Prevalence of Type 2 Inflammatory Diseases in Pediatric Patients With Atopic Dermatitis: Real-world Evidence

Abstract

To analyze the prevalence and incidence of type 2 inflammatory disease in pediatric patients with atopic dermatitis (AD) compared with a similar population of patients without AD, utilizing a large United States commercially insured database.The study included 244 776 pediatric AD patients matched 1:1 with non-AD patients from the IBM MarketScan Commercial Database from 2013 to 2017. This US database included representation from a variety of commercial insurance plan types throughout the country with 44% of patients from the South.This was a retrospective study conducted over a 12-month period that used a commercial database of select the patient populations. This database included diagnostic codes entered by clinicians for billing from a variety of providers, including primary care (pediatricians and family practitioners), subspecialists (allergy and dermatology), and emergency department or urgent care clinicians. Date of AD diagnosis was used for patients with AD and a random date for non-AD patients with required insurance 6 months prior and 12 months after date. AD patients were defined as <18 years with ICD 9 code (691.8) or ICD 10 code (L20×) and matched controls had no diagnosis of AD for the entirety of the study. Severity of disease was divided into 3 levels based on treatment. Level 1 included low potency topical steroids (class 5–7) or calcineurin inhibitors, level 2 included mid potency topical steroids or calcineurin inhibitors, and level 3 includes systemic steroids, phototherapy, immunosuppressants, or intravenous immunoglobulin. The prevalence of type 2 diseases including asthma, allergic conjunctivitis, chronic rhinosinusitis, nasal polyps, eosinophilic esophagitis, allergic rhinitis and urticaria was evaluated in 2 comparison groups: severity of disease (level 3 AD versus level 1 and 2) and presence or absence of AD. Furthermore, food allergy was omitted from the evaluation as it was felt it was poorly defined and without standardization. Logistic regression models were used for statistical comparison. Kaplan-Meier curves were used to show time to development of type 2 disease. Hazard ratios using a Cox proportional hazard model were reported for the risk of type 2 disease in AD versus non-AD patients.In this population database, patients who had AD at time of diagnosis (index) were more likely to develop type 2 inflammatory disease compared with patients without AD (odds ratio [OR] of 2.17, [2.15–2.20], P < .001). Asthma, eosinophilic esophagitis, urticaria and rhinitis were among conditions that were twice as likely diagnosed in AD patients. Chronic rhinosinusitis and nasal polyps had a slightly higher likelihood in AD patients (OR 1.10, [1.06–1.13]; P < .001). Patient with level 3 severity had 4 times the likelihood of asthma (OR 4.43, [4.32–4.54]; P < .001). Infants with AD demonstrate a higher risk for development of type 2 disease, with a hazard ratio of 1.47 (1.45–1.50) compared with non-AD patients. Based on the Kaplan-Meier curves, patients with AD had a 51.7% probability of developing type 2 inflammatory disease compared with non-AD patients who had a 38.9% probability.The relative prevalence of type 2 inflammatory disease was higher in pediatric patients with AD in a very large insurance-based database. In addition, the risk of developing type 2 inflammatory disease was higher in infants with AD over a 1-year period. Severity of AD was found to increase both the prevalence and risk.This study used a large national insurance database to examine the prevalence and incidence of type 2 inflammatory diseases (asthma, rhinitis, conjunctivitis, urticaria, eosinophilic esophagitis, and chronic rhinosinusitis or nasal polyps) in pediatric patients with AD. Authors analyzed an impressive cohort size and found a statistically significant higher incidence and prevalence of other allergic diseases within pediatric AD. There were a few limitations to further consider. First, authors were unable to quantify food allergy in this cohort due to concerns of nonstandardization of food allergy diagnostic criteria and associated codes. Though quantifying food allergy was a challenge, perhaps adding epinephrine prescription or food anaphylaxis may have been helpful. Secondly, the scoring of AD severity was based upon treatment regimens which, consequently, grouped all ages into 1 cohort. Patients on a high potency topical steroid were all considered level 2 (on a 1 to 3 scale), whereas in many cases an infant or toddler receiving this treatment would be considered severe. In addition, it was difficult to know if eczema was indeed the indication for the prescriptions. For example, in patients considered severe or level 3, treatments such as systemic steroids may have been given for asthma exacerbation and immunoglobulin replacement may have been used to treat an underlying immunodeficiency. Finally, extending the 1-year observation period would have provided a more complete evaluation of the incidence of other atopic conditions in infants and toddlers less than 2 years of age. This retrospective study added to work establishing an increased burden of type 2 inflammatory diseases found in patients with AD.URL: www.pediatrics.org/cgi/doi/10.1542/peds.10.1016/j.jaad.2021.10.038