Community Health Services: ESHCRU Policy Research Unit projects

The Economics of Health Care and Interface with Social Care II (ESHCRUII) is one of 15 Policy Research Units (PRUs) funded by the NIHR Policy Research Programme. Its predecessor, the Economics of Health Care and Interface with Social Care (ESHCRU) programme, was re-awarded for a further five years starting from January 2019. The aim of ESHCRUII is to inform and guide policy making in the health and social care sectors by undertaking high quality, robust and policy-relevant research, based on the discipline of economics, thereby helping to improve the health and well-being of the population, reflecting distributional concerns and population diversity. ESHCRUII involves researchers at the University of York and the Care Policy and Evaluation Centre (CPEC) within the London School of Economics (LSE).

For individual ESHCRUII projects, either LSE or the University of York takes the lead. University of York has no involvement in research projects conducted by LSE, and vice versa. For the purposes of processing requested in this data sharing agreement, University of York will not have access to the data provided under this agreement. LSE will be solely responsible for determining the purposes for and means of the processing of the personal data and decision-making related to the data and its analysis. University of York has no involvement in determining the purpose and means of processing and is not carrying out any data controllership activities.

The NIHR/DHSC has established an Oversight Group for ESHCRUII, as for each of the other PRUs, meeting twice a year to receive progress reports on projects and to discuss strategy. This Oversight Group does not in any way control the analysis of data. The NIHR/DHSC are funders of the research and as such are not directly involved nor responsible for decision-making related to the data, their analysis, and outputs.

Community health services (CHS) constitute an important part of the NHS. They provide community nursing, therapy and other services to people in the own homes or elsewhere in the community. They are important for the care of people living in the community with a range of health conditions that require nursing or therapy services in their own homes and can have a valuable role in promoting independence, preventing hospital admission, and expediting hospital discharge. This role is especially important as the NHS faces a huge waiting list for hospital elective care.

There is currently limited research evidence on the CHS, relative to evidence on primary care, hospital care and adult social care. The NIHR/DHSC has therefore outlined their support for ESHCRUII to conduct the two related projects for which NHS England data are sought through this application.

The purposes outlined in this data sharing agreement refer to two specific projects under ESHCRUII:

- PR-PRU-1217-20301(10): Demand for community health services for adults

This project will examine the use of the Community Health Services (CHS) by different groups of patients and by geographical area. The project will explore how service use varies by age, gender, ethnicity and health condition (broadly defined) and also by features of the patient’s area of residence, such as its rurality (that is, if the area is rural or urban) or how disadvantaged it is. This evidence can help policy makers to tackle health inequalities.

The project will also produce projections of demand for CHS nationally and locally for the next 10 years. It will involve analyses of linked data from the Community Services Data Set (CSDS) and Hospital Episodes Statistics (HES) and the production of a simulation model to produce the projections. The projections will reflect official populations projections on the numbers of people by age and gender and will not take account of future changes in policy or future patient preferences.

- PR-PRU-1217-20301(11): Interaction between community health services (CHS) and hospital care

This project will focus on examining use of the Community Health Services (CHS) by hospital inpatients and outpatients with health conditions for which their use is important. It will also consider the use of adult social care (ASC) by these patient groups by age and gender, drawing on findings from other studies (e.g. work conducted by the Adult Social Care Policy Research Unit – see https://www.ascru.nihr.ac.uk/), so that use of CHS and ASC can be compared.

The NIHR has stated that this is important work that will help inform central and local government policy on community

healthcare and having access to linked data is vital to further this understanding. This is also a sector the NIHR believe will play a crucial role in recovery from the COVID-19 pandemic.

The project will involve analyses of linked data from the Community Services Data Set (CSDS) and Hospital Episodes Statistics (HES). It will include consultation with clinical experts and policymakers about the health conditions for which use of the CHS is likely to be especially effective.

Since there is limited data available on the CHS in England, the CSDS is one of the few data sources on these services. It would be impossible to address the research questions in the public interest without use of CSDS data linked to HES data. The LSE researchers request access to pseudonymised, record-level data focused around use of community health services, admitted patient care, outpatient care, A&E attendances, service users’ characteristics (e.g. age, gender, employment status, receipt of social care, date of death where applicable) and area characteristics (e.g. deprivation and rurality).

The core data set LSE request is the Community Services Dataset (CSDS). CSDS is a nation-wide data set which includes care contacts recorded from a range of community settings. These activities could take place in settings such as health centres, community centres, mobile facilities, or a patient's own home. As such the scope of the CSDS is vast, for example, in 2021-22 in CSDS there were 17.7 million referrals relating to 8.2 million people. There were also 94.4 million care contacts.

