Comparing Self-Reported and Physician-Assessed Acne Severity Using Real-World Data

Background: Real-world data (RWD) offers access to vast amounts of health information; however, limited knowledge exists regarding its reliability in clinical and research practices. Objectives: To analyse the agreement between self-assessed severity of self-reported acne vulgaris (AV), and physician-assessed severity based on photographic evaluation using consumer-generated RWD. Methods: RWD was accessed through a skincare company's (NØIE) database. We collected data from individuals with self-reported AV who had submitted a questionnaire, including self-assessments of AV severity, and photographs of affected skin taken at least 14 days apart. A resident dermatologist, trained in assessing AV lesions from photographs, evaluated the photographs, counting visible AV lesions to determine the severity. Agreement in the direction of change between self-assessed and physician-assessed severity was calculated using data sets from all pairs of consecutive assessments, and Pearson correlation coefficients (PCC) were subsequently applied. Results: Out of 1691 eligible individuals, 276 (16%) were included in the study. The selected individuals were mainly white females in their late twenties. Agreement between self-assessment and physician-assessment regarding the direction of change in severity was observed in 48% of cases. The overall PCC between physician-assessment and self-assessment was 0.22 (95% CI 0.12–0.32), indicating a weak agreement. A slightly higher PCC was observed in females, individuals > 35 years old, individuals with darker white skin, and when photographs showcased AV lesions on the forehead. Conclusions: The study reveals a weak agreement between self-assessed severity of self-reported AV and physician-assessed severity using consumer-generated RWD. This emphasizes the need for caution when interpreting such consumer-generated RWD in clinical and research dermatology.