Understanding decision-making in psychosis: a case series of psychological assessment and formulation of impaired treatment decision-making; and, A systematic review and meta-analysis of the attribution–self- representation model of persecutory delusions

Abstract

PURPOSE: A systematic review and meta-analysis was conducted to test key predictions of the widely-studied ‘paranoia as defence’ model (more formally known as the ‘attribution–self-representation cycle’) proposed by Bentall, Corcoran, Howard, Blackwood, and Kinderman (2001), as applied to people with psychosis with persecutory delusions. A novel case series was also conducted to examine the feasibility and acceptability of collaborative psychological assessment and formulation of impaired treatment decision-making capacity (TDMC) among patients with psychosis, and produce preliminary data on safety and efficacy.

METHODS: With regard to the systematic review and meta-analysis, people with psychosis with persecutory delusions were compared to healthy controls, people with depression and people with psychosis without persecutory delusions (and, if specified, grandiose delusions) on a number of outcomes: externalising attributional bias, explicit self-esteem, implicit self-esteem and discrepancy between implicit and explicit self-esteem. Correlations between paranoia severity and each of these outcomes and self-esteem instability were also examined. In regards to the case series, a formulation of impaired TDMC for 5 patient participants was developed and shared with 13 clinician participants. Acceptability, utility, working alliance and safety were assessed through pre and post self-report and interview measures.

RESULTS: Sixty-three studies were included in the meta-analysis and systematic review, of which 33, 36, 10, 10 and 4 were used to test hypotheses on externalising attributional bias, explicit self-esteem, implicit self-esteem, implicit-explicit self-esteem discrepancy and selfesteem instability, respectively. Key model-consistent findings included the following: people with psychosis with persecutory delusions had a greater externalising attributional bias compared to all the other groups and a greater implicit-explicit self-esteem discrepancy than people with depression, and paranoia severity was positively correlated with externalising attributional bias and self-esteem instability. Key model-inconsistent findings included the following: people with psychosis with persecutory delusions had lower explicit self-esteem than healthy controls, and paranoia severity was negatively correlated with explicit self-esteem. There were also some model-inconclusive findings. Regarding the case series, 3 of the patient participants collaborated in the development of their formulation. They found the intervention safe and acceptable, following which they provided a much richer understanding of the factors that may impair their TDMC (Cohen’s d = 2.16). Two patient participants only partially adhered to the intervention protocol, but a psychological formulation was still feasible to produce and no adverse effects were reported. Clinician participants provided a much richer understanding of the factors that may impair the patient participants’ TDMC (Cohen’s d = 1.36; 95% CI = 0.63 to 2.07) after the presentation of the case formulations. Increases in knowledge, confidence and positive attitudes regarding supporting the TDMC of patients were observed. They strongly believed that the formulations cohered with their knowledge of the patient participants and were comprehensive and accurate.

CONCLUSIONS: The findings of the systematic review and meta-analysis support a ‘weak’ version of the paranoia as defence model, which suggests persecutory delusions are only partially effective at protecting low implicit self-esteem from reaching awareness. The findings of the case series suggest that patients with psychosis, and their clinicians, can be engaged in a collaborative psychological assessment and formulation of factors that may impair their TDMC. Initial data from the case series also suggests this process is acceptable, safe and helpful.