—Dr Nielsen analyzed hospitalized patients with HSP in Denmark and did not find statistical evidence of space-time clustering. In our study we identified all children given the diagnosis of HSP in a geographic area, and we described a case cluster that was clearly defined in space, time, and race/ethnicity. There are many differences between the two studies, and we agree with Dr Nielsen that there is no contradiction between the two. However, Dr Nielsen's statistical findings do not preclude the existence of a group of etiologically linked cases occurring with unusual frequency in one population subgroup during one period of time. It is this type of epidemiologic cluster that we have described.
We concur with Dr Nielsen's speculation that several different infectious agents may trigger HSP. However, even if many different agents have this capability, the cluster that we investigated may still have been caused by a single agent. We