Infection and lockdowns resulting from COVID-19 have been suggested to increase the prevalence and treatment rates of Attention Deficit/Hyperactivity Disorder (ADHD). To accurately estimate the pandemic's effects, pre-pandemic data can be used to estimate diagnosis and treatment rates during the COVID-19 years as if the COVID-19 pandemic did not occur. However, accurate predictions require a broad dataset, both in terms of the number of cases and the pre-pandemic timeframe. In the current study, we modeled monthly ADHD diagnosis and treatment rates over the 18 years preceding the COVID-19 pandemic. The dataset included ∼3 million cases for individuals aged 6 to 18 from the Clalit Health Services' electronic database. Using a trained model, we projected monthly rates for post-lockdown and post-infection periods, enabling us to estimate the expected diagnosis and treatment rates without the COVID-19 pandemic. We then compared these predictions to observed data, stratified by age groups, gender, and socioeconomic status. Our findings suggest no influence of the COVID-19 pandemic on ADHD diagnosis or treatment rates. We show that a narrower timeframe for pre-COVID-19 data points can lead to incorrect conclusions that COVID-19 affected ADHD diagnosis rates. Findings are discussed, given the assumed impact of the COVID-19 pandemic on ADHD.
Keywords: ADHD; Adolescents; COVID-19; Children; Diagnosis; Machine-learning; Prevalence.
Copyright © 2024. Published by Elsevier B.V.