In Reply.—I am pleased to learn of the project Dr Baker performed during his residency, and my bias is that the conclusion is probably correct: "If family history discloses a recent minor viral infection, then the probability is increased that the infant is not septic."
Translation into action is tougher, however. The following observations seem warranted.
The "informal" retrospective review of cases creates opportunities for bias. Since the crux of the matter is the absence of a family history of minor viral infection in infants with bacterial infection and/or pneumonia, it is important that those infants "easily identified as being 'sick'" had the same likelihood of a family history of minor illness being recorded and thus ascertained by chart review. No mention of family illness cannot be equated with an absence of family illness. Confidence would be greatly increased by prospective questioning, in which a response regarding presence or