LSE analysts request linked Index of Multiple Deprivation (IMD) data and rurality data to address a key objective of the first study, to examine variation in use of the CHS by deprivation and type of area. LSE request linked HES admitted patient care and outpatient data to address a key objective of the second study, examining the interrelationship with hospital services for those receiving CHS. For this purpose LSE request CSDS data for both those who use hospital inpatient or outpatient services and those who do not use these hospital services so that the study may examine how the characteristics and use of CHS varies between these two groups of patients.

The following NHS Digital data will be accessed:

• Community Services Dataset - necessary as this is the primary resource for assessing patient use of community services nationally.

• Hospital Episode Statistics Datasets & Emergency Care Dataset (ECDS) - necessary to provide information on use of secondary care services by patient included within the CSDS data.

• Civil Registration Mortality – necessary so that the project can be sure not to include people in the analysis after their deaths, which would adversely affect the quality of the analysis.

LSE have considered carefully the limitations feasible to the data requested without undermining the ability to conduct the analyses. LSE have limited the request in the following ways:

• Age group: LSE do not require data for children (aged under 18) as initial analysis will focus on adults use of CHS. The project currently focuses on use of CHS by adults only as the policy areas the project intends to support assessment and change within focus on those aged 18 & over and has been referenced within the projects NIHR funding grant.

• Years: will limit analysis to three years – 2019/20 to 2021/22 - which will suffice to achieve the studies’ objectives while including one pre-pandemic year,

• Variables: LSE have selected the appropriate variables specified in this agreement.

• Geography: LSE require national data as the analysis will include assessing the use of community services across different geographies.

• Episodes: LSE require access to all episodes within the periods selected in order to accurately assess patient use of CHS over time, view patient journeys through these services and capture any variations in the use of these services that will require indicating within the analysis.

LSE plan to investigate whether the use of community health services varies by age, gender and other patient characteristics and how the use of CHS is correlated with use of admitted care and outpatient care. These analyses are hoped to enable LSE to understand (1) what the future trends in the use of community health services will be as the demographic structure of England continues to change, and (2) to what extent the use of CHS interacts with, affects, or is affected by use of inpatient and outpatient hospital care.

The purpose of the research cannot be achieved in a less intrusive way, as record-level data is necessary to explore how use of services varies by individual-level characteristics. Conducting analyses on the basis of record-level data is a crucial step to understand variations in use of the CHS and interaction between the use of CHS and hospital services as well as a prerequisite to making projections of future service use. Without record-level data, the academic rigour of the analyses and projection modelling cannot be achieved.

This request to process NHS England data is made on the legal basis outlined in the GDPR, Article 6.1.(e). That is, the purpose establishes the legal basis for processing special categories of personal data being necessary for archiving purposes in the public interest, Article 9(2)(j) of the GDPR. The processing of sensitive personal data is in the public interest as the results of this work will help to identify the ways in which services can be improved and patients and treatments better matched, informing evidence-based health policy on how to improve the effectiveness of community health services in the UK.

There will be no flow of data from the London School of Economics into NHS England. The following pseudonymised data will flow from NHS England to LSE:

• Community Services Dataset (CSDS)

• HES – Admitted Patient Care

• HES – Outpatient

• Emergency Care Dataset

• Civil Registration – Deaths (Secondary Care Cut)

The data will be processed by the LSE Secure Research Computing Governance Group. Amazon Web Services (AWS) provides the hosting environment of the LSE Secure Research Environment. All processing is done by the end user, namely researchers at the LSE who analyse data on the server after connecting.

Amazon Web Services (AWS) supply support to the system, but do not access the data. Therefore, any access to the data held under this agreement would be considered a breach of the agreement. This includes granting of access to the database[s] containing the data. The Amazon Web Services (AWS) datacentre provides cloud storage of the data only - the processing is conducted by individuals substantively employed by LSE only.

Once the data has been received at the LSE Secure Research Computing Governance Group, the data will then be analysed in the LSE Secure Research Computing Governance Group Environment, exclusively by the LSE researchers. All outputs produced will be aggregated with small numbers suppressed.

The data will be cleaned, with duplicate records removed. Outliers and impossible values will be carefully detected and checked before they are dropped. Researchers will take the following steps to examine missing values in the dataset. First, the proportion of missing values for each variable will be calculated. Second, analyses will be conducted to examine how the patterns of missingness are correlated with variables without missing values. This step will allow researchers to evaluate whether the mechanism of missingness is missing completely at random, missing at random, or missing not at random. Finally, examination of patterns will be used by analysts to substitute values for missing data. Researchers will build imputation models based on the missing mechanisms evaluated in the second step. Such a strategy will help researchers make full use of the information available and minimise the risk of biased estimates caused by missing values. The researchers will then implement the agreed methodology by applying the appropriate econometric techniques and using statistical software packages.

