Search

Background: The evidence on factors that may influence treatment decisional capacity (‘capacity) in psychosis has yet to be comprehensively synthesised, which limits the development of effective strategies to improve or support it.Aims: To determine the direction, magnitude and reliability of the relationship between capacity in psychosis and a range of clinical, demographic and treatment related factors, thus providing a thorough synthesis of current knowledge.Method: We did a systematic review, meta-analytical and narrative synthesis of factors that help or hinder treatment decision-making capacity in psychosis (PROSPERO registration CRD42015025568), assessing the direction, magnitude, significance and reliability of reported associations.Results: We identified 23 relevant studies (N=1823). Psychotic symptoms had small, moderate and strong associations with appreciation, understanding, and reasoning, respectively. Both verbal cognitive functioning and duration of education had small to moderate correlations with understanding and reasoning. Better capacity was also associated with better insight, better metacognitive ability, higher anxiety and lower perceived coercion. No linear relationship with depression was observed. Interventions linked to improved capacity over time were inpatient care (including antipsychotic treatment), information-simplification, shared decision-making and metacognitive training.Conclusion: This synthesis of over 25 years of research provides a comprehensive assessment of factors that may help or hinder treatment decision-making capacity in psychosis. Although much is known about the role of symptoms and other clinical variables, effective and acceptable psychological interventions to support capacity in this group are lacking.

Purpose: A systematic review and meta-analysis was conducted to identify what factors have been investigated as correlates of the 4 key domains of treatment decision making capacity (TDMC) in people who have experienced psychosis (understanding, reasoning, appreciation, communication) and to provide estimates of the magnitude of these correlations, taking into account study quality. A novel empirical study was conducted to test the hypothesis that variance in psychosis-specific cognitive biases (including the well-established ‘jumping to conclusions’ bias) would account for unique variance in TDMC domains in those with psychosis, after taking into account the known contribution of symptoms and insight. A secondary aim of the empirical study was to examine for the first time the relationship between TDMC and personal recovery in this group, and post hoc analyses of the relationship between cognitive biases, emotional distress and TDMC were also conducted. Methods: Electronic databases were systematically searched for literature on the schizophrenia and psychosis and treatment decision making capacity. Pooled estimates of correlation were estimated for factors with data from three or more studies, and both study and outcome quality were systematically assessed. A cross-sectional observational study was conducted, and individuals with psychosis completed measures of TDMC, cognitive biases, psychotic symptoms and recovery. Multiple regression was used to examine the primary and secondary hypotheses, and mediation analyses were used to conduct the post hoc analyses. Additional data from a parallel study was incorporated to increase power. Results: Twenty-four studies met inclusion criteria for the systematic review and meta-analysis. Low to moderate quality evidence suggested that the ability of people with psychosis to understand treatment-relevant information was strongly associated with overall psychotic symptom severity, verbal cognitive functioning and years of education, but not depression (moderate quality evidence). Low quality evidence suggested reasoning was strongly associated with verbal cognitive functioning and moderately associated with symptoms. Appreciation was associated with symptoms, but it and communication were generally poorly studied. Findings from the empirical study suggest that cognitive biases, and the Jumping to Conclusions bias in particular, predicts a moderate amount of the variance in the understanding and reasoning TDMC domains, but did not add predictive power to a model containing symptoms, insight, and cognition. The appreciation domain was strongly predicted by cognitive biases, insight, and cognition. TDMC was not found to be correlated with personal recovery and post hoc analyses did not find that emotional distress mediated any relationship between cognitive biases and TDMC. Conclusions: The meta-analysis confirms there is a robust association between symptoms and TDMC in psychosis, as currently conceived. The empirical study suggests cognitive biases may be related to TDMC, even after taking into account the contribution of symptoms. Larger studies, perhaps employing experimental procedures, are required to clarify the exact nature of this relationship. The lack of any relationship between TDMC and service-user defined recovery from psychosis is notable, and lends support to those calling for a conceptualisation of TDMC that takes greater account of this concept.