Evaluation of community-based health and social care multi-disciplinary teams (MDTs) - data linkage and comparison patients

On 1 February 2023, NHS Digital merged with NHS England. NHS England has assumed responsibility for all activities previously undertaken by NHS Digital. The merger was completed by a statute change. Any reference made to NHS Digital within this Data Sharing Agreement is in reference to the merged organisation known as NHS England.

The Policy Innovation & Evaluation Research Unit (PIRU), based at the London School of Hygiene & Tropical Medicine (LSHTM), is carrying out a long-term programme of research on the Integrated Care & Support Pioneers (Pioneers) in order to identify factors that enable or inhibit progress towards the joining up (integration) of health and social care services and to assess whether such integrated services lead to better outcomes for patients in a more patient-centred and cost-effective way. Since the current project began in 2015, there has been an increasing number of initiatives in England designed to better integrate the health and care systems, most recently the development of Integrated Care Systems (ICSs) in all parts of England which will be placed on a statutory basis from June 2022. Thus the need for solid evidence on how best to deliver integrated systems and integrated care is more urgent than ever, particularly in light of the White Paper (“Integration and Innovation: working together to improve health and care for all”) released by the Government in February 2021 and the further health and social care integration White Paper published in February 2022 (https://www.gov.uk/government/publications/health-and-social-care-integration-joining-up-care-for-people-places-and-populations) setting out legislative proposals for integrating health and social care services, the financial austerity of recent years, and the likely impacts of the Covid-19 pandemic on health and care professionals, patients, informal carers, service delivery and coordination between NHS and social care providers. This is the precise focus of the current evaluation. It is unique in being able to examine integration activities in specific areas over a timeframe of over six years, with most previous studies having much shorter timescales; thus, this project is able to look at the extent to which barriers can be surmounted over time, the extent to which integration initiatives can be sustained, as well as any unanticipated side-effects of these initiatives.

The PIRU evaluation is funded by the Department of Health and Social Care (DHSC)/National Institute for Health Research (NIHR) Policy Research Programme (PRP) and covers all of the 25 Pioneers in England. The Pioneer Programme (2013-2018) was a national scheme covering 25 geographically and socio-demographically different areas of the country chosen by DHSC to develop and test new ways of joining up health and social care services https://www.england.nhs.uk/new-care-models/integrated-care-pioneers/#:~:text=Integrated%20Care%20Pioneers%20were%20local,to%20efficiently%20deliver%20integrated%20care). The Pioneer Programme was, in many respects, the precursor to the current policy to establish ICSs throughout England. Further details are available on the evaluation’s website: https://piru.ac.uk/projects/current-projects/integrated-care-pioneers-evaluation.html.

The current research programme, which is due to be completed in autumn 2022, follows on from an earlier evaluation of the Pioneers which was undertaken by the same research team (2014-15), and involves three work packages.

• Work package 1: Implementation and progress – Pioneer level process evaluation and (limited) impact evaluation in all 25 Pioneers via interviews, web based panel surveys and analysis of performance indicators relevant to integrated health and social care;

• Work package 2: Impacts, costs and patient outcomes – impact and economic evaluations of selected Pioneer initiatives using mixed methods, and designed to follow patients, carers and staff over the longer-term;

• Work package 3: Lessons learned – Working with Pioneers, national policy makers and partners, patient/user organisations and experts to derive and spread learning on improving integrated care.

The second work package comprises a quasi-experimental impact and economic evaluation, comparing costs and outcomes for patients on the caseload of community-based health and social care multi-disciplinary teams (MDTs) in two of the former Pioneer areas and a comparison group of patients not receiving MDT care. Such MDTs are the most widely reported integration initiative undertaken by the Pioneers and are common to many health and social care integration initiatives elsewhere in the country. For the current study, MDTs had to: 1) include both primary and social care, as well as allied healthcare professionals, and in some instances, the community and voluntary sector (CVS); 2) bring these professionals together in a shared process of care coordination; and 3) have a target caseload that included people aged 55 and over with multiple long-term conditions, i.e. those who often need complex care coordination and high levels of health and social care support. As well as focusing on service use-related outcomes, this prospective, mixed methods evaluation also involves surveys and qualitative interviews with MDT staff, MDT patients, and their nominated informal carers, as well as observation of MDT meetings.

