The COMMAND trial of cognitive therapy for harmful compliance with command hallucinations (CTCH): A qualitative study of acceptability and tolerability in the UK

Abstract

Objectives To explore service user experiences of a 9-month cognitive behavioural therapy for command hallucinations in the context of a randomised controlled trial including their views on acceptability and tolerability of the intervention. Design Qualitative study using semistructured interviews. Setting The study took place across three sites: Birmingham, Manchester and London. Interviews were carried out at the sites where therapy took place which included service bases and participants' homes. Participants Of 197 patients who consented to the trial, 98 received the Cognitive Behavior Therapy for Command Hallucinations (CTCH) intervention; 25 (15 males) of whom were randomly selected and consented to the qualitative study. The mean age of the sample was 42 years, and 68% were white British. Results Two superordinate themes were identified: participants' views about the aspects of CTCH they found most helpful; and participants' concerns with therapy. Helpful aspects of the therapy included gaining control over the voices, challenging the power and omniscience of the voices, following a structured approach, normalisation and mainstreaming of the experience of voices, and having peer support alongside the therapy. Concerns with the therapy included anxiety about completing CTCH tasks, fear of talking back to voices, the need for follow-up and ongoing support and concerns with adaptability of the therapy. Conclusions Interpretation: CTCH was generally well received and the narratives validated the overall approach. Participants did not find it an easy therapy to undertake as they were challenging a persecutor they believed had great power to harm; many were concerned, anxious and occasionally disappointed that the voices did not disappear altogether. The trusting relationship with the therapist was crucial. The need for continued support was expressed.