Letter to the Editor
1 Centers for Disease Control and Prevention, Respiratory Virus Branch, Atlanta, GA 30333
We agree with Dr Weiss that our finding of increased odds of steroid exposure among 19 patients with severe varicella and no known immune supression should be confirmed in a larger study. A case-control design, such as the one we used,1 and not a prospective study as proposed by Reiches and Jones in their commentary2 is the proper approach because of the rarity of severe varicella.
As Dr Weiss points out, the criticism of our choice of controls,2 although true technically, is irrelevant in a practical sense.