Specific research questions include:

• Is the health, well-being and life expectancy of patients improved by a community-based MDT approach when compared with similar patients not managed by a community-based MDT?

• Does a community-based MDT approach reduce health and social care service use and costs?

• Do these improvements justify the cost of developing and implementing community-based MDTs?

• How much are the cost and quality of providing health and social care services to patients through a community-based MDT affected by the design of teams (e.g., its service functions, target patient population, team composition and competencies, enabling structure and resources and other contextual factors at the local and national level)?

PIRU is requesting NHS Digital Hospital Episode Statistics (HES) and Mortality data to examine whether patients treated by community-based MDTs are admitted to hospital less frequently than a matched comparison group of non-MDT patients and whether there are any survival advantages of being managed by MDTs. Another aim is to look at any differences in health care costs between these two groups. For the MDT patient group, HES and mortality data will be linked to survey data which have already been collected from patients. For the comparison group, HES data will be used both to match comparison patients with MDT patients, and to examine the outcomes of the comparison patient group.

This application for HES and mortality data has three purposes:

1. To obtain a group of matched comparison patients. Comparison patients will be obtained by using HES data to find matches with the MDT patients. Matches will be established on the basis of selected demographic variables (sex, age, etc) along with hospital use over the past 5 years. The matched comparison patients will be taken from geographic areas not covered by the Pioneer programme or by other national integrated care initiatives such as the NHS New Care Models (Vanguards).

2. To compare the MDT and matched comparison group on key outcomes (emergency admissions, inpatient admissions, mortality, etc);

3. To compare these outcomes between patients on the caseloads of different models of community-based MDTs. The HES data will also allow for the calculation of the comparative service use costs of MDT and non-MDT patients by applying national reference costs to hospital activity.

The COVID-19 pandemic provides an unanticipated opportunity to examine how the pandemic may have affected patient care and outcomes, and whether these differ between MDT and comparison patients. With the policy support of the Department of Health and Social Care, the study has been amended to collect additional primary data for MDT patients through surveys and interviews which, in combination with HES data, will allow an exploration of possible COVID-19-related outcomes after March 2020. The pandemic also highlights the importance of this study, as collaboration between the NHS and social care providers increased in both scale and pace as a result of the pandemic, and has been a significant contributor to current proposals to introduce legislation on integrating these services, as set out in the Government White Paper “Integration and Innovation: working together to improve health and social care for all” (February 2021).

The study is concerned with the experience and outcomes of patients managed through health and social care integrated, community-based MDTs, in two Pioneer areas (Islington and Vale of York) in England. MDT patients who meet the study inclusion criteria are aged 55 years and over, have multiple chronic conditions, live at a private residential address, and were added to the caseload during the recruitment period. MDT patients were sent an invitation to participate in the study together with a baseline postal questionnaire by MDT administrative staff and asked to return it to the LSHTM research team in a freepost envelope, starting in October 2018. Patients returning the questionnaire and agreeing to be followed-up were sent a second questionnaire about 9 months later. At baseline, MDT patients were asked to provide written consent to link their survey data with their hospital data.

To meet the study aims, data is requested for two specific purposes:

Firstly, data covering a period of up to 5 years preceding the start of the study in order to have sufficient information to enable the identification of suitable matches as comparison patients.

Secondly, data that will enable comparisons between the MDT and the comparison patient groups, as well as between patients from different MDTs, on outcomes such as hospital admissions, emergency admissions, re-admissions, mortality, etc. over the approximately 9-month follow-up period. The MDT patient recruitment period ran from December 2018 to September 2019, and data is requested covering the period from January 2014 (ideally) to the end of March 2021.

The GDPR Article 6(1)(e)and Article 9(2)(j) provide the legal basis for the processing of the data:

- Article 6(1)(e): The LSHTM is a public authority as described under Schedule 1 of the FOI Act 2000 and has a basis in law for lawfulness of processing as they have a Royal Charter, and public interest provides justification for data to be released in a pseudonymized format.

