• Graham Thornicroft, Simone Farrelly, George Szmukler, Max Birchwood, Waquas Waheed, Clare Flach, Barbara Barrett, Sarah Byford, Claire Henderson, Kim Sutherby, Helen Lester, Diana Rose, Graham Dunn, Morven Leese, and Max Marshall.
• Health Service and Population Research Department, Institute of Psychiatry, King's College London, UK. graham.thornicroft@kcl.ac.uk
• Lancet. 2013 May 11; 381 (9878): 163416411634-41.

BackgroundThe CRIMSON (CRisis plan IMpact: Subjective and Objective coercion and eNgagement) study is an individual level, randomised controlled trial that compared the effectiveness of Joint Crisis Plans (JCPs) with treatment as usual for people with severe mental illness. The JCP is a negotiated statement by a patient of treatment preferences for any future psychiatric emergency, when he or she might be unable to express clear views. We assessed whether the additional use of JCPs improved patient outcomes compared with treatment as usual.MethodsPatients were eligible if they had at least one psychiatric admission in the previous 2 years and were on the Enhanced Care Programme Approach register. The study was done with 64 generic and specialist community mental health teams in four English mental health care provider organisations (trusts). Hypotheses tested were that, compared with the control group, the intervention group would experience: fewer compulsory admissions (primary outcome); fewer psychiatric admissions; shorter psychiatric stays; lower perceived coercion; improved therapeutic relationships; and improved engagement. We stratified participants by centre. The research team but not participants nor clinical staff were masked to allocation. This study is registered with ClinicalTrials.gov, number ISRCTN11501328.Findings569 participants were randomly assigned (285 to the intervention group and 284 to the control group). No significant treatment effect was seen for the primary outcome (56 [20%] sectioned in the control group and 49 [18%] in the JCP group; odds ratio 0·90 [95% CI 0·58-1·39, p=0·63]) or any secondary outcomes, with the exception of an improved secondary outcome of therapeutic relationships (17·3 [7·6] vs 16·0 [7·1]; adjusted difference -1·28 [95% CI -2·56 to -0·01, p=0·049]). Qualitative data supported this finding.InterpretationOur findings are inconsistent with two earlier JCP studies, and show that the JCP is not significantly more effective than treatment as usual. There is evidence to suggest the JCPs were not fully implemented in all study sites, and were combined with routine clinical review meetings which did not actively incorporate patients' preferences. The study therefore raises important questions about implementing new interventions in routine clinical practice.FundingMedical Research Council UK and the National Institute for Health Research.Copyright © 2013 Elsevier Ltd. All rights reserved.

20 keywords...

hide…