Answers are grounded in this paper and cite their sources.
BACKGROUND: Precise skin phenotypic data are indispensable in accurately diagnosing atopic dermatitis (AD). Therefore, this study examined the interobserver concordance for AD and non-AD diagnoses between two dermatologists. AD prevalence determined by the self-reported physician diagnoses and the diagnoses determined from the United Kingdom (UK) diagnostic criteria were compared with the diagnoses made by the two dermatologists, using data from a skin health survey. METHODS: This study included 1,638 children that participated in the skin health survey, which was part of the Tohoku Medical Megabank Project Birth and Three-Generation Cohort Study. AD was assessed using dermatologist assessments, self-reported physician diagnoses, and the UK diagnostic criteria. The concordance for diagnoses was evaluated using kappa. The sensitivity and specificity of the self-reported physician diagnoses and the UK diagnostic criteria were calculated by comparing them with the two dermatologists' diagnoses. RESULTS: Among the 1,638 children, 393 (24.0 %), 194 (11.9 %), and 597 (37.2 %) were diagnosed with AD by the two dermatologists, physicians, and the UK diagnostic criteria, respectively. The kappa (95 % CI) of the interobserver concordance for AD or non-AD diagnoses between the two dermatologists was 0.78 (0.75-0.81). The sensitivity and specificity of the self-reported physician diagnoses were 26.7 % and 94.1 %, respectively. The sensitivity and specificity of the UK diagnostic criteria were 85.0 % and 82.4 %, respectively. CONCLUSIONS: Interobserver concordance for AD or non-AD diagnoses between the two dermatologists was substantial. Self-reported physician diagnoses exhibited low sensitivity that potentially indicated underdiagnosis of AD, whereas the UK diagnostic criteria might overdiagnose AD.
No takes yet. Share an insight, caveat, or question.
Kawaguchi et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: