Frequent users of the Emergency Department: Improving and standardizing services- a mixed methods study

The University of Leeds requires access to the Data to undertake research to understand patterns of frequent use, health costs and impact of current services on frequent users of Emergency Departments (EDs) in England. This research is one of four workstreams of a mixed methods project of frequent users of EDs and the services provided for this group of patients, the FUsED study (Frequent users of the ED: Improving and standardising services - a mixed methods study).

Approximately 2.5% ED users account for 10% of total attendances. The NHS Long Term Plan in 2019 outlined an Urgent and Emergency Care Framework to try to reduce the frequency of attendance at EDs but there is little evidence to suggest that alternatives within the urgent and emergency care framework have achieved the intended effect of reducing ED visits. Evidence suggests the proportion of frequent to non-frequent users is rising in the UK.

Frequent users of EDs are a heterogeneous population and groups of frequent users have been described based on: a) pattern of attendance e.g. persistent frequent users (>12 months), multisite attenders (25% of all frequent users, more likely to have mental health problems) or 'burst' attendances (may resolve spontaneously); b) reasons for attendance e.g., illness concerns, suicidal thoughts/self-harm, recurrent emotional crises, social difficulties, substance misuse, and c) demographics and/or diagnosis e.g., elderly physically unwell.

There is no agreed definition of frequent use of the ED; for the purposes of this study, the research team will define frequent attendance as 5 or more attendances at the ED in a 12-month period which was recommended in a recent systematic review of ways to define frequent use. National data show that 31,492 people made 10 or more ED visits in England in 2017-18. Total hospital costs of frequent users are 3-4 times greater than routine ED users.

Frequent attendance at the ED is often a sign of distress and unmet need. Frequent users have high rates of multimorbidity, psychiatric co-morbidities and psychosocial problems. Failure to recognise and address frequent users' needs can potentially lead to a cycle of re-attendance. Frequent users have high rates of 12-month mortality (10-13%), most commonly attributed to suicide or alcohol /drug abuse.

Recent systematic reviews have examined the effectiveness of interventions to decrease ED visits by adult frequent users. The evidence for reductions in attendance from the small number of randomised controlled trials that have been conducted is very weak. Uncontrolled evaluations of frequent user interventions have suggested these interventions generate large cost savings to the NHS. However, such uncontrolled evaluations usually result in large overestimates of service impact, most likely because of regression to the mean.

Considerable investment (of the order of £120 million) has been put into ED liaison mental health services, so that all EDs in England now have a Liaison mental health service. It is known from scoping work conducted by one of the programme team members that services for frequent users have been rapidly developing and that most services for frequent users in the UK are currently delivered by Liaison mental health services. The last annual survey of Liaison mental health services (April 2019) showed that out of 170 acute hospitals with EDs in England, 80 now have specific services for frequent users, with others planning service developments. The programme team have identified 4 types of service: basic care planning with no designated staff; services with ED designated staff; services with designated Liaison mental health service staff; and services based outside ED. The team found that services for frequent users are developing in a piecemeal fashion, providing a range of non-evidenced based interventions for frequent users including case management, social prescribing, and specific frequent user clinics.

The aims of the broader FUsED mixed methods study are to improve frequent user services for Urgency and Emergency Care (UEC) by:

-Describing current patterns and costs of frequent UEC use of the services for frequent users and the impact of frequent user services on attendance; identify prior and early predictors of persistent frequent use

-Identifying which interventions appear to work (provide a safe and appropriate response, whilst reducing UEC use) for which types of frequent user in what settings and why.

-Testing these findings by in-depth comparative studies of four different types of services for frequent users and their UEC networks

-Producing and disseminating an implementation framework for frequent user services to optimise care.

The overall objectives of the FUsED study are to produce:

-A framework for describing and characterising frequent user UEC services (Workstream 1a & 1b)

-An analysis of use by frequent users of the ED and wider UEC networks, including different patterns of attendance (Workstream2a), multisite use (Workstream 2b) and healthcare costs of frequent users (Workstream 2c).

-An analysis of the impact of current services for frequent users on overall ED use (Workstream 2d) and a sub analysis of the impact of COVID-19 on frequent use of ED (Workstream 2d)

-A realist synthesis, which describes Programme theories about how interventions may work for frequent users in different contexts (Workstream 3)

-An implementation framework to help plan and optimise frequent user services (Workstream 4)

This data sharing agreement (DSA) relates to workstream 2. The specific aim of workstream 2 is to identify patterns of attendance, multisite use, and health care costs of frequent users and to evaluate the impact of current services for frequent users on overall ED use.

