Compassion-focused therapy for psychosis


There is growing evidence that compassion-focused therapy (CFT) is effective in activating self-soothing and affiliative processes (Millard et al., 2023). As the authors of this paper summarise, it may lead to improved outcomes in people with eating disorders (Gale et al., 2014), personality disorders (Lucre & Corten, 2013), and other complex difficulties (Gilbert & Procter, 2006).

The paper reviewed here seeks to extend the evidence for CFT for psychosis. This is based on previous theoretical work suggesting that the processes targeted in CFT may be especially relevant for this group given that people with psychosis often have threat-focused cognitions (Heriot-Maitland, 2023). To date, the evidence for CFT in psychosis is provided by Mayhew and Gilbert’s (2008) promising case series with people who hear hostile voices, and Braehler et al.’s (2013) feasibility trial of a group CFT intervention. The present study also builds on encouraging empirical investigations of CFT techniques for paranoia in both clinical (Forkert et al., 2022) and non-clinical populations (Lincoln et al., 2012; Brown et al., 2020).

The authors of this study developed a manualised individual Compassion-Focused Therapy for psychosis (CFTp) intervention and investigated its acceptability, effectiveness, and the processes of change in a small sample of seven participants. The results showed that it is feasible to deliver CFTp in a manualised form for people experiencing distressing psychosis symptoms. Outcomes related to distress, stress, depression, voices, and delusions all improved during the study.

The processes targeted in Compassion-Focused Therapy may be beneficial to people with psychosis who have threat-focused experiences.

The processes targeted in Compassion-Focused Therapy may be beneficial to people with psychosis who have threat-focused experiences.


This study had a single-case, experimental, multiple-baseline design. Participants received 26 sessions of the CFTp intervention, which included standard CFT content with some psychosis-specific adaptations, such as making sense of voices. Fidelity to the therapy protocol was measured throughout.

Participants were adults invited to participate by their secondary care mental health teams in either London or Glasgow. They had a psychosis-related diagnosis and distressing psychotic experiences.

The following outcomes were measured, mostly by self-report, at five timepoints (with final follow-up 6-8 weeks after sessions ended): psychotic symptoms, depression and anxiety symptoms, psychological distress and dissociation. The authors also measured social-rank threat, safeness, compassion and heart rate variability.


Feasibility and acceptability

  • 21 referrals to the study were received. Nine (43%) of these were willing and eligible to participate. Eight people began the intervention, and there were seven completers. One participant did not complete the intervention for mental health reasons unrelated to the study.
  • All seven completing participants gave full process and outcome measures at all assessment timepoints.
  • The majority of participants were male (n=5), whilst the sample included White British (n=4), Black British (n=2) and White Irish (n=1).

Process and outcome measures

Due to the small sample size and uncontrolled nature of the study, analysis of change was conducted at the individual participant level as well as across the group. A Reliable Change Index (Jacobson & Truax, 1991) looked at greater than expected changes in outcome measures for individual participants. Changes in group mean scores across study phases were compared using a Wilcoxon signed-rank test. Significant differences between baseline and intervention phases were analysed with a Tau-U statistic.

Hypothesis 1: “Significant changes in outcome and process measures will occur during the intervention and not the baseline phase.”

This hypothesis was supported at both the individual and group levels overall. At the group level, significant improvements were observed in outcome measures of auditory hallucinations, delusions, depression, stress, and distress, but not of anxiety and dissociation. Moreover, there were significant improvements at the group level during intervention on process measures of social comparison, self-criticism, self-hatred, feeling shamed by others, self-compassion, and personal beliefs about illness. Self-reassurance and heart rate variability were processes that showed some improvement but not to a level that reached significance.

Hypothesis 2: “Significant changes in process measures will precede changes in outcome measures.”

At the individual level, this hypothesis was not supported. A similar number of participants showed reliable changes in the first half of therapy as in the full therapy. At the group level, there was some evidence that most process changes preceded changes in delusions and depression. Self-criticism, however, showed clear changes only later in therapy.

Hypothesis 3: “Significant changes in outcome and process measures will remain post-intervention.”

