The "fourth disease" of childhood: reevaluation of a nonexistent disease.

Observed and described between 1884 and 1900, "fourth disease" (Dukes disease) followed measles, scarlet fever, and rubella as the fourth clinically characterized childhood exanthem. Like rubella ("third disease") and erythema infectiosum ("fifth disease"), accepted by the medical community at about the same time, the existence of fourth disease was initially controversial. Over the following decades descriptions of hundreds of cases, outbreaks, and laboratory studies were published in the indexed medical literature. Unlike rubella and fifth disease, however, fourth disease was not subsequently proven to exist by either epidemiologic criteria or isolation of an etiologic agent. By the 1930s, it was infrequently recognized and by the 1960s had been dropped from textbooks. In this study, the authors use epidemiologic methods to reevaluate published data on English schoolchildren from 1892 to 1900 upon which the original fourth disease claim of Dukes was based. The authors conclude that fourth disease never existed. Reinterpretation of the original data suggests that cases can be completely explained as misdiagnosed rubella and scarlet fever. Misidentification of fourth disease is attributed to failures in the critical abilities of the medical and scientific communities at the time. The implications of erroneously identifying a nonexistent disease suggest that modern scientific approaches to disease identification are sound.