IN THIS issue of the ARCHIVES, Firooz et al1 present some data on the validity of the U.K. Working Party's Diagnostic Criteria for Atopic Dermatitis (AD),2- 4 which are based on the criteria of Hanifin and Rajka,5 in a sample of 416 patients of all ages attending a private dermatology clinic in Iran. Firooz et al1 found a sensitivity and specificity of 10% and 98%, respectively, compared with the clinical diagnosis of one senior private dermatologist. The authors state that although the high specificity of the criteria makes them suitable for clinical trial work, the poor validity of the criteria in this study makes them unsuitable for epidemiological use. For example, if the validity indices of this study are true, then in a population setting in which the true prevalence of AD is 10%, the positive predictive value of the criteria (proportion of those with a positive test result who really have the disease) would be only 36%.6 In other words, around two thirds of those diagnosed as having AD using the criteria would not actually have AD. So, where do we go from here?
Williams HC. Diagnostic Criteria for Atopic DermatitisWhere Do We Go From Here?. Arch Dermatol. 1999;135(5):583–586. doi:10.1001/archderm.135.5.583