- Design
- Systematic review and meta-analysis of 23 RCTs
- Population
- 2,016 adults with schizophrenia-spectrum disorders and auditory hallucinations
- Primary outcome
- Hallucination severity at end of treatment
- Effect
- SMD −0.18 (−0.31 to −0.05); NNT 16; avatar therapy SMD −0.37
This meta-analysis pooled 23 randomised trials covering 2,016 adults with schizophrenia-spectrum disorders, testing five types of psychological or psychosocial intervention designed specifically to target auditory hallucinations, against control conditions.
At the end of treatment, targeted interventions reduced hallucination severity by a small margin (standardised mean difference −0.18, 95% CI −0.31 to −0.05), equivalent to treating about 16 people for one to benefit. GRADE certainty was low. Effects extended to frequency and distress of voices, positive symptoms, depression and anxiety. Avatar therapy — in which the person converses with a computer-rendered representation of the voice — showed the largest effect (SMD −0.37). Dropout was low.
These therapies are an add-on to antipsychotics, not a substitute, and avatar therapy is not widely available. The practical point is that persistent distressing voices despite adequate medication are worth a referral for targeted psychological work rather than another dose increase alone.
- For voices that persist despite adequate antipsychotic treatment, consider referral for hallucination-focused psychological therapy.
- Ask about distress and control over voices, not just their frequency — therapy shifted both.
- Expect a modest effect: about one in 16 treated patients benefits.
- Avatar therapy had the largest effect but is available in few centres.
Why it matters
Persistent voices are often met with more medication; this gives a non-drug option with evidence behind it.
Don't overread it
The avatar therapy advantage comes from a subgroup comparison, which is weaker than a head-to-head trial.
The statistics, in plain English
A standardised mean difference of −0.18 is a small effect by convention. An NNT of 16 means treating 16 people to produce one extra meaningful improvement. Low GRADE certainty means further trials could change the estimate.
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