Diagnostic criteria are made of frail stuff, especially when used in the analysis of a single case. More suitable for population descriptions and case selection in clinical trials, criterial formulations must be revised frequently as diagnostic techniques are refined and knowledge of disease processes changes.
Acute rheumatic fever, especially its presentation in the adult, is one such disease in which the evolving clinical features strain the existing criteria (1). The difficulties in applying the revised Jones' criteria to adults are emphasized in the paper by McDanald and Weisman in this issue (2). That acute rheumatic fever is a sequel to