Does better quality of primary care influence admissions and health outcomes for people with serious mental illness (SMI)? A linked patient-level analysis of the full patient care pathway

Date of ISAC Approval: 
Lay Summary: 
Serious mental illness (SMI) covers a range of conditions including schizophrenia, psychoses and bipolar affective disorder. Patients with these conditions are at higher risk of developing a range of physical conditions and tend to have poorer health outcomes than patients without SMI. Our key research question addresses whether better quality of primary care - as delivered by general practices - reduces the risk of attending accident and emergency (A&E) at hospital, and having poorer outcomes for patients with SMI. This research project will enable us to investigate the entire care pathway for SMI patients for the first time in an English study. We will examine the care that patients receive in primary care in addition to their care in hospital and the community. We will also investigate whether the introduction of national guidelines and quality improvement schemes was associated with improved patient outcomes. Finally, we will make recommendations to NHS organisations on improvements that can be made within primary care in order to deliver better outcomes for patients with SMI.
Technical Summary: 
In England, general practitioners (GPs) play a central role in the care of most people with serious mental illness (SMI). The Quality and Outcomes Framework (QOF) offers financial rewards to GP practices for good quality care, including payments for treating people with SMI. Our key research question is whether better management of SMI in general practice improves health outcomes and health care utilisation for people with SMI. We will measure the quality of care using two methods: 1) indicators from the QOF; and 2) non-QOF quality measures also derived from primary care patient records. Our two outcome measures will include: 1) Number of A&E attendances (found in Hospital Episodes Statistics (HES) A&E data), and 2) Health of the Nation Outcome Scale (HoNOS) scores (found in Mental Health and Learning Disabilities Data Set (MHLDDS) data). We will use quantitative approaches (e.g. regression models, survival analysis, time-to-event analysis) to model the impact of the two types of quality indicators on the two outcome measures, conditional on a number of covariates which may affect A&E attendances and outcome scores. We already hold a dataset of SMI patients (CPRD data protocol 14_168RMn) and we seek to link this to two non-standard linkages (HES A&E and MHLDDS) to answer the two research questions.
Health Outcomes to be Measured: 
Number of Accident and Emergency attendances Health of the Nation Outcome Scale scores

Rowena Jacobs - Chief Investigator - University Of York
Anne Mason - Collaborator - University Of York
Ceri Owen - Collaborator - Service User of MHLDDS
Christoph Kronenberg - Collaborator - University Of York
Giuseppe Moscelli - Collaborator - University Of York
Professor Hugh Gravelle - Collaborator - University Of York
Katja Grasic - Collaborator - University Of York
Lauren Aylott - Collaborator - Service User of MHLDDS
Maria Goddard - Collaborator - University Of York
Mark Wilson - Collaborator - University Of York
Nigel Rice - Collaborator - University Of York
Nils Gutacker - Collaborator - University Of York
Panagiotis Kasteridis - Collaborator - University Of York
Ruth Helstrip - Collaborator - University Of York
Simon Gilbody - Collaborator - University Of York
Dr Tim Doran - Collaborator - University Of York
Professor Tony Kendrick - Collaborator - University Of Southampton

HES A&E;Mental Health;Non-standard;Patient IMD