitz BP, et al. Corticosteroids for the prevention and treatment of post-extubation stridor in neonates, children and adults. Cochrane Database Syst Rev. 2008;(2):CD001000.
2. Young D, Watkinson P. Preventing postextubation airway complications in adults. BMJ. 2008;337:a1565.