Cognitive behavioural therapy works well for OCD. A study of 120 people treated for the condition adds an important qualification: not every kind of improvement holds.
Researchers followed patients for a full year after they finished therapy. Those who had reduced their OCD symptoms without reaching full remission were nearly nine times more likely to slide backwards than those who had recovered fully. Remission here meant symptoms had fallen to a level low enough that they no longer interfered significantly with daily life.
In plain terms, finishing therapy feeling somewhat better carries a considerably higher chance of the OCD returning. Working through to genuine remission makes the gains far more likely to last, still measurable a year on.
This challenges a comfortable assumption, which is that good enough is good enough. For OCD the evidence points towards full recovery as the appropriate goal. In practice that can mean staying in therapy a little longer, or working on the harder exposures that are easiest to postpone.
If OCD is affecting your life, a psychologist experienced in exposure-based CBT can help you reach that point rather than stopping short of it.
What the study did
Elsner et al. followed 120 people with obsessive-compulsive disorder (OCD) for a full year after they finished individual cognitive behavioral therapy (CBT) at a university outpatient clinic. This was an intention-to-treat study, meaning everyone who started treatment was included in the analysis, even if they dropped out early — a design that gives a more realistic picture of how treatment performs in practice, rather than only looking at people who completed everything perfectly.
Symptom severity was measured using the Yale-Brown Obsessive Compulsive Scale (Y-BOCS), a standard clinician-rated interview that scores the severity of obsessions and compulsions. The researchers applied two internationally agreed benchmarks: 'response' (a Y-BOCS score reduction of at least 35% from baseline) and 'remission' (a Y-BOCS score of 12 or below, roughly indicating symptoms are no longer clinically significant). They then tested how well each of these end-of-treatment states predicted what happened over the following year.
What it found
People who had achieved only 'response' — a meaningful drop in symptoms, but not remission — were far more likely to relapse than those who had reached remission. Specifically, the odds of deteriorating during the follow-up year were 8.8 times higher in the response-only group. An odds ratio like this compares the chance of an outcome (here, deterioration) between two groups; a value of 8.8 means the odds were nearly nine times greater for those who hadn't fully remitted.
The team also used a statistical tool called a receiver-operating characteristic (ROC) curve, which helps identify the score that best separates two outcomes — in this case, who would later deteriorate and who wouldn't. This analysis showed that a post-treatment Y-BOCS score of 13 or higher best distinguished those at risk of later worsening. A slightly different cut-off, a score of 12, best identified who was likely to hold onto any gains at all (whether response, remission, or both) a year later. Encouragingly, the study also reconfirmed something seen in earlier research: OCD symptom improvements after CBT are generally quite stable over time.
Why it matters
These findings give clinicians a concrete, evidence-based marker to aim for during treatment, rather than relying on a vague sense that someone is 'doing better.' Because the consensus remission threshold (Y-BOCS ≤ 12) so closely matched the cut-offs that best predicted long-term stability, it suggests this isn't just an arbitrary number — it reflects a meaningful difference in how durable someone's recovery is likely to be. For therapists, this supports continuing or intensifying treatment (for example, working through harder exposure exercises) when a client has improved but hasn't yet reached this threshold.
Its limits
The abstract doesn't report on what happened beyond one year, so it can't tell us about longer-term outcomes. It also doesn't specify why some people reached remission and others didn't, so this study identifies a useful predictor without explaining the underlying mechanism. As with any single study, these cut-offs would benefit from replication in other clinics and populations.
If OCD is part of your life, know that structured, exposure-based CBT with a psychologist can help you work toward full remission — not just partial improvement — and reaching out for that support is always a reasonable and worthwhile step.