To the Editor.
—Wheeler et al1 reported a "real increase in the incidence of cases of severe group A streptococcus sepsis in children in Denver" based on an increase in "cases per 10 000 patient discharges." A case was defined as a patient "with one or more positive blood cultures." It is important to determine whether the incidence of discovery of septicemia in this defined medical care environment has changed over the 10-year period of this study and is therefore responsible for the observed increase in incidence rather than an actual increase in cases in Denver. Similar increases in the incidence of septicemia cases have been observed in our institution, but these disappear when intensity of discovery is controlled for.In order to control for changes in length of stay, it is necessary to calculate the incidence density for each period, which can be defined as follows: (number of