Multi-centre cluster trial in primary care comparing a community group exercise programme with home based exercise and with usual care for people aged 65 and over

DESIGN A multi-centre cluster controlled trial in primary care to compare two exercise interventions with usual care to increase PA in older people. The three arms of the study: FaME vs OEP vs usual care. General practices recruited from two centres, and allocated to the three arms in each centre by the London co-ordinating centre using minimisation on centre, practice size and index of multiple deprivation. Patients within practices will be stratified and recruited from two age bands. SETTING Participants recruited from general practices in two centres: London and Nottingham/Derby. TARGET POPULATION Patients 65-74 and 75+ excluding those with contraindications for exercise. HEALTH TECHNOLOGIES BEING ASSESSED FaME comprises one hour PSI delivered group exercise class in a community venue (maximum 15 participants) and two 30 minute home exercise sessions/week for 24 weeks. Participants are also advised to walk at least twice per week for up to 30 minutes, their progress and adherence monitored by the PSIs. OEP comprises at least three 30 minute home exercise sessions, supplemented by a walking plan at least two times per week for 24 weeks. Participants are supported by an assigned mentor (volunteer working with practices) who starts the home exercise plan with them and monitors and visits as necessary. Both programme objectives will be to improve functional ability, exercise confidence and encourage increasing PA to reach and maintain the recommended targets of at least 30 minutes of moderate PA on at least five days per week. MEASUREMENT OF COST AND OUTCOMES Primary outcomes PA participation measured to ascertain type, intensity, duration and frequency (Hauer 2006), and perceived health status will be measured by EuroQol EQ-5D scores transformed into QALYs. Secondary outcomes will include direct costs of delivering the programmes; health and social service utilisation; direct health benefits i.e. functional and psychological status, nature and number of falls; predictors of long term exercise: self efficacy, social support, physical self perception and exercise value. Outcomes will be measured at baseline, 24 weeks and at 6, 12, 18 and 24 months after the programmes to enable an accurate record of continuing PA participation. SAMPLE SIZE is based on numbers needed to detect differences in proportions reaching PA targets and health status measured by the EQ-5D. Under individual randomisation a small effect size (0.3) equivalent to mean differences of 5.2 points on the EQ-5D (VAS) and 0.05 on the EQ-5D index in community samples requires 176 patients per study group (Roset 1999), and 215 patients per group are required to detect the difference between study groups of 14.6% and 4.9% achieving PA targets (Elley 2003) (2-sided a=0.05, 1-b=0.90). Data from 24 practices in the British Regional Heart Study suggested an intra-class correlation coefficient (ICC) not exceeding 0.02 for PA among middle aged men (Morris 2001). ICCs collected in primary care settings have typically averaged 0.01 (Adams 2004). With a minimum practice sample after losses to follow up of 32 patients (16 in each age band); a design effect of 1.31 based on an ICC of 0.01 to account for cluster randomisation and 30% attrition, approximately 1200 patients (400 in each study arm) will be recruited from 30 practices (three groups of 200 patients/5 practices on each site in London and Nottingham/Derby). Differences in proportions reaching the PA threshold and differences in means on the EQ5D will be analysed using multi-levelling modelling which allows for the practice effect, and stratified by age group on an intention to treat basis. PROJECT TIMETABLE Months 0 to 12: ethical and R&D approval; recruitment of general practices; briefing of staff on the study; recruitment of PSIs and training of peer mentors; recruitment of patients; start baseline assessments; staggered enrolment of participants onto FaME and OEP; Months 13 to 24: complete exe