The relationship between patient suitability, therapeutic alliance, homework compliance and outcome in cognitive therapy for psychosis

Objective. This paper examines the relationships between patient suitability, therapeutic alliance, homework compliance and outcome in cognitive-behaviour therapy (CBT) for psychosis. Method. 29 patient and therapist dyads completed patient and therapist versions of the California Psychotherapeutic Alliance Scale (CALPAS) at session 3 and 21 at session 9. The level of patient homework compliance was rated by therapist and patient, using visual analogue scales, at therapy sessions 3 (29 dyads) and 9 (20 dyads). The Positive and Negative Symptoms of Schizophrenia (PANSS) scale was used to measure psychotic symptoms pre- and post-treatment. Patient suitability for the treatment was measured at session 3, using the Suitability for Short-Term Cognitive Therapy Scale. Results. Only two of the specific hypotheses were supported. There was no significant agreement between patients and therapists on the quality of the therapeutic alliance at session 3, but by session 9 two (patient commitment and working strategy consensus) of the four sub-scales were significantly associated, although there was little evidence that the quality improved over time. Lower patient suitability and insight at session three predicted a lower therapeutic alliance. Therapeutic alliance predicted level of homework compliance. The only significant predictor of outcome was the PANSS positive symptoms sub-scale pre-treatment score. Conclusions. These results suggest that patients who experience psychotic symptoms can be successfully engaged early in therapy as can those with affective disorders but that therapeutic alliance may not have a direct effect on outcome. Copyright (c) 2006 John Wiley & Sons, Ltd.