Long‐Term Treatment Patterns in Norwegian ADHD ‐Medicine Users: A Cohort Register‐Based Study With 5 Years of Follow‐Up, 2012–2021

ABSTRACT Purpose To improve interpretation of long‐term ADHD medication treatment trajectories in Norway by examining sex‐ and age‐specific persistence, distinguishing within‐substance product switching from active‐ingredient (ATC substance) switching, and describing the medication and prescriber context of the last observed dispensation before non‐persistence. Methods This study used population‐based data from the Norwegian Prescription Database (NorPD) to identify 33 706 incident users initiating ADHD medication in 2012–2016, defined using an individual‐specific 5‐year look‐back. Participants were followed for up to 5 years until 2021. We estimated dispensing‐based persistence, defined non‐persistence as the first dispensing gap of ≥ 396 days/13 months, characterised switching as either within‐substance product switching or active‐ingredient (ATC substance) switching, and described the last dispensed medication and prescriber specialty recorded before non‐persistence. Results The cohort included 19 898 men (59%) and 13 808 women (41%); 97%–98% initiated methylphenidate. Five‐year persistence was 39% in men and 35% in women. During follow‐up, 41% had no switch and 26% had ≥ 4 switches. Among first switches, within‐substance product switching predominated (80.9% in men; 77.5% in women), while active‐ingredient switching was less common (19.1% and 22.5%, respectively). Among those who met the non‐persistence definition during follow‐up (62.4%), the last dispensed medication was methylphenidate (76%), atomoxetine/guanfacine (12%), lisdexamfetamine (7%), and dexamphetamine (3%), respectively; the last prescriber specialty recorded on the last dispensation was a general practitioner for 51.9% and a specialist for 48.1%. Conclusions In Norway, most incident ADHD medication users met a dispensing‐based definition of non‐persistence within 5 years. Switching was frequent, but most first switches occurred within the same active substance rather than between substances, suggesting that many observed changes in dispensing data may reflect product substitution or formulation change rather than a true change in pharmacological strategy. These findings may improve interpretation of long‐term treatment trajectories.