Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the child anxiety research overview.
Short answer. Anxiety disorders and OCD follow a relapsing-remitting course for many children, and a return of symptoms after a good stretch is not evidence that the original treatment failed (Ginsburg et al., 2018; Piacentini et al., 2014). The research-backed response is a brief booster course — not a full restart from baseline. Most relapses respond to four to eight booster sessions of the same evidence-based treatment, often with quicker gains the second time.
The CAMS long-term follow-up (Ginsburg et al., 2018) tracked children who had completed the original CBT, sertraline, or combination treatment six years later. Roughly half maintained remission; the rest experienced at least one return of symptoms during the follow-up window. The pattern was not failure of treatment — it was the underlying recurrent nature of paediatric anxiety disorders. POTS long-term data (Piacentini et al., 2014) shows a parallel pattern for OCD.
The clinical implication is important. Albano and colleagues' work on relapse prevention (Kendall et al., 2004; Albano et al., 2018) emphasises that a child who has previously responded to evidence-based treatment is statistically the most likely to respond again — and to respond faster. The skills are largely intact; what is needed is reactivation, not rebuilding.
The literature also distinguishes between three patterns that often get conflated: a brief return of symptoms during a stressor (transition, illness, family event); a partial relapse that erodes some functional gains; and a full recurrence that meets diagnostic criteria again. The response is different for each.
When parents describe a relapse, they usually mean one of three patterns:
1. A specific symptom that has returned. The compulsion that disappeared at end-of-treatment is back. The school refusal that resolved is returning. The reassurance loop that closed is reopening. 2. A general softening of functional gains. Nothing dramatic, but the child is doing less, withdrawing more, looking more like they did pre-treatment. 3. A frank crisis. A meltdown, a refusal, a clinical-level escalation that prompted the question "is this the whole disorder coming back?"
A return of one or two symptoms during a clearly identifiable stressor — start of school, illness, family transition, exam period — is the most common pattern and the most benign. The literature (Ginsburg et al., 2018) suggests that brief, focused use of the skills the child has already learned often resolves it within weeks without a return to full treatment. The protocol the parent should run is the one the family wrote at end-of-treatment.
A slower, more diffuse softening — some compulsions returning, accommodation creeping back in, functional participation declining over weeks — usually warrants a brief booster course (four to eight sessions). The child has the skills but needs structured reactivation. Most evidence-based clinicians will see returning patients for short booster courses.
A return that meets diagnostic criteria again — sustained, multi-symptom, functionally impairing — usually warrants a more complete treatment course, sometimes with medication added if it wasn't before, sometimes with a different therapy modality. The research suggests this group still responds well to a second course, often with faster gains than the first (Albano et al., 2018).
The strongest predictor of a benign relapse is early detection. Families that had an explicit relapse-prevention plan at end-of-treatment — written, with named early-warning signals and a pre-agreed call-back protocol — usually catch returns in pattern 1 territory before they become pattern 3. Families without a plan more often present in pattern 3.
From recent parent threads:
The third quote captures the parent's emotional reality. The research-backed answer is reassuring on the practical question: the second course is rarely as long as the first. The skills are largely intact, the family knows the language and protocols, and the response to treatment is typically faster.
1. Don't conclude failure. A relapse is a feature of the disorder, not a verdict on the original treatment. The skills the child learned are still there. 2. Run the agreed protocol. If end-of-treatment included a written relapse-prevention plan, this is the moment to use it. If it did not, write one now with the clinician. 3. Re-engage briefly, not fully. Four to eight booster sessions of the same modality is the most common evidence-based response (Ginsburg et al., 2018; Piacentini et al., 2014). 4. Track which symptoms returned and when. A simple log distinguishes stress-linked spikes from durable erosion. Patterns 1 and 2 look identical for a week and diverge over three. 5. Re-tighten accommodation reduction. Accommodation creep is one of the most common causes of slow relapse; the SPACE programme (Lebowitz et al., 2020) frames this as the maintenance work. 6. Address co-parent re-alignment. Families that allowed accommodation drift after end-of-treatment usually need the same alignment conversation again.
---
Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full child anxiety research overview for the complete framework.