- Article 9(2)(j): The data are required for research purposes in the public interest meeting the conditions in the DPA 2018 Schedule 1 Part 1(4) - which GDPR Recital 52(2) determines is an appropriate derogation from the prohibition on processing special categories of personal data, demonstrated in the Objectives

The data requested have been minimized by requesting only those variables that meet the criteria for the analysis from patients aged 55 years or older. The request is for access to mortality data, and HES datasets (inpatient, emergency, and critical care), starting 5 years before the study period, up to the most recent data available . These datasets will provide the key outcome measures to compare MDT and matched comparison patients. The use of HES data is the only means of identifying a suitable matched comparison group; this will be accomplished by examining data 5 years previous to the study recruitment period. Likewise, HES and mortality data are the only means available to examine the key outcome measures and to estimate health care costs for both MDT and comparison groups.

Individual record level data is required as this is the only means by which HES data can be linked to the MDT patient survey data collected by the present study and to identify matched comparison patients. Data for the comparison patients will be pseudonymised to the research team. For MDT patients, consent was obtained for data linkage, but the dataset will be pseudonymised and linked only by a unique study number. There will be no attempts to re-identify these individuals.

LSHTM is the only organisation involved in the process of obtaining the HES and mortality data. LSHTM will provide NHS Digital with unique study number, full name, and date of birth for MDT patients who have consented to data linkage. LSHTM will be the sole data controller who also processes the data, and will hold the linked HES, mortality and survey data for MDT patients, and all the pseudonymised HES and mortality data for the comparison group patients. LSHTM have made all decisions on the way that data would be processed. All data analysis will be carried out by LSHTM staff. Only substantive employees of LSHTM will have access to the pseudonymised data requested and supplied by NHS Digital. Additionally, while the DHSC is commissioning this evaluation, they do not carry out data controllership activities.

In order to obtain HES and mortality data for MDT patients within the two Clinical Commissioning Group (CCG) areas included in the study, LSHTM will provide NHS Digital with unique study number, name, and date of birth for those 441 MDT patients who consented to data linkage of their survey data with HES and mortality data. The patient consent form said that this information would be passed to the NHS Commissioning Support Unit (i.e., NEL CSU), which was the intention at the start of the study. However, by the time the study ended, this route for obtaining data linkage was no longer an option, and it was decided that these details should be passed by LSHTM directly to NHS Digital, without the intervention of NEL CSU. Following advice from IGARD, the research team carried out a consultation to determine whether patients had any objections to their details being passed by LSHTM directly to NHS Digital, and all contacted patients agreed with this proposed change.

NHS Digital will use name, date of birth, and CCG to trace NHS numbers for consenting MDT patients and then use NHS numbers to link to these consenting patients' HES and mortality data. After removing NHS number, name, date of birth and CCG, NHS Digital will provide the HES and mortality data directly to LSHTM, who will link this data with the survey data for individual MDT patients using the unique study number.

For the comparison patients, NHS Digital is asked to provide the data items requested for all patients born before 1st October 1964 (and who were still alive on 1st October 2018 when the study started data collection). For the comparison patients, the request does not include any identifiers or potential identifiers, so the data provided by NHS Digital will be pseudonymised. The data will be provided directly to LSHTM. Only variables needed for the research have been requested, and only for patients relevant for the research, i.e., those aged 55 and over at baseline. The comparison (control) group comprises patients aged 55 years or older living at a private residential address located in an area not covered by the Pioneer programme and by other national integrated care initiative. The years requested cover the time period needed to identify a suitable comparison sample and to allow follow-up (i.e., 5 years before baseline to latest available).

To identify the matched comparison group, common matching algorithms (e.g., nearest neighbour matching) will be applied to a pool of patients aged 55 years or older living at a private residential address located in an area not covered by the Pioneer programme and by other national integrated care initiative. The use of this matching approach will ensure that the requested key observable socio-demographic variables are balanced between the comparison groups and that individuals in the matched comparison group are not subject to the influence of similar interventions. The balance on the requested key observable variables (e.g., age, sex, ethnicity, deprivation) is a necessary condition for the unbiased identification of the MDT’s treatment effects. Data from patients younger than 55 years old and not living at a private residential address have been excluded from this request as they were not the target population of MDTs and there is no expected effect of the intervention in this population. The years requested cover the time period needed to identify a suitable comparison sample and to allow follow-up.

