Acute Respiratory Failure in Patients With Toxic Epidermal Necrolysis: Clinical Features and Factors Associated With Mechanical Ventilation

Objectives: Stevens-Johnson syndrome and toxic epidermal necrolysis are severe adverse cutaneous drug reactions characterized by widespread skin and mucous membrane detachments, including bronchial mucosa, which may be associated with respiratory failure requiring mechanical ventilation. The presentation and outcome of patients requiring mechanical ventilation and the characteristics of bronchial epithelial lesions among ventilated patients are reported. Predictors of mechanical ventilation available on hospital admission were identified using univariate and multivariate logistic regressions. Design: Retrospective cohort study. Setting: Medical ICU and dermatology department of a tertiary care hospital, which hosts the French national referral center for toxic epidermal necrolysis. Patients: Patients admitted for Stevens-Johnson syndrome/toxic epidermal necrolysis over a 14-year period were included. Interventions: None. Measurements and Main Results: Of the 221 patients included in the study, 56 patients (25.3%) required mechanical ventilation. None of the patients received noninvasive ventilation. Patients requiring mechanical ventilation had a larger baseline detached body surface area, higher Logistic Organ Dysfunction score, and Simplified Acute Physiology Score II, and they presented more often with shock, pulmonary infiltrates, and renal dysfunction (p < 0.0001 for all comparisons). Among patients receiving mechanical ventilation, 57% of the patients died; those having bronchial epithelial lesions (22 of 56) required intubation earlier than others (1 [1-4] vs 4 [1-6] d after hospital admission; p = 0.027). Variables associated with mechanical ventilation in multivariate analysis included serum bicarbonates less than 20mM (odds ratio, 4.9 [95% CI, 1.1-22.7]; p = 0.041), serum urea greater than 10mM (odds ratio, 7.0 [95% CI, 2.2-22.8]; p < 0.001), a detached body surface area between 10% and 29% (odds ratio, 3.7 [95% CI, 1.0-13.8]; p = 0.048) or greater than or equal to 30% (odds ratio, 19.7 [95% CI, 4.4-87.4]; p < 0.0001), WBCs more than 12,000/mm(3) (odds ratio, 11.6 [95% CI, 2.8-48.1]; p < 0.001), blood hemoglobin less than 8g/dL (odds ratio, 8.1 [95% CI, 1.2-55.2]; p = 0.032), and more extensive pulmonary infiltrates (odds ratio, 9.7 [95% CI, 3.6-25.9]; p < 0.0001). Conclusions: Mechanical ventilation is required in one of four Stevens-Johnson syndrome/toxic epidermal necrolysis patients and is associated with a poor outcome. Prompt identification of Stevens-Johnson syndrome/toxic epidermal necrolysis patients at higher risk of intubation could help guide their early management, particularly for those having bronchial epithelial lesions.