The specific objectives of workstream 2 (this DSA) are to

-Characterise patterns of frequent ED use and their associations with patient characteristics; identify early predictors of persistent frequent use.

-Examine use of the whole UEC network by ED frequent users

-Describe the costs (ED and inpatient) associated with frequent use

-Conduct interrupted time series analysis of frequent ED use to understand the impact of (i) initiation of services for frequent users (ii) the COVID-19 pandemic.

The following Data will be accessed:

• Hospital Episode Statistics Accident & Emergency and the Emergency Care Data Set (ECDS) – necessary to show ED attendances at hospital (or hospital Trusts) in England. Attendance data will include date, time, ED mode of arrival, reason for attendance, primary and secondary diagnosis, primary and secondary investigations, and hospital frailty score. Basic demographic characteristics (excluding identifiers), reasons for attendance, diagnosis, associated hospital admissions and other healthcare use data will also be collected.

• Hospital Episode Statistics Admitted Patient Care, Critical Care and Outpatients – necessary to determine wider use of healthcare and costs associated with frequent use of emergency services

• Mental Health Services Data Set – necessary to determine wider use of healthcare and costs associated with frequent use of emergency services and frequent use of emergency services is associated with high rates of mental health problems

• Medicines Dispensed in Primary Care – necessary to give a more complete measure of the costs of patients and to check whether certain types of medication may act as flags in predictor models

• Civil Registrations (Deaths) Secondary Care Cut – necessary because frequent use of emergency services is associated with high mortality

The datasets will be used to examine different patterns of attendance at A&E (e.g. bursts of attendance, persistent attendance or multisite attendance, different thresholds for defining frequent use, wider healthcare resource use associated with frequent attendance, impact of frequent user services on ED attendance). All episodes are required from the datasets above to 1) build an unbiased predictive model to identify people at high risk of reattendance from the general population and 2) analyse the wider healthcare usage and cost associated with frequent attenders, and frequent attender services (as previous research indicates a trust-wide effect after implementation of such services).

The level of the data will be pseudonymised.

The data will be minimised as follows:

• Limited to data for a study cohort meeting the following criteria: adults aged 18 years and above, attendance at an A&E department during the study period.

• Limited to data between April 2016 – March 2021

• Limited to pseudonymised data only;

Only fields instrumental to the analysis of the key research questions have been requested. The University of Leeds wish to use the data to carry out analyses to examine patterns of attendance at A&E for the whole adult A&E population, in order to determine appropriate thresholds for defining frequent use and to undertake predictor analyses to identify frequent use at an early stage and also persistent use. The whole adult population is also required for health economic evaluations in order to compare costs and healthcare use resource associated with frequent use in comparison with non-frequent use. Nation-wide data is required to track patients who attend at multiple sites, as previous studies have shown some evidence of people moving between multiple hospitals – particularly when frequent attender services are in operation. Multiple years of data are required to establish a baseline number of visits in each month, and look at the impact after frequent attender services have been introduced.

The University of Leeds is the research sponsor and the data controller as the organisation responsible for ensuring that the data will only be processed for the purpose described above.

The lawful basis for processing personal data under the UK GDPR is:

Article 6 (1)(E): Processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller.

The lawful basis for processing special category data under the UK GDPR is:

Article 9 (2)(J): Processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject.

This processing is in the public interest because it adheres to the UK Policy Framework for Health and Social Care Research and aims to produce generalisable and publicly available evidence for planning future services for frequent users of Emergency Departments, and optimising current services to improve outcomes for patients and generate cost-savings for the NHS.

This study forms part of a National Institute for Health Research (NIHR) funded research Programme, from the Health and Social Care Delivery Research (HS & DR) funding stream (132852) which seeks to improve and standardise services for frequent users of emergency services.

The University of Leeds use cloud hosting services provided by Microsoft Limited.

Academics from the University of Sheffield, University of York, Yorkshire Ambulance Service NHS Trust and Cornwall Partnership NHS Foundation Trust will not have access to the Data. These organisations are involved in the wider research project and lead on the other workstreams unrelated to the processing of the Data.

The research oversight committee consists of academics from University of Manchester, University of West England, University of Cardiff, University of Oxford, King’s College London and an expert by experience. They will not have access to the Data.

All individuals who will access the Data are substantive employees of the University of Leeds.

No data will flow to NHS England for the purposes of this Agreement.