Only pseudonymised data will flow to LSHTM and there will be no attempts to re-identify these individuals. The pseudonymised dataset used for analysis will not include identifiers and will be stored on the secure server at LSHTM. (The paper consent forms are kept securely in a locked office at LSHTM.)

The dataset which includes HES and mortality data will be saved on LSHTM’s secure server, which has been certified as meeting NHS Digital’s security standards. The dataset will only be accessible by authorised members of the LSHTM project research team. All LSHTM staff accessing these data have undergone training in GDPR and data protection regulations.

The planned outputs include briefings and presentations to policy and management staff at DHSC and other national policy agencies such as NHS England and Improvement (NHSE/I) as well as published papers in high-profile, peer-reviewed scientific journals (previous articles on this evaluation have been published in BMC Health Services Research and the Journal of Public Health) and presentations at national/international academic, policymaker and practitioner seminars/ conferences/workshops (e.g., briefings with DHSC and NHSE/I, possibly joint events with the Nuffield Trust and the Health Foundation). These will form part of a package of outputs covering the wider long-term evaluation of the Pioneer programme, and all journal articles will be brought together within an over-arching ‘portfolio’ report highlighting the policy implications of the various strands of the evaluation and which will be made available to policymakers within the DHSC, NHS England, local authorities, Clinical Commissioning Groups (CCGs) and other relevant stakeholders. The research team has regular contacts with DHSC staff and other researchers involved in integrated care research and provides an annual report to DHSC covering progress and key findings of the research programme. Presentations will also be offered to the NHS and local authority social services organisations in the two localities involved in this particular component of the wider Pioneer Programme evaluation.

To facilitate dissemination among a wider audience, lay summaries of key outputs will be provided and the research team will write blogs and articles for newsletters (e.g., Health Service Journal) and newspapers, as well as engage with the mainstream media (e.g., press releases, interviews). LSHTM has an active press office, with good connections to both specialist and mainstream media outlets, and the project team will work closely with the press office to promote the project among both key stakeholders and the wider public. Moreover, all outputs will be available on the project’s website (https://piru.ac.uk/projects/current-projects/integrated-care-pioneers-evaluation.html).

All publications derived from record level data will contain only aggregate level data without patient identifiers and with suppression of small numbers in line with the HES analysis guide.

The wider Pioneer evaluation already has a number of publications and other outputs. Outputs based on patient HES and mortality data will start to be produced in the Second Quarter of 2022 and are likely to be published later in 2022. One paper on the MDTs was submitted in late 2021 and is currently in revision with a journal and others will be submitted during 2022.

The specific component of the wider evaluation which this application supports focuses on the most common local integration intervention in England, i.e., the implementation of community-based MDTs, of which there are different models. By having a matched comparison group of non-MDT patients, and using HES and mortality data, it should be possible to examine whether MDT patients have better outcomes than the comparison patients. Also, by looking at a number of different models of MDTs within our evaluation, it should also be possible to look at whether some models lead to better outcomes than others, and which models are most cost-effective. By combining the quantitative analysis with the extensive qualitative data collected within the MDT sites, it should enable a better understanding of the mechanisms generating any different outcomes within the different models of MDTs. The qualitative data generated during the evaluation should provide a unique combination of perspectives from patients, informal carers, and strategic and frontline staff on the receipt, planning and delivery of care respectively, including since the start of the Covid-19 pandemic, which could contribute to the future development of MDTs, as a key intervention designed to provide integrated care. In turn, better designed MDTs involving health and social services staff coordinating care could benefit patients living in the community who have multiple long-term conditions that require the skills and services of a range of health and local authority staff.

The ultimate aim of the wider evaluation, of which the analysis of patient outcomes using the HES and mortality data is a part, is to provide policymakers with better evidence on how best to integrate health and social care services in a cost-effective way. Given limited resources within both NHS and local authority social care budgets, these results should be of clear public benefit to the extent they can contribute to improved patient outcomes, and more efficient use of resources.

The research team at LSHTM is part of a Policy Research Unit (PIRU), funded by the NIHR Policy Research Programme. Its goal is to improve evidence-based policymaking and implementation across DHSC, NHS England and other arm’s-length bodies, and it has a wealth of experience working with policymakers in government.

Not stated in the register.