Reading focus
Syllabus: How Does Psychotherapy Produce Change?
Compare the active treatments, confidence intervals, follow-up and harm measures; use the linked correction for affected details.
Why you’re reading it. Bring development, duration and uncertainty into the distinction between benefit and mechanism.
A question to consider. Why does a lack of superiority in IMPACT answer a different question from HAP’s advantage over enhanced usual care, and what can neither trial establish about how change occurs?
Why this reading
Different approaches may yield similar average outcomes while leaving questions about mechanisms, individual needs and durability unresolved. IMPACT compares three treatments in adolescent specialist services and follows outcomes beyond treatment.
Before you read
A superiority trial asks whether a treatment produces better outcomes under specified comparisons. Failure to find superiority is not proof of identical effects. An active comparator provides care; improvement within every group cannot establish what would have happened without it.
Read for
- Compare planned treatment, attendance, fidelity, co-interventions and follow-up availability.
- Interpret the primary symptom estimates and confidence intervals separately from diagnostic remission, harms and costs.
Consider afterwards
- Which causal question can randomization answer here, and which requires a different comparator?
- What additional evidence would help explain why an individual improves or does not improve?
Editorial context
The trial randomized 470 adolescents in England; 465 remained in the analysis population, with primary follow-up data from 392. There was no untreated group, and some participants received antidepressants. The 2017 correction changes secondary-outcome directions, a table legend and the CBT relapse count. The linked PMC copy retains pre-correction details: consult the notice alongside it. These comparisons do not establish a shared mechanism or rule out unmeasured harms.