Severe varicella associated with steroid use.

OBJECTIVE: To evaluate whether corticosteroid use is associated with severe varicella.
DESIGN: The odds of corticosteroid exposure were compared among 35 children with severe varicella and 10,000 control subjects.
RESULTS: Five (26.3%) of 19 case patients without known immunosuppression had received steroids within 30 days prior to the onset of their rash compared with 20 of the 10,000 control subjects (0.2%), giving an odds ratio of 178 (95% confidence interval 59 to 541). If 16 case patients with leukemia or other known immunosuppressive conditions were included, the odds ratio was larger (odds ratio 420, 95% confidence interval 189 to 935). Of the 13 case patients whose dosage was recorded, 7 received less than the equivalent of 2 mg/kg per day of prednisone. The timing of the steroid use in those who had severe varicella clustered within the incubation period for the virus.
CONCLUSIONS: Systemic corticosteroid use appears to increase substantially the risk of severe or fatal varicella. The timing of corticosteroid exposure, in addition to dose and duration, may be an important factor in determining vulnerability. Further studies are needed to determine whether short-course, low-dose, or inhaled steroids are associated with similarly increased risk. Physicians should be aware that varicella-susceptible patients receiving corticosteroids are at considerably increased risk for severe varicella and should consider how to counsel their patients.