After the study has ended, the LSE would like to store the research data for a further 18 months. This will enable LSE to respond to any comments or peer-review suggestions on the research and to make the necessary adaptations. The data will be stored in the LSE Secure Research Computing Governance Environment with the usual security arrangements in place at all times, and only the researchers will have access to the data.

Data will only be accessed and processed by substantive employees of the LSE and will not be accessed or processed by any other third parties not mentioned in this agreement.

There will be no other data linkage undertaken with NHS England data provided under this agreement that is not already noted in the agreement. There will be no attempts made by the LSE to re-identify individuals involved in this project as there is no requirement to do so.

The main outputs from the data analysis will be reports to DHSC and submissions to peer-reviewed academic journals. Target journals include the Health and Social Care in the Community and Health Economics (or similar journals). LSE will seek to publish in journals that have an open-access agreement with the LSE, as the research output is intended to be made freely available to the general public. If this is not possible and the journals require a fee from readers for access, LSE will make the full content available for free via LSE Research Online, as part of LSE’s commitment to the Open Access model. The target dates for the completion of reports and submission of journal articles will be between 31 December 2023 and 30 June 2024.

In addition to the reports and journal submission, the researchers plan to present research findings in seminars or workshops organised by the Care Policy and Evaluation Centre (CPEC) or the ESHCRU Policy Research Unit. The researchers also plan to present in national or international conferences such as the International Long-term Care Policy Network or Health Economists’ Study Group.

The researchers may also write blogs or other short articles summarising the findings to ensure they communicate the findings of the research to wider audiences, including policy makers, practitioners, service users, and other key stakeholders. These outputs will be published on the official website for this research project.

All figures produced will be aggregate figures. These figures will include proportions, total number of users, coefficients, means, standard deviations, and predicted probabilities. Each figure will contain a sufficient cell count in line with the HES Analysis Guide to avoid disclosure of patient data. All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide.

LSE will consult public advisers, organisations supporting patients, and commissioners and providers of CHS, as LSE conduct the project. The findings will provide evidence to inform national and local planning of CHS. They hope to inform Spending Reviews, policy development and planning of CHS at national and local level, including informing ways for the CHS to address health inequalities. LSE will promote the impact of project findings through offering presentations and discussions to the Department of Health and Social Care, NHS England and relevant professional and voluntary sector organisations.

The findings of the research will also be communicated beyond academia to the public sphere and to citizens through engagement with politicians and stakeholders, seminars, blogs and articles, and social media.

The dissemination is in the public interest in part due to the role CHS can play an in promoting independence, preventing hospital admission, and expediting hospital charge. Yet, there is very little information about the use of the CHS in the public domain. The distribution of the CHS resources in the population and the association between the CHS and admitted/outpatient hospital care have not been well understood so far. It is estimated that around £10 billion of the NHS budget is spent on community health services every year and the demand is expected to keep growing. Without rigorous evidence, it is challenging to accurately assess the impacts and value of the CHS and project potential future changes in demand. The importance of CHS is likely to be underappreciated, these projects aim to expand understanding on the effectiveness of the Community Health Services (CHS) by different groups of patients and by geographical area. The project will explore how service use varies by age, gender, ethnicity and health condition (broadly defined) and also by features of the patient’s area of residence, such as its rurality (that is, if the area is rural or urban) or how disadvantaged it is.

The findings of this research study are expected to contribute to evidence-based decision-making for policy-makers, local decision-makers involved in CHS, 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 outputs (peer reviewed journal submissions and seminars) are hoped to expand the evidence base on the CHS. The information will be valuable to government and NHS policy makers to inform planning of CHS and implementation of policy associated with these services and its links with inpatient and outpatient care. It is hoped and expected that it will be used to inform current policy debates on the integration of different services so that existing resources can be effectively utilised to meet the care needs of the population. The outcomes of this project are hoped to help to identify the effectiveness of the interface of hospital and community healthcare services and the role CHS can play in a patient’s journey. These projects hope to identify the strengths of CHS not currently realised, and how they can support patients following discharge from hospital services, as well as contribute to evidencing how CHS can support demand of health services more effectively/efficiently.

The use of the data could:

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

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

• inform decisions on how to effectively allocate and evaluate funding according to health needs.

• 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.

It is hoped that through publication of findings in appropriate media, the findings of this research will add to the body of evidence that is considered by the bodies, organisations and individual care practitioners charged with making policy decisions for CHS or treatment decisions in relation to specific patients.

No yielded benefits to date due to delays in receiving and analysing data.