PSY17 The Prevalence and Clinical Profile of Atopic Dermatitis (AD) in England: A Population Based Linked Cohort Study Using Clinical Practice Research Datalink (CPRD) and Hospital Episode Statistics (HES)

Abstract

To describe the prevalence and clinical profile of Atopic Dermatitis in England by severity. A descriptive population-based database analysis was conducted using CPRD and HES linked data between 2008-2018. Adults (≥18y) and adolescents (12-17y) with a diagnosis of AD were identified using read codes, ICD-10 codes and the receipt of AD therapies. Comorbidities and symptoms were defined using read codes or ICD-10 codes. AD severity is not routinely recorded therefore severe disease was defined as receipt of ≥1 systemic immunosuppressant, phototherapy or referral for AD, moderate as ≥2 potent topical corticosteroid prescriptions in 1 year or ≥1 topical calcineurin inhibitor, and mild as any patient not meeting moderate or severe criteria. Patients may move across age groups and severity categories. There were 252,589 patients with AD, with a mean of 2,065 days follow-up. Period prevalence of AD was 2.01%, (4.99% adolescent, 1.86% adults) of which 52.23% was mild, 46.56% moderate and 3.11% severe. The most common comorbidities recorded were asthma (23.95%) depression (22.67%) and conjunctivitis (22.38%); all of which increased with AD severity (34.21%, 25.28% and 27.09% in severe AD respectively). Notably, 1.17% had inflammatory bowel disease recorded (0.71% in mild to 7.99% in severe AD). Over the cohort 31.27% had a skin infection recorded. There was a mean of 0.32 (Standard deviation ±1.13) skin infections per patient, increasing from 0.20 (SD± 0.80) in mild AD, to 0.44 (SD± 1.47) in severe AD. Recorded symptoms included pruritus (21.45%), xerosis (17.13%) and sleep disturbance (13.03%). This analysis of linked national datasets provides an initial overview of the epidemiology and clinical burden of AD indicating an increased clinical burden, including symptoms and comorbidities, with increasing severity. The limitations associated with the analysis mean that the results presented are likely an underestimate of the true burden of AD.