Risks, benefits and NHS costs associated with antipsychotic medications prescribed to individuals diagnosed with serious mental illness

The proposed research is relevant to people experiencing serious mental illnesses (SMI) including schizophrenia, schizoaffective disorder and bipolar affective disorder. SMIs are mental disorders which have a substantial impact on people's lives. Those with SMIs are at increased risk of suicide, accidents, violence, heart disease and stroke. The impact of SMIs on life expectancy is substantial and generally higher than that of smoking, diabetes and obesity. Our research has estimated that individuals with SMI die between 10-15 years earlier than the general population. However, reasons for this remain unclear.

Despite improving healthcare, people with SMIs continue to have frequent relapses and reduced ability to function. Antipsychotic medications, a mainstay of treatment for SMIs, are used to manage psychotic symptoms, including hallucinations, delusions and disturbed thoughts. The effectiveness of antipsychotic medications has been assessed in numerous experimental clinical trials; however, a number of adverse effects have been reported, including, diabetes, lowered white blood cell count and heart disease. Also experimental clinical trials are performed under ideal conditions and over relatively short time periods. In practice, antipsychotics are used in a variety of ways that differ from how they are tested in trials. For example, often more than one antipsychotic is prescribed or clinicians may switch medications or increase the dose in an effort to maximise the benefit. Currently, there is very limited information on the risks and benefits of these decisions in real-life clinical settings.

Detailed electronic patient records provide an ideal source of data to investigate the risks and benefits associated with antipsychotic medication use in real-life. The South London and Maudsley NHS Foundation Trust (SLAM) provides all mental healthcare to 4 London boroughs and is the largest such provider in Europe. Electronic clinical records have been used across all SLAM services since 2006, and the Clinical Record Interactive Search (CRIS) system allows researchers to search and retrieve anonymised information from the records efficiently. There are currently over 180,000 cases in the CRIS system providing in-depth information which I have used in previous investigations.

Aims: My aim is to investigate antipsychotic profiles (antipsychotic use over time) and examine the impact of these profiles on the mental health, physical health and risk of mortality in people with SMIs. This will be achieved using CRIS linked to routine mortality tracing and death certificates, routine hospital care information and a local General Practice dataset making it possible to assess primary, emergency and acute care, causes of mortality, physical and mental health outcomes and associated costs for a large group of SLAM patients with SMIs.

Impact: This research will improve our understanding of the physical and mental health consequences of antipsychotic medications and the ways in which they are used in secondary care. This information would prove invaluable for clinical practice allowing clinicians to make better informed decisions regarding the risks and benefits of commonly used medications. Understanding associated risks, benefits, service use and costs can inform healthcare policy, allow health service providers to make more effective use of available resources and assist the NHS in providing the highest level of patient care.

Researchers: As the fellowship applicant, I will coordinate the study activities in this multidisciplinary team under the supervision of Dr Robert Stewart, and mentored by Professor Matthew Hotopf at Kings Collage London, Institute of Psychiatry. The collaborators will contribute expertise and guidance in their respective fields.

Part 1: I will model cross-sectional and longitudinal antipsychotic medication profiles to identify prescribing patterns exploring polypharmacy, medication switching, antipsychotic load and antipsychotic action ( based on known target receptors). I will explore associations between antipsychotic profiles and cause-specific mortality in people with SMI. Text mining algorithms and structured fields in CRIS will provide data on antipsychotics and potential confounders making it possible to control for confounders including disease severity (using HONOS and level of mental health service contact) and other medications. Routine mortality tracing already provides data on deaths, and cause of death will be retrieved by the Medical Research Information Service.
Part 2: I will investigate associations between cross-sectional/longitudinal antipsychotic medication profiles (identified in part 1) and physical illness in people with SMI. CRIS will provide data on use of mental health services, psychotropic medications and mental health outcomes. A linkage between CRIS and Hospital Episode Statistics (HES), will provide data on emergency and acute hospital care, diagnosis and length of stay for physical illness in those with SMIs and control for potential confounders including disease profile/management and pre-existing illness. In addition, in a subsample of SMI cases I will examine primary care outcomes through a link between CRIS and Lambeth DataNet, a primary care database for one of four SLAM catchment boroughs.
Part 3: Using the CRIS to HES and Lambeth DataNet linkage, I will calculate costs associated with different antipsychotic medication profiles, allocating costs to admissions and medications prescribed while in contact with SLAM services as well as costs associated with acute and emergency hospital care. Benefit will be estimated with longitudinal observational data in nested studies comparing good and poor disorder control under different pharmacotherapy profiles.

This fellowship proposal is relevant to people with serious mental illnesses (SMIs) including schizophrenia, schizoaffective disorder and bipolar affective disorder and the clinicians who treat them. Despite medical advancements SMIs continue to be associated with a chronic relapsing course and notable health disparities including a substantially reduced life expectancy. In real-life clinical settings antipsychotic medications designed to treat SMIs are used in a variety of ways that go beyond the scope of most randomised controlled trials. Additional information is required to understand fully the potential consequences of antipsychotic pharmacotherapy in real-life settings. This information is essential to improving outcomes for people with SMIs. In this fellowship proposal I will draw on electronic secondary mental health care records linked to mortality tracing and primary, emergency and acute care databases to examine the impact of antipsychotic prescribing practices and medication profiles on mental health, physical health and risk of mortality in people with SMIs. This information would allow clinicians to make more informed decisions based on a clear understanding of risks and benefits of antipsychotic prescribing practices for improved patient outcomes. As the planned research progresses clinicians will be able to access this information from a variety of sources including presentations at scientific conferences, open access peer-reviewed journals and on the BRC-MH website where results summaries will be featured. Much attention has been paid to service user involvement in the work undertaken at the SLAM BRC-MH including the setting up of a Stakeholder Participation Theme. In addition there is an ongoing programme set up to publicize our research to staff and service users in SLAM and neighbouring Trusts. These efforts are carried out in conjunction with the IoP media support unit and the BRC-MH Stakeholder Participation Theme.

The proposed research aims increase our understanding the risks, benefits and costs of current prescribing practices which can inform healthcare policy and treatment guidelines. Timescales for this type of impact are likely to be relatively rapid, within say 5-10 years of the output, although may depend on the availability of confirmatory findings from other samples. Because of the observational nature of the data (as distinct from randomized controlled trial data) findings and evidence will be contributory rather than definitive, but the objective will be a more evidence-based approach to medication use in the disorders of interest. As well as assessment of mental and physical health consequences of prescribing practices this proposal includes a health economics component where costs will be attributed to medications and primary, emergency, acute and mental health care service use. This information will allow health service providers to make more effective use of available resources and assist the NHS in providing the highest level of patient care, again with a similar timescale. Result summaries will be fed back to relevant organizations such as NICE with the aim of directly impacting NHS policy and current patient care.