Sir.—In the July 1989 issue of AJDC Farley et al1 reported a case cluster of Henoch-Schönlein purpura (HSP) occurring in Connecticut during 1987-1988. They suggested that the cluster may have been caused by person-to-person spread of an infectious agent of the respiratory tract to susceptible hosts.
They cited my conclusion from an epidemiological study of 1222 Danish children with HSP2 that there was no clustering. They continued: "In contrast, we were able to demonstrate a case cluster limited in time and space and centered in one ethnic group."
There is no contrast between our findings. The lack of clustering found in my study does not imply that clusters do not exist, but only that they do not occur more frequently than expected by chance.
The strong association between HSP and previous upper respiratory tract infections,1-3 the lack of contagiousness of HSP, and the negative serological findings