Diagnosis validation and clinical characterization of atopic dermatitis in Nurses' Health Study 2. (18th February 2019)

• Record Type: Journal Article
• Authors: Drucker, A.M.; Cho, E.; Li, W.‐Q.; Camargo, C.A.; Li, T.; Qureshi, A.A.
• Is Part Of: Journal of the European Academy of Dermatology and Venereology. Volume 33:Number 3(2019)
• Page Start: 588
• Page End: 594
• Publication Date: 2019-02-18
• Subjects: Dermatology -- Periodicals; Sexually transmitted diseases -- Periodicals; 616.5
• DOI: 10.1111/jdv.15360
• Languages: English
• ISSNs: 0926-9959

Abstract

Abstract: Background: Epidemiologic studies of atopic dermatitis (AD) are often limited by case definitions that have not been validated. Objective: In this study, we assessed the accuracy of self‐report of AD in a large cohort of US female nurses, the Nurses' Health Study 2 (NHS2). We also provide clinical characteristics of AD in the cohort. Methods: We sent an electronic questionnaire to NHS2 participants who previously reported ever having a diagnosis of AD. This questionnaire was designed to confirm cases of AD using previously validated algorithms with >85% specificity. We assessed the association of AD with asthma, comparing the results when different definitions of AD were applied. We also inquired about various aspects of participants' AD. Results: Responses were received from 2509 of 5126 (49%) nurses who were sent the questionnaire, with an average age of 62. Most participants (1996/2509, 80%) reiterated their previously reported clinician diagnosis of AD. Application of the two diagnostic algorithms yielded confirmation of 1538 and 1293 prevalent cases, respectively. The association of AD with asthma was stronger when more stringent AD case definitions were applied. Participants generally reported mild disease (92% with ≤10% maximal body surface area involved) and a high proportion (57%) reported adult‐onset disease. Conclusions: Self‐report of AD diagnosis has good reliability, and future analyses will be strengthened by our ability to conduct sensitivity analyses with refined confirmed AD subgroups.