NHS England will provide the relevant records from the HES, ECDS, Mortality, Mental Health and Prescribing datasets to the Leeds Analytic Secure Environment for Research (LASER Platform), a cloud-based Trusted Research Environment system hosted in Leeds Institute of Data Analytics (LIDA). The data will contain no direct identifying data items. The data will be pseudonymised and individuals cannot be reidentified through linkage with other data in the possession of the recipient.

The LASER Platform provides a secure environment for the handling, processing, analysis and storage of sensitive and confidential research data, segregates projects at network level, using separate Virtual Research Environments (VRE), protecting data from the rest of the University's computing facilities and from internet access. It is managed and maintained by University of Leeds IT Services and operated and supported by LIDA's Data Analytics Team (DAT).

The University of Leeds stores data on the Cloud provided by Microsoft Limited.

No other data flow to or from NHS England will be undertaken as part of this research project.

Flags for when and what interventions were introduced will be integrated into the NHS England data, in order to analyse the impact of hospital trusts introducing interventions for people who were frequent attenders. These flags will be generated as part of another workstream from this programme, and combined with the record level NHS England data.

The aggregated information derived from the NHS England data will be analysed in parallel with aggregated data from the CUREd database, which covers all hospitals with an ED in the Yorkshire and Humber Region plus associated 999/111/and ambulance data. CUREd has 3.8M attendances over 3 years, covering approximately 88k frequent users.

The data will be accessed onsite at the premises of University of Leeds. The data will also be accessed by authorised personnel via remote access. The data will remain on the servers at Microsoft Limited at all times.

Personnel are prohibited from downloading or copying data to local devices.

The data will not leave the UK at any time.

Access is restricted to employees or agents of University of Leeds who have authorisation from the Principal Investigator.

All personnel accessing the data have been appropriately trained in data protection and confidentiality.

There will be no requirement and no attempt to reidentify individuals when using the data.

Researchers from the University of Leeds will process the data for the purposes described in ‘Objectives for Processing’.

Results are expected to be published in open access and peer reviewed journals, including publication via the National Institute for Health Research's own journal library, and a paper on “Frequent users of the emergency department: improving services and identifying evidence-based interventions-a mixed methods study” for submission to other appropriate journals. The analysis of the datasets requested are planned to be presented at research meetings, such as national and international liaison psychiatry meetings. These conference presentations and papers will report aggregated results across patient episodes and the results will be based on statistical analysis generated from the data (typically in the format of tables, graphical representations, and text).

The outputs will not contain NHS England data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.

Longstanding and ongoing engagement with stakeholders, including both scientific and policy-making audiences is expected to provide a direct pathway to impact for the outputs of this research. The Experts by Experience in the team should help inform the dissemination strategy and play an active role in the format and content of academic papers, as well as supporting presentation at local, regional, and national conferences and wider stakeholder meetings.

The results and outputs of the study are planned to be communicated further via the study team's websites, social media accounts and through other public promotion of research utilizing the researchers' networks, including clinical networks, scientific networks, and charitable organisations.

The main outputs that are intended to be produced from this study are:

-Characterization of the different patterns of frequent use of UEC services and associated factors; Detailed costs related to the pattern and use of UEC services by frequent users.

-A robust evaluation of the impact of frequent user services including cost-effectiveness.

-Identification of specific interventions for certain sub-types of frequent users and the mechanisms by which they achieve their desired outcomes.

-An understanding of the impact of COVID-19 on frequent use of UEC services.

-Publications of the main research findings/conference presentations, information across different media outlets including the study website, social media (using podcasts and twitter), PPI organisations links, relevant charities.

The programme team plan to use a multi-level approach to transfer information about the study findings to relevant stakeholders. The top-down approach will involve targeting six key groups: NHS England, The Royal Colleges of Psychiatry, Emergency Medicine and General Practice, Ambulance Services, and the NHS Confederation. The intention is to convene a workshop of relevant national stakeholders and PPI representatives to present the findings from the Programme, and to discuss the recommended improvements to services.

At an intermediate level, the plan is to hold 4 webinars (each one focusing on a particular service type) which will be jointly presented by members of the programme team and key personnel from the case site of relevance, and which will be targeted at the frequent user services that participated in work stream 1 of FUsED.

At the bottom-up level, the intention is to hold 4 further webinars which will again focus on a particular service type. All remaining frequent user services would be invited to attend one of the webinars, according to their type of service. It is intended that any relevant feedback from the webinars would be incorporated into the team's implementation framework.