All follow-up measures were taken 6-8 weeks after intervention. Overall, this hypothesis was supported at single-case and group levels. In terms of process, reliable improvements on feeling shamed by others were extended from three participants during therapy, to five at follow-up. Self-criticism improved for an additional participant at follow-up, to five in total. All seven participants showed improvements in beliefs about illness at follow-up. Four (of five) participants maintained their improvements in auditory hallucinations and one in delusions. Five participants had improved depression scores. Four participants had improved stress scores, and one deteriorated. Three participants had improved anxiety scores, but two deteriorated. There was no evidence that improvements in heart rate variability remained after therapy.

Hypothesis 4: “Session-by-session measures will improve significantly in the intervention, compared to the baseline, phase.”

Overall, social safeness, but not dissociation, improved session-by-session in the intervention phase as compared to the baseline phase.

Social safeness improved session-by-session in the intervention phase.

Social safeness improved session-by-session in the intervention phase.


The authors concluded that “CFTp is a feasible and acceptable intervention for people with distressing psychotic experiences.” It was also deemed that there were “enough indicators of effectiveness” to justify proceeding to a pilot randomised controlled trial (RCT) of the intervention. With a bigger sample size, this pilot RCT will be more adequately powered to detect statistical effects. Given the expense of running even a pilot RCT, the results of this case series at the participatory level are encouraging, showing that the research procedures, as well as the intervention, were feasible to carry out with this sample of people.

“Compassion-focused therapy for psychosis is a feasible and acceptable intervention for people with distressing psychotic experiences.” argued the authors.

“CFTp is a feasible and acceptable intervention for people with distressing psychotic experiences.” argued the authors.

Strengths and limitations

 This paper makes a valuable contribution and provides encouraging results for the future of CFTp. Particularly rigorous and patient-centred aspects of the study include:

  • The theoretically-driven design, hypothesising that changes in outcomes would follow changes in associated processes.
  • The development of a 50-page treatment manual in collaboration with experts by experience.
  • The use of a treatment fidelity measure (the CFTp Adherence and Competence Measure).
  • Transparency about why not all participants completed the intervention, for example reporting that one participant found the weekly measures irritating and confusing!
  • The inclusion of the Template for Intervention Description and Replication (Hoffmann et al., 2014).
  • The authors emphasise that the intervention is targeted at people experiencing distressing voices and delusions. This is important as not all people experiencing psychosis are distressed by their experiences, and treatment development needs to target the areas that patients most want to be treated.

Inevitably the authors made compromises to maximise the strengths of the study, resulting in the following limitations:

  • Most obviously, the sample size is of course small and only one therapist (who was also the intervention developer) provided the intervention.
  • Although the collection of both process and outcome measures is a strength, process improvements did not precede outcome improvements at the single-case level, leaving uncertainty about the mediators and mechanisms of change. The 26-session intervention was comprehensive, but given its length, it is difficult to distill down to the most helpful active ingredients to test in an RCT. This is important when thinking forward to implementation within mainstream mental health services.
  • In the same vein, it might have been useful to see a qualitative phase to the study, hearing from participants about how acceptable they found the therapy and which elements they found subjectively most helpful or unhelpful.
  • Mere participation in a safe, supported intervention, aside from therapy factors, may have accounted for some of the changes observed. This may be especially the case for social safeness, since leaving the house to attend sessions with a responsive listener and even just participating in research assessments may have increased participants’ sense of safety around others. A future RCT design will be better equipped to account for this confound.
  • It could have been interesting to discuss in the paper the rationale for certain key decisions, such as choosing social safeness and dissociation as the session-by-session measures over other potential variables, and excluding participants if they had had more than 12 sessions of psychological therapy within the past three years.
The development of a therapeutic manual in consultation with experts by experience is a particular strength of this feasibility study.

The development of a therapeutic manual in consultation with experts by experience is a particular strength of this feasibility study.

