Evaluating the effects of community treatment orders (CTOs) in England using the Mental Health Services Dataset (MHSDS): Protocol for a national, population-based study
Weich S., Duncan C., Bhui K., Canaway A., Crepaz-Keay D., Keown P., Madan J., McBride O., Moon G., Parsons H., Singh S., Twigg L.
Introduction Supervised community treatment (SCT) for people with serious mental disorders has become accepted practice in many countries around the world. In England, SCT was adopted in 2008 in the form of community treatment orders (CTOs). CTOs have been used more than expected, with significant variations between people and places. There is conflicting evidence about the effectiveness of SCT; studies based on randomised controlled trials (RCTs) have suggested few positive impacts, while those employing observational designs have been more favourable. Robust population-based studies are needed, because of the ethical challenges of undertaking further RCTs and because variation across previous studies may reflect the effects of sociospatial context on SCT outcomes. We aim to examine spatial and temporal variation in the use, effectiveness and cost of CTOs in England through the analysis of routine administrative data. Methods and analysis Four years of data from the Mental Health Services Dataset (MHSDS) will be analysed using multilevel models. Models based on all patients eligible for CTOs will be used to explore variation in their use. A subset of CTO-eligible patients comprising a treatment group (CTO patients) and a matched control group (non-CTO patients) will be used to examine variation in the association between CTO use and study outcomes. Primary outcome will be total time in hospital. Secondary outcomes will include time to first readmission and mortality. Outputs from these models will be used to populate predictive models of healthcare resource use. Ethics and dissemination Ethical approval has been granted by the National Health Service Data Access and Advisory Group and Warwick University. To ensure patient confidentiality and to meet data governance requirements, analyses will be carried out in a secure microdata laboratory using de-identified data. Study findings will be disseminated through academic channels and shared with mental health policy-makers and other stakeholders.