Background Antipsychotic medication is beneficial for people with psychosis or schizophrenia in the short term, but the balance of risks and benefits in the long term is less clear. Many patients remain functionally impaired, experience significant and distressing side effects and physical health problems. Evidence in people with first episode psychosis suggests that social functioning may be improved for some patients following a gradual reduction or discontinuation of antipsychotics, but there is no evidence in people with recurrent conditions. Objectives To assess patients’ attitudes to long-term use of antipsychotic medication (work package 1a). To develop a gradual strategy of antipsychotic reduction and discontinuation and to design a trial to assess its benefits and harms (work packages 1a and 1b). To evaluate the antipsychotic reduction strategy in a randomised pilot trial (work package 2). To conduct a full randomised trial (work package 3a) of the antipsychotic reduction strategy in patients with multie pisode psychosis. To explore the experiences of people enrolled in the antipsychotic reduction strategy (work package 3b). Methods Design Objective 1: A mixed-methods survey of attitudes to long-term treatment and (work package 1a). Objective 2: Survey of attitudes to participation in a randomised trial, expert consultation, lived experience consultation, literature review, including a systematic review of definitions of relapse in previous trials of antipsychotics, focus groups (work packages 1a and 1b). Objectives 3 and 4: Multicentre, pragmatic, open, parallel group randomised controlled trial (with blinded assessors) (work packages 2 and 3a). Objective 5: qualitative study using semistructured interviews (work package 3b). Participants Patients with schizophrenia or recurrent, non-affective psychotic episodes, taking antipsychotics. Setting English community mental health services. Interventions A gradual strategy of antipsychotic reduction, with discontinuation where possible (evaluated in work packages 2 and 3, developed during work package 1). Main outcome measures In work package 1a: Patients’ views of antipsychotic medication and participation in a randomised trial of antipsychotic reduction. In work package 2: Recruitment rates and adherence to trial procedures. In work package 3a: Primary outcome was the Social Functioning Scale. The principal secondary outcome was severe relapse (admission to mental health inpatient care). Other outcomes included any relapse, symptoms, and a full economic analysis was performed. Results Work package 1a – 269 participants were interviewed. Only 33% were content with taking antipsychotic medication for long term, and 31% and 45% wished to try and stop or reduce it, respectively, with clinical support. Seventy-nine per cent indicated that they would or might be interested in taking part in a future trial. Work package 1b – Trial recruitment strategy and intervention protocol, procedures and adherence protocols were developed, along with definitions and procedures for determining relapse. The systematic review failed to identify a previous definition of relapse that was reliable, clinically relevant and feasibly to apply. Work package 2 – Recruitment was 65% of the projected target and the trial was approved to continue to a full trial. Work package 3a – 253 participants were randomised, 126 to supported reduction, 127 to maintenance antipsychotic treatment. At 24 months, there was no statistically significant difference between groups in change on the Social Functioning Scale ( b : 0.19, 95% confidence interval −1.94 to 2.33, p = 0.859). The rate of severe relapse was significantly higher in the intervention group (25%) compared to the maintenance group (13%) (odds ratio, b : 2.39, 95% confidence interval 1.25 to 4.56). The risk of any relapse was also higher, but there was no difference in other outcomes. There were 93 serious adverse events in 49 individuals in the reduction arm (mostly admissions to hospital for a mental health relapse) and there were 64 in 29 individuals in the maintenance arm. In the economic analysis, at 24 months, there was no significant difference between the two arms in costs (£5619; 95% confidence interval −£386 to £11,625) or quality-adjusted life-years (−0.035, 95% confidence interval −0.125 to 0.054). There were significantly fewer years of full capability in the supported reduction arm (−0.103, 95% confidence interval −0.191 to −0.015). Work package 3b – 26 patients were interviewed. Analysis indicated that patients experienced improvements in adverse effects, social functioning and sense of self. However, some experienced increased symptoms, challenging emotional intensity and relapse. For some patients, supported medication reduction provided an opportunity for learning and empowerment. Work package 3c – 15 clinicians provided feedback on their experiences. They felt that antipsychotic reduction had been beneficial for some but had led to negative outcomes for other patients. Some reported that the trial enabled more collaborative relationships with patients. Limitations Recruitment was challenging, and some participants did not adhere to their randomised treatment programme. The number recruited was a small proportion of the number screened, which means the sample may differ from the general population of people considered likely to be suitable to take part in a trial of this sort. The COVID pandemic affected the social functioning measure. Conclusions Quantitative analysis showed that antipsychotic reduction and discontinuation over a period of months does not improve social functioning at 24 months, and increases the risk of severe relapse. There is a low probability that supported reduction is cost-effective compared to maintenance. In qualitative findings, patients and clinicians noted positive and negative effects of antipsychotic reduction. Future work Future work should include long-term follow-up of the trial cohort and investigation of more gradual reduction strategies. Trial registration The trial is registered as ISRCTN90298520 on 7 February 2017 and at ClinicalTrials.gov on 18 June 2018. Funding
Moncrieff et al. (Tue,) studied this question.