Implications for practice

This study paves the way for a randomised-controlled trial of individual Compassion-Focused Therapy for psychosis. There are considerable challenges when conducting research with this group, so this is an incredibly exciting development both in terms of psychosis treatment and the clinical evaluation of CFT. The authors lay out a clear theoretical framework for why CFT may be particularly well-poised to address the social-threat processes inherent in psychosis. Although this theory was not fully supported by the results of this small study, the detailed and theoretically-driven design of the study holds great merit. Which processes the treatment aims to target and how these are measured and linked to outcomes may need refining in the pilot RCT, but the simultaneous assessment of potential processes of change alongside measuring outcomes will give important information about active ingredients that will ultimately be useful in a full-scale trial. The tighter these links can be made and evaluated, the more likely it is that we may see CFT (if shown to be effective) offered as a treatment for people with distressing psychotic experiences within future mainstream mental health services.

Compassion-Focused Therapy for psychosis needs to be fully evaluated with larger sample sizes before being widely implemented in the NHS.

Compassion-Focused Therapy for psychosis needs to be fully evaluated with larger sample sizes before we consider implementation in the NHS.

Statement of interests

Ellen Iredale declares no conflicts of interest. Poppy Brown, while having no involvement with the present study, has conducted research and published work in the area of compassion and paranoia, which is cited in the primary paper. She declares no other conflicts.


Primary paper

Heriot‐Maitland, C., Gumley, A., Wykes, T., Longden, E., Irons, C., Gilbert, P., & Peters, E. (2023). A case series study of compassion‐focused therapy for distressing experiences in psychosisBritish Journal of Clinical Psychology62(4), 762-781.

Other references

Braehler, C., Gumley, A., Harper, J., Wallace, S., Norrie, J., & Gilbert, P. (2013). Exploring change processes in compassion focused therapy in psychosis: Results of a feasibility randomized controlled trial. The British Journal of Clinical Psychology,52(2), 199–214.

Brown, P., Waite, F., Rovira, A., Nickless, A., & Freeman, D. (2020). Virtual reality clinical-experimental tests of compassion treatment techniques to reduce paranoia. Scientific Reports, 10(1), 8547.

Egan, S. Self-compassion for anxiety and depression in young people #ActiveIngredientsMH. The Mental Elf, 27 Nov 2020.

Forkert, A., Brown, P., Freeman, D., & Waite, F. (2022). A compassionate imagery intervention for patients with persecutory delusions. Behavioural and Cognitive Psychotherapy, 50(1), 15–27.

Gale, C., Gilbert, P., Read, N., & Goss, K. (2014). An evaluation of the impact of introducing compassion focused therapy to a standard treatment programme for people with eating disorders. Clinical Psychology & Psychotherapy, 21(1), 1–12.

Gilbert, P., & Procter, S. (2006). Compassionate mind training for people with high shame and self-criticism: Overview and pilot study of a group therapy approach. Clinical Psychology & Psychotherapy, 13(6), 353–379.

Heriot-Maitland, C. (2023). Position paper – CFT for psychosis. Psychology and Psychotherapy: Theory, Research and Practice, 1–15.

Hoffmann, T. C., Glasziou, P. P., Boutron, I., Milne, R., Perera, R., Moher, D., … Michie, S. (2014). Better reporting of interventions: Template for intervention description and replication (TIDieR) checklist and guide. BMJ, 348, g1687.

Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59(1), 12–19.

Laws, K. Third wave CBT for psychosis: how reliable is current evidence? The Mental Elf, 5 Jul 2018.

Lincoln, T. M., Hohenhaus, F., & Hartmann, M. (2012). Can paranoid thoughts be reduced by targeting negative emotions and self-esteem? An experimental investigation of a brief compassion-focused intervention. Cognitive Therapy and Research, 37(2), 390–402.

Lucre, K. M., & Corten, N. (2013). An exploration of group compassion-focused therapy for personality disorder. Psychology and Psychotherapy, 86(4), 387–400.

Mayhew, S. L., & Gilbert, P. (2008). Compassionate mind training with people who hear malevolent voices: A case series report. Clinical Psychology & Psychotherapy, 15(2), 113–138.

Millard, L. A., Wan, M. W., Smith, D. M., & Wittkowski, A. (2023). The effectiveness of compassion focused therapy with clinical populations: A systematic review and meta-analysis. Journal of Affective Disorders, 326, 168–192.

Photo credits

Source link

We will be happy to hear your thoughts

Leave a reply

Nyc Health Store | Amazon Affiliate Store
Enable registration in settings - general
Compare items
  • Total (0)