Design: multicentre prospective cohort study (BioDay Registry), up to 3 years of follow-up
Population: 309 children aged 16 or under with moderate to severe atopic dermatitis starting dupilumab at 7 Dutch hospitals
Primary outcome: effectiveness by EASI and IGA, safety, and drug survival
Effect: 3-year drug survival 80.1%; mean EASI 3.4 (95% CI 2.4 to 4.5); ocular surface disease in 36.6%

The Dutch BioDay registry followed 309 children aged 16 or under with moderate to severe atopic dermatitis who started dupilumab across seven hospitals, with up to three years of follow-up. Median follow-up ranged from 45 weeks in the youngest group to 89.5 weeks in the 12 to 16 year olds.

Disease control held: mean Eczema Area and Severity Index was 3.4 (95% CI 2.4 to 4.5) and mean itch score 3.6 (3.1 to 4.1) at three years. Three-year drug survival was 80.1%. Fifty children (16.2%) stopped, most within the first 18 months, for ineffectiveness (38% of stoppers), adverse events (30%) or administration problems (22%). Dupilumab-associated ocular surface disease was the commonest adverse event at 36.6%. Allergic asthma, allergic rhinitis and being aged 6 to 11 were each associated with lower discontinuation.

The age-specific signal is the part to act on. Children aged 6 months to 5 years consistently reported more itch and worse quality of life than older children throughout, and stopped more often because of the injections themselves. That is a fixable problem: injection technique, topical anaesthesia, who administers it and where, reviewed before concluding the drug has failed. And with ocular surface disease in over a third, ask about the eyes at every visit rather than waiting for a complaint.

Ask about dry, gritty or red eyes at every review; ocular surface disease affected 36.6%
In under-fives, review injection technique and setting before calling the drug ineffective
Expect most discontinuations in the first 18 months - persistence after that is high
Track itch and quality of life separately in young children; severity scores alone understate their burden
Consider dose reduction where control is sustained, as this registry did when clinically indicated

Why it matters

Most paediatric dupilumab failures in this cohort were not the biology failing - they were injections, eyes and the first 18 months.

The statistics, in plain English

Drug survival of 80.1% at three years is a composite measure of effectiveness, tolerability and practicality - a patient stays on a drug only if all three hold, which is why it predicts real-world use better than a response rate. Mean EASI of 3.4 with an interval of 2.4 to 4.5 is mild disease on average, but a mean hides the minority doing badly. This is an observational registry with no comparator, so it describes what happened on treatment rather than what treatment achieved.