Design: Nationwide retrospective cohort with longitudinal trajectory clustering
Population: 10,078 biologic-naive UK adults with severe asthma meeting biologic exacerbation criteria
Primary outcome: Exacerbation trajectories after first meeting eligibility
Effect: Remitting-stable 70%; chronic-frequent 20%; refractory-frequent 10%

A nationwide UK electronic records study, published in Chest on 5 October, followed 10,078 biologic-naive adults with severe asthma from the first year they had three or more steroid-treated exacerbations, the point at which they became eligible for a biologic.

Three trajectories emerged. Seventy per cent improved to fewer than three exacerbations a year on standard care, even though inhaled steroid doses fell. Twenty per cent kept exacerbating, about five times a year. Ten per cent had refractory disease with more than ten oral steroid courses a year. The persistently frequent group had higher blood eosinophils, and this was visible years before they met biologic criteria.

This matters for two reasons. Much of the improvement seen after starting a biologic may be regression to the mean, so response should be judged against that background. And a single bad year is not, on its own, a reliable sign of who needs a biologic.

For practice, that means optimising standard care, adherence, inhaler technique and comorbidities, and reviewing eosinophil history, before and alongside a biologic decision. In India, where biologic cost is a real barrier, this supports careful selection.

Check adherence, inhaler technique and comorbidities before starting a biologic for severe asthma.
Review the patient's eosinophil history over several years, not just the latest count.
Expect some improvement in the year after a bad year even without a biologic.
Set clear response criteria when starting a biologic and reassess at a fixed point.
Prioritise biologics for persistent frequent exacerbators with raised eosinophils.

Why it matters

It challenges how much of early biologic response is the drug, and refines who most needs one.

The statistics, in plain English

Trajectory groups come from clustering patterns in records, not from randomisation. The 70% who improved reflects regression to the mean as well as standard care: people selected in their worst year usually do better the next.