The programme team intend to co-develop with the patient and public involvement (PPI) reference group plain English summaries, voice-recordings and podcasts of findings and make these available online through blogs and social media.

This programme is expected to provide evidence as to which groups of ED users should be targeted; what kind of interventions should be delivered; how those interventions achieve their desired outcomes; and how outcomes should be measured.

The expectation is that the outputs of this study will reach the health and care system via established links with NHS England, the Royal Colleges of Psychiatry and Emergency Medicine and via direct contact with all frequent user services by a cascading series of webinars as detailed in ‘Expected Outputs’.

The findings are hoped to enable UEC networks to make informed planning choices about services for frequent users. The programme team intend to provide best evidence on how interventions work (or fail) in order to guide on-the-ground delivery. The large investment in liaison mental health services (LMHS) over the last 4 years means that NHS stakeholders should be keen to receive evidence of the impact of LMHS on frequent user services. By incorporating patient and public involvement (PPI) in the research, it is expected that awareness of constructive approaches to this important issue will increase.

The findings of this research study are expected to contribute to evidence-based decision-making for policy-makers, local decision-makers such as doctors, and patients to inform best practice to improve the care, treatment and experience of health care users relevant to the subject matter of the study.

The use of the data could:

• help the system to better understand the health and care needs of populations.

• lead to the identification or improvement of treatments or interventions, or health and care system design to improve health and care outcomes or experience.

• advance understanding of regional and national trends in health and social care needs.

• inform planning health services and programmes, for example to improve equity of access, experience and outcomes.

• provide a mechanism for checking the quality of care. This could include identifying areas of good practice to learn from, or areas of poorer practice which need to be addressed.

The aim is to improve services for UEC frequent attenders by understanding which activities are carried out by services and what difference they make. This is expected to enable us to develop programme theories which includes which aspects of services are important, how they work and which patients they help most. An implementation framework can then be developed to improve services planning and focus on activities likely to bring benefits to the patients.

A Public and Patient Information and Engagement group was consulted regarding the collection of the data for the purposes described above. Two experts by lived experience make up part of the programme team. They attended the design study meetings, commented on all drafts of the study application for NIHR funding and had individual meetings with the principal investigator of the study, contributing to all workstreams of the programme and helping to set up the dissemination strategy.

As part of the study plan, a PPI reference group was set up to follow the progress of the study and assure PPI input in all stages.

To inform and advertise findings to a wider audience and optimise the potential public benefits from the use of the data, the programme team intend to use strong and established links with:

NHS England; the Current Chair of the Faculty of Liaison Psychiatry; the Royal College of Emergency Medicine and its mental health committee; The Association of Ambulance Chief Executives (AACE) who provide ambulance services with a central organisation that supports, coordinates, and implements nationally agreed policy; the Royal College of General Practice; and the European Association of Psychosomatic Medicine to reach a wider European audience.

The principal investigator and experts by lived experience co applicants plan to lead on developing a strategy to inform users and carers on a local and national basis about the findings. The intention is to approach NHS Choices and offer to help with updates as a result of the study.

The work to data has contributed to the growing evidence base around frequent users (FUs) of emergency departments (EDs) small group accounting for disproportionate demand in overstretched NHS systems.

To date, the following benefits have been delivered:

- Informing integrated care models: Interrupted time series analyses show that community-based integrated FU services see greater reductions in ED attendances and potentially lower costs especially among high-intensity users and those with mental health needs. These findings support business cases for community-led provision.

- Targeting care through subgrouping: Latent class analysis identified three core FU subgroups that were consistent across settings, reflecting different patterns of need and service use. Characterising these subgroups is important for developing more tailored interventions, beyond the current focus on psychosocial needs, to include, for example, older adults with frailty or people with multiple long-term conditions. This supports the extension of current frequent user services to better meet the diverse needs of high-intensity service users and improves identification strategies across the system.

System engagement and impact: Results have been shared with NHS teams delivering FU services across the region, contributing to service planning. We have also involved Public and Patient Involvement and Engagement (PPIE) contributors to explore how findings can enhance patient experience and care coordination.

- Policy engagement: A summary of findings has been presented to the Department of Health and Social Care. Planned cost impact modelling of FU services will further inform commissioning and national strategy.

- This work supports NHS aims to improve urgent care performance, reduce health inequalities, and promote more person-centred, integrated care. It also demonstrates the value of NHS England data in delivering impactful, policy-relevant research.