In order to improve the use of limited health care resources, there is interest in assessing the value for money of treatments for psychotic disorders, a group of serious mental illnesses. However, most studies have assessed the value for money of medications, psychological therapies or community services and relatively little is known about the value of specialist inpatient care. Data that is routinely collected in electronic health records can be the basis for generating relatively inexpensive and timely evidence to support policy relevant questions.Thus, the aim of this thesis was to assess the value for money of specialist inpatient care for people with psychosis using data from electronic health records. My objectives were (1) To identify approaches to handle unmeasured confounding and measurement error; (2) To conduct an economic evaluation of admission to child and adolescent inpatient care compared to admission to adult wards for young people with psychosis (Analysis 1) and (3) to conduct an economic evaluation of referral to inpatient rehabilitation compared to usual care for adults with persistent forms of psychosis (Analysis 2). All analyses are based on data derived from the South London and Maudsley Biomedical Research Centre (BRC) clinical records interactive search (CRIS) database. In addition to three approaches to handling confounding that are well-known in health economics, I identify the front-door adjustment as an approach relevant to Analysis 2. I distinguish between four types of measurement error assumptions with respect to non-outcome variables and discuss five potential different strategies to support or enable these measurement error assumptions.The results of Analysis 1 suggest that a regression discontinuity design is not suitable to compare the impact of admission to child and adolescent and adult ward. The results of Analysis 2 suggest the costs of inpatient rehabilitation are not offset by substantial savings in other service use and there was little evidence to suggest that patients benefit clinically from referral to inpatient rehabilitation.
|Date of Award||1 Feb 2021|
|Supervisor||Paul McCrone (Supervisor) & Richard Hayes (Supervisor)|