Sequencing decisions in paediatric anxiety are made without evidence: which treatment first, and what to do when three months of it has not produced remission. This pragmatic sequential multiple assignment randomised trial addressed both.
It ran in primary care and mental health clinics over 24 weeks in two stages. In stage one, 316 youths aged 8 to 17 with a DSM-5 anxiety disorder, not already in treatment, were randomised to fluoxetine or exposure-based cognitive behavioural therapy. At 12 weeks, non-remitters were randomised again to continue the initial treatment or to add the other modality. The primary outcome was the youth-reported 41-item SCARED. The cohort had severe anxiety, high sociodemographic disadvantage and frequent co-occurring diagnoses.
Youth-reported SCARED scores fell 31.7% overall. Initial treatment did not matter significantly, though CBT held a non-significant advantage (24-week difference in mean change 1.45, 95% CI -2.25 to 5.16). Among week-12 non-remitters, combination treatment was no better than continuing monotherapy (difference -2.74, 95% CI -6.53 to 1.05). Starting with CBT and then adding medication did separate from other sequences on a subset of measures.
A subgroup finding deserves flagging without being over-read: non-Hispanic White youths did better starting and continuing fluoxetine, while youths from racial and ethnic minority groups did better transitioning to combination therapy after week 12.
The practical message is liberating rather than deflating. If the options perform similarly, the choice can be made on what the family will accept and what is actually available — and in most of India, that means fluoxetine, because exposure-based CBT delivered by a trained therapist is not obtainable for most children.
- Start with whichever of fluoxetine or exposure-based CBT the family prefers and can access; outcomes were similar
- Do not automatically add the second modality at 12 weeks for non-remitters — combination was not superior to continuing monotherapy
- Expect roughly a third reduction in youth-reported symptoms over 24 weeks with any of these approaches
- Treat the ethnicity subgroup finding as hypothesis-generating rather than as a prescribing rule
- Where CBT is unavailable, fluoxetine monotherapy is a defensible first choice on this evidence
The statistics, in plain English
Both key confidence intervals cross zero and are narrow enough to be informative rather than merely inconclusive: the initial-treatment comparison spans -2.25 to 5.16 SCARED points, and the combination comparison -6.53 to 1.05. On a 41-item scale these ranges exclude any difference a clinician would notice. The 31.7% overall decline is not a treatment effect — there was no untreated arm, so it includes natural improvement, regression to the mean and the effect of being in a trial. The claim that CBT followed by combination 'significantly separated' on a subset of measures is a secondary finding across many outcomes and should be treated as exploratory.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free