Randomised controlled trial of a new palliative care service: Compliance, recruitment and completeness of follow-up
1 King's College London, School of Medicine, Department of Palliative Care, Policy and Rehabilitation, Denmark Hill, London SE5 9RJ, UK
2 King's College Hospital NHS Trust, Department of Neurology, Denmark Hill, London SE5 9RS, UK
BMC Palliative Care 2008, 7:7 doi:10.1186/1472-684X-7-7Published: 28 May 2008
Palliative care has been proposed for progressive non-cancer conditions but there have been few evaluations of service developments. We analysed recruitment, compliance and follow-up data of a fast track (or wait list control) randomised controlled trial of a new palliative care service – a design not previously used to assess palliative care.
An innovative palliative care service (comprising a consultant in palliative medicine, a clinical nurse specialist, an administrator and a psychosocial worker) was delivered to people severely affected by multiple sclerosis (MS), and their carers, in southeast London. Our design followed the MRC Framework for the Evaluation of Complex Interventions. In phase II we conducted randomised controlled trial, of immediate referral to the service (fast-track) versus a 12-week wait (standard best practice). Main outcome measures were: compliance (the extent the trial protocol was adhered to), recruitment (target 50 patients), attrition and missing data rates; trial outcomes were Palliative Care Outcome Scale and MS Impact Scale.
69 patients were referred, 52 entered the trial (26 randomised to each arm), 5 refused consent and 12 were excluded from the trial for other reasons, usually illness or urgent needs, achieving our target numbers. 25/26 fast track and 21/26 standard best practice patients completed the trial, resulting in 217/225 (96%) of possible interviews completed, 87% of which took place in the patient's home. Main reasons for failure to interview and/or attrition were death or illness. There were three deaths in the standard best practice group and one in the fast-track group during the trial. At baseline there were no differences between groups. Missing data for individual questionnaire items were small (median 0, mean 1–5 items out of 56+ items per interview), not associated with any patient or carer characteristics or with individual questionnaires, but were associated with interviewer.
This is the first time a fast track (or wait list) randomised trial has been reported in palliative care. We found it achieved good recruitment and is a feasible method to evaluate palliative care services when patients are expected to live longer than 3–6 months. Home interviews are needed for a trial of this kind; interviewers need careful recruitment, training and supervision; and there should be careful separation from the clinical service of the control patients to prevent accidental contamination.
Clinical Trials.Gov NCT00364963