CBT-I Trials with Psychosis Conversion as the Outcome

confidence: high · 2 sources

⚠
Working research — not a recommendation.

This page records what the evidence says a policy does, not what anyone should do about it. It has been synthesized from the cited sources but has not been independently reviewed.

What the evidence says

Cognitive behavioral therapy (CBT) shows no measured effect on conversion to psychosis in people at clinical high risk (CHR-P). The strongest direct test is a randomized trial by Morrison and colleagues (2012) in the BMJ. It compared cognitive therapy with monitoring of mental state only. Transition to psychosis was 6.9% in the cognitive therapy group and 9.0% in the monitoring group. The proportional odds ratio was 0.73 (95% CI 0.32 to 1.68, P=0.45). The interval spans both a reduction and an increase in odds, so the trial cannot say which direction the true effect runs. Confidence in the direction of "no measured effect" is high, but the magnitude is imprecise.

A 2019 review by Fusar-Poli in Frontiers in Psychiatry reaches the same result from a wider base. It reports that two large randomized trials found neither cognitive behavioral therapy nor Omega-3 interventions reduced progression to psychosis from a CHR-P state. It also reports that, across seven independent meta-analyses, there was no evidence to favour any preventive intervention over any other or over a control condition. That claim covers CBT as one of several interventions, so it speaks to CBT only as part of a group.

A simulated living meta-analysis in the same review compared CBT with needs-based intervention. At both 6 and 12 months, every update reported results that never crossed the monitoring boundaries. The accrued sample also never reached the information size needed to detect the assumed risk ratio reduction of 0.5. This shows the available trials are underpowered for large effects. It does not rule out a small effect.

The same Morrison trial measured symptoms, and the picture there differs by outcome. At 12 months, cognitive therapy decreased the severity of psychotic symptoms on the CAARMS scale by 3.67 points (95% CI −6.71 to −0.64, P=0.018) relative to monitoring only. The claims do not say whether that gap changes later outcomes. Distress from psychotic symptoms showed no measured effect. The estimated difference was −3.03 points (95% CI −6.95 to 0.94, P=0.14).

The claims name no causal mechanism by which CBT would lower or fail to lower conversion. They establish only the measured outcomes.

Where the evidence splits

No contradictions in the current sources.

Thresholds and dose-response

No threshold established in the current sources.

Open empirical questions

Findings

CBT shows no measured effect on transition to psychosis in people at clinical high risk

confidence — direction: highconfidence — magnitude: lowconsensus: settled. All sources point to no measured effect, though the trial evidence is imprecise.
mechanism
not established in the sources
effect
no measured effect. In the Morrison trial, transition was 6.9% with cognitive therapy and 9.0% with monitoring only (proportional odds ratio 0.73, 95% CI 0.32 to 1.68, P=0.45). The review and the simulated living meta-analysis also report no measured effect.
conditions / moderators
the trial's confidence interval crosses the null, so a wider sample is needed to narrow the true effect. The simulated meta-analysis assumed a risk ratio reduction of 0.5, and the available trials were underpowered for it. Population: CHR-P individuals.

No preventive intervention, including CBT, outperforms another or a control for preventing psychosis onset

confidence — direction: highconfidence — magnitude: highconsensus: settled. The review reports no evidence to favour any intervention.
mechanism
not established in the sources
effect
no measured effect separating CBT from other interventions or control conditions, based on convergent findings across seven meta-analyses with AMSTAR ratings of 8-10/11
conditions / moderators
not established

Cognitive therapy decreases the severity of psychotic symptoms at 12 months

confidence — direction: highconfidence — magnitude: mediumconsensus: thin. Only one trial is reported here.
mechanism
not established in the sources
effect
decreases CAARMS severity by 3.67 points relative to monitoring only (95% CI −6.71 to −0.64, P=0.018)
conditions / moderators
the treatment-by-time interaction showed no difference. The effect is the 12-month gap between arms. Replication in independent samples is needed to confirm the size.

Cognitive therapy shows no measured effect on distress from psychotic symptoms

confidence — direction: highconfidence — magnitude: lowconsensus: thin. Only one trial is reported here.
mechanism
not established in the sources
effect
no measured effect. The estimated difference at 12 months was −3.03 points (95% CI −6.95 to 0.94, P=0.14).
conditions / moderators
the confidence interval crosses the null