Quality of life, autonomy, satisfaction, and costs associated with mental health supported accommodation services in England: a national survey.

BACKGROUND Little research has been done into the effectiveness of mental health supported accommodation services. We did a national survey to investigate provision and costs of services and assess service user quality of life and outcomes across England. METHODS We randomly sampled three types of services from 14 nationally representative regions-residential care, supported housing, and floating outreach-and recruited up to ten service users per service. Service quality and costs and service users' quality of life, autonomy, and satisfaction with care were assessed in a standardised manner with validated tools and compared by multilevel modelling. FINDINGS 619 service users were recruited from 22 residential care, 35 supported housing, and 30 floating outreach services. Those in residential care and supported housing had more severe mental health problems than those in floating outreach. 348 (57%) were assessed as being at risk of severe self-neglect and 229 (37%) as being vulnerable to exploitation in the previous 2 years. Residential care was most expensive but provided for people with the greatest needs. The mean annual budget was £466 687 for residential care (range £276 000-777 920), compared with £365 452 for supported housing (£174 877-818 000), and £172 114 for floating outreach (£17 126-491 692). Quality of care was best in supported housing. People in supported housing and floating outreach were more socially included but experienced more crime than those in residential care. After adjustment for service quality and service user sociodemographic and clinical factors, quality of life was similar for service users in residential care and supported housing (mean difference -0·138, 95% CI -0·402 to 0·126, p=0·306) and lower for those in floating outreach than in residential care (-0·424, -0·734 to -0·114, p=0·007). However, autonomy was greater for those in supported housing than for those in residential care (0·145, 0·010 to 0·279, p=0.035). Satisfaction with care was similar across services. INTERPRETATION Supported housing might be cost-effective, but the benefits need to be weighed against the risks associated with increased autonomy. FUNDING National Institute for Health Research.