Predictors of disengagement from cognitive behavioural therapy for psychosis in a National Health Service setting: A retrospective evaluation

Abstract

To evaluate whether demographic and clinical variables are related to disengagement rates in cognitive behavioural therapy (CBT) for psychosis in a clinical setting. The medical records and symptom severity data (from Health of the Nation Outcome Scales) were analysed retrospectively for 103 referrals for CBT for psychosis in a National Health Service secondary care and Early Intervention in Psychosis team. Overall, 42.7% (n=44) disengaged from CBT. There was no impact of gender or ethnicity, and no impact of clinical variables such as risk history and comorbid diagnosis. However, risk of disengagement was significantly higher for those who were younger, F=6.89, partial η2 =.064, p=<.05; those with greater total HoNOS scores, F=4.22, partial η2 =.04, p<.05; more severe symptoms on the HoNOS items of overactive, aggressive, disruptive, or agitated behaviour, χ2 =6.13, p<.01; problem drinking or drug taking, χ2 =7.65, p<.05; depressed mood, χ2 =7.0, p<.01; and problems with occupation and activities: χ2 =3.68, p<.05. There was a non-significant trend for shorter waiting times to be associated with greater levels of disengagement. These results indicate that it may not be psychosis per se that disrupts engagement in CBT, but linked behavioural and emotional factors. A more assertive approach to these factors - overactive, aggressive, disruptive, or agitated behaviour, problem drinking or drug taking, depressed mood, and problems with occupation and activities, particularly in younger people - may be valuable prior to or early on in therapy as a means of increasing engagement in CBT for psychosis. Risk of disengagement from CBT for psychosis increases with overactive, aggressive, disruptive, or agitated behaviour (54.9% vs. 30.8%), problem drinking and drug taking (61.1% vs. 32.8%), depressed mood (56% vs. 30.2%), and problems with occupation and activities (53.3% vs. 34.5%), with a trend for younger age. An assertive and motivational approach to engagement and a focus on addressing low mood and problematic behaviours, prior to or early in therapy, may be warranted, particularly for younger people. This evaluation is limited by small sample size and being retrospective. These results speak to the question of whether psychosis itself renders people inappropriate for CBT for psychosis, or whether problems arise due to behavioural and emotional factors that might be addressed to increase access to CBT for psychosis.