EVALUATION OF RISK-FACTORS FOR LARYNGEAL EDEMA AFTER TRACHEAL EXTUBATION IN ADULTS AND ITS PREVENTION BY DEXAMETHASONE - A PLACEBO-CONTROLLED, DOUBLE-BLIND, MULTICENTER STUDY

Because laryngeal edema (LE) after tracheal extubation is likely to result from an exudative response, corticosteroids often are given routinely as a preventive treatment. No adequate controlled study supports this strategy, however. A prospective, randomized, placebo-controlled, double-blind, multicenter trial that included 700 consecutive patients requiring tracheal intubation and mechanical ventilation was conducted to determine risk factors for LE occurrence after tracheal extubation in adults and to evaluate the efficacy of corticosteroids in its prevention. One hour before extubation, patients were given either an intravenous bolus of 8 mg dexamethasone or a placebo. Patients were divided into two groups: 1) those in whom short-duration intubation (SDI, less than 36 h) was administered; and 2) those in whom long-duration intubation (LDI, more than 36 h) was administered. Minor LE was diagnosed when either stridor or laryngeal dyspnea, or both, occurred; major LE was diagnosed when reintubation due to LE was required, with LE evidenced during direct laryngoscopy. The overall incidence of LE was 4.2% and varied among the six participating centers from 2.3 to 6.9% (not significant). In only seven patients (1%), all with LDI, was tracheal reintubation required for LE. Laryngeal edema occurred more frequently after LDI than after SDI (7.2 vs. 0.9%; P < 0.001). It also was more frequent in female than in male patients (20/284 vs. 8/379; P < 0.05), irrespective of intubation duration and treatment. There was no association between LE and either difficulty/route of intubation or admission diagnosis. Its overall incidence was not different between patients given dexamethasone and those given placebo, whether they were in the SDI or LDI group (0.7 vs. 1.2% and 5.7 vs. 8.7%, respectively). Dexamethasone did not affect the number of patients reintubated. Assuming an incidence of LE in each group similar to that found in our study, more than 7,000 patients would need to be studied to evaluate adequately the presence of a type 2 error. Therefore, we conclude that the overall incidence of postextubation LE in adults is small and that risk factors for LE include an intubation duration of more than 36 h and female gender. In addition, 8 mg dexamethasone given as an intravenous bolus 1 h before extubation does not appear to prevent either minor or major forms of LE.