The acceptability, effectiveness and cost-effectiveness of Enhanced Patient Observation in reducing the risk of self-harm and suicide on psychiatric wards
Every day on psychiatric wards across England, staff sit with patients for hours at a time to prevent them from harming themselves—yet no one knows whether this practice, called Enhanced Patient Observation (EPO), actually works. This matters because EPO consumes significant NHS resources and is recommended by NICE guidelines, but there is no good evidence it reduces suicide or self-harm. Patients often find it intrusive, coercive, and unsettling. The research will fill this gap by analysing routine electronic health records from two mental health trusts, using a target trial emulation approach that has equivalent validity to a randomised controlled trial but costs far less. If the research succeeds, it will produce the first robust evidence on whether EPO is effective and cost-effective. The team will also co-create staff training, patient information, and revisions to clinical guidelines with inpatients and ward staff. The findings could change how psychiatric wards balance patient safety with patient dignity, potentially reducing the use of a resource-heavy practice that may not be helping—and may be harming—the people it is meant to protect.
View original technical description
Research questions Is Enhanced Patient Observation (EPO) acceptable to psychiatric inpatients and ward staff? Is it used differentially for specific patient groups? Is EPO effective and cost-effective in reducing risk of self-harm (SH) and suicide during psychiatric admission? Could EPO be improved? Background EPO of patients by staff is used widely on psychiatric inpatient wards in England, consuming significant resources. Although recommended in NICE guidelines there is no good evidence it reduces the risk of suicide and SH. Qualitative evidence and our patient/staff consultation shows that patients can find it intrusive, coercive and unsettling. Newer methods of analysing observational data can overcome the cost and time implications of randomised controlled trials (RCTs) and have equivalent validity to RCT findings, yet have not been applied to this context. Aims and objectives Our aim is to evaluate the cost-effectiveness of EPO for suicide and SH, establish its acceptability and explore suggestions for change. Our objectives are: to collect quantitative (cost/activity) data on the practice of EPO on general adult psychiatric wards in England, and qualitative data on its acceptability to patients, carers and ward staff; to analyse routine clinical data to establish the effectiveness and cost-effectiveness of EPO in reducing suicide/SH; to bring this evidence together to co-produce staff training, patient information, revisions to clinical guidelines, and policy advice to improve the safety and experiences of staff and inpatients. Methods We will collect EPO cost/practice data for all general adult NHS psychiatric units in England, for use in our CEA. We will conduct individual qualitative interviews with current inpatients and clinical staff nationally with experience of EPO to explore their experiences and potential suggestions for change. Using electronic health records (EHRs) for two mental health trusts, we will develop an algorithm that can identify patients on EPO (and indication) for each bed day and describe their characteristics. We will test whether neurodivergent patients and those from ethnic minority backgrounds are more likely to be placed on EPO. Using a target trial emulation approach we will analyse EHR data in two trusts to establish the effectiveness and cost-effectiveness of EPO in reducing risk of SH and suicide on psychiatric wards. Bringing all this evidence together in a set of separate workshops for inpatients and ward staff, we will start a process of co-creating training, patient information, and policy briefings, and suggesting revisions to local/national clinical guidelines to improve EPO’s therapeutic value. Timelines for delivery In Y1 and Y2 we will collect national cost/activity/qualitative data and create our algorithm. In Y2 and Y3 we will model target trial emulation. In Y3 workshops we will discuss findings to create staff training, patient information, policy briefings, and suggested revisions to clinical guidelines, using diverse methods of dissemination to reach all stakeholders. Anticipated impact and dissemination Our close working with clinical leaders and commissioners will ensure that the staff training package we create will be integrated into existing training infrastructures. This, and our wider dissemination of findings, will ensure that the evidence we create will translate rapidly into benefits to patients, staff and all stakeholders.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know