Sir.—The report by Halsey and associates (Journal 134:521-522, 1980) deserves critical review. They observed an infant with clinical pertussis, the culture being positive; after only seven days of erythromycin therapy, the infant was found on the eighth day to have a positive culture for Bordetella pertussis. The fluorescent antibody (FA) test on the same sample was negative. This does not necessarily indicate an erythromycin treatment failure.
Short courses of erythromycin are often followed by bacteriologic relapse. We observed such a relapse in the first of ten patients treated with erythromycin in our study of pertussis.1 This patient had negative culture and FA test after 48 hours of treatment. She was treated for ten days, and one day after stopping erythromycin therapy she again had positive culture and FA test; this persisted for four days. We did not (and have not since) observed a bacteriologic relapse in patients treated for