Nuffield Trust Primary DSA: Renewal 2024 - 2027

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees with a wide range of experience including perspectives from outside the NHS as well as senior clinicians, NHS managers and academics. It aims to improve the quality of health and social care in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

The data from NHS Digital is vital to the Trust's work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. Data is primarily used in projects relating to quality and equity but has also been used for projects on models of care and legislative reform, for example integrated care policy. The Trust's work aims to serve the public interest, and for some topics, research analysis using patient data is vital.

Under the Health Research Authority's (HRA) GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject". The Nuffield Trust's legitimate interest is carrying out health and care research in the wider public interest of improved policy, services and care for NHS patients. Examples of how the Nuffield Trust process the data to pursue their legitimate interests can be found in the DPIA at https://www.nuffieldtrust.org.uk/about/corporate-policies/schedule-of-current-data-protection-documentation

The Trust relies on Article 9 Condition 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) as the condition for processing "Special" categories of personal data.

5ai. RATIONALE FOR STRATEGIC PRIORITIES AND PROJECTS

The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence applied, and the recommendations made take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable - between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled. The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised. The Trust focuses on policy areas that are relatively neglected, yet of clear significance to the health and care sector.

Current Nuffield Trust projects on quality of care and equity work are described later in this section.

The Nuffield Trust’s other strategic priorities for 2020 to 2025 are Workforce, Primary Care, Small Hospitals, Technology and Digital and Politics, Legislation and Governance.

5aii. USE OF HOSPITAL EPISODE STATISTICS (HES) DATA AND COMMUNITY SERVICES DATASET (CSDS) WITHIN PROGRAMMES AND PROJECTS

The number of concurrent projects using HES data will vary, but typically there are 5-10 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Although the methods for use of HES and CSDS will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken i.e. analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as A&E attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, such as avoidable harm, and patients with particular needs (e.g., frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient follow-up models, digital delivery of care, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis - comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, panel data and cohort studies;

o Use of the above methods in combination for particular research projects;

o Producing visualisations of analysis and results from the above methods.

Monthly HES and Emergency Care Dataset (ECDS) data:

Monthly data ensures that at any point in time the Nuffield Trust have access to the most recent data as soon as becomes available. This is beneficial as the Nuffield Trust have a number of projects in progress to investigate the impact of Covid-19 on the use of health services which has led to rapid changes in health services, and for which monthly data is required.

The data will not be linked with other record level datasets and there will be no attempt to re-identify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g., indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset. Any necessary risk assessment would be carried out as part of establishing the analysis plan during the course of the project planning process. The DPO (who is a member of the Project Planning Group) would advise on any potential issues. Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital will be made.

5aiii. DECISIONS ABOUT PROJECTS AND USE OF NHS DIGITAL DATA

This Agreement permits the Nuffield Trust to use the data for the purposes of projects through the following process:

1. Projects intended will be conceived and planned through an iterative process involving the senior researchers with appropriate input from the Data Protection Officer (DPO). A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. The project lead will ensure:

o Projects have clearly defined objectives and operational plans;

o An analysis plan is prepared for each project, setting out the data requirements and methods;

o In each case, the use of the HES is necessary and proportionate to the purpose of the project and the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

o Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Group (PPG). The PPG consists of the Chief Executive, Directors and Deputy Directors, and representatives from across the organisation including the DPO. PPG provides a forum for discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust. The project template remains valid throughout a project's life, serving as the central control document in the management and delivery of the project. The PPG is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound and

• Draw fully on the expertise within the Trust including making connections to other related work;

The PPG draws on expertise from other organisational committees to ensure that obligations are met in specific areas:

• The Data Strategy Group (responsible for Information Governance)

• The Leadership Team (responsible for Patient / Public Involvement and Equality, diversity and inclusion)

The strategic aims of the Nuffield Trust to inform decisions made to improve health and social care are set out in the following document:

https://www.nuffieldtrust.org.uk/files/2020-08/nutj7957-leaflet-200220-web-pages.pdf

After the project has been approved there will be follow up meeting with the project team and a member of the Leadership Team where details of the project are looked at in more detail including a section of questions to understand the data requirements of the project. A member of the team will be appointed as the IG lead if the size of the project suggests that this is necessary.

3. The team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales: the time frame for undertaking each project will vary according to project resource, extent of the research and data analysis required. This is always considered as part of the Project Planning Group's review and approved based upon the detail of each individual project.

Project Funding: some projects are funded by the Nuffield Trust, but some receive funding from other partners. At the current time, the Trust are receiving funding for projects which use HES data from the National Institute for Health Research (NIHR), the NHS Race and Health Observatory (RHO) and the Health Foundation. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose, as outlined below.

Data Governance: the Nuffield Trust has independently determined the purposes for which it requires and will process data under this Agreement in terms of its priorities and programmes. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the sole data controller named in this Agreement, The Trust cannot be compelled by any third party to process the data for any purpose in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

The Trust do not undertake events for commercial purposes and the data requested as part of this application is not being requested for commercial purposes.

Public Participation: The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence the Trust applies, and the recommendations made, take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust's secure, ISO27001 certified data environment. Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital will be made.

Results may be shared in aggregate form to meet the objectives of the project, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust and will not be shared with any other third parties. The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities. Details will also be available on the Nuffield Trust's website.

Current projects using HES, ECDS or CSDS data include:

• QualityWatch, a programme jointly funded by the Health Foundation which provides independent scrutiny of the quality of care in the NHS and social care, in order to highlight to policy makers and the public about services where quality of care is at risk.

• Prisoner health, a programme funded by the Health Foundation which in this phase will compare the effect of being in prison on use of hospital care, by comparison with a matched case control group (this project is also covered by DARS agreement NIC-195377-M9L8Z). This will inform policy for prisoner health.

• Evaluation of integration pioneers, a long running evaluation funded by NIHR for which the Nuffield Trust is undertaking quantitative analysis as part of a wider project with the London School of Hygiene and Tropical Medicine. The analysis of HES to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data nor does LSHTM determine the purposes or means of processing data under this agreement.

• Rapid Service Evaluation Team funded by NIHR and jointly delivered with University College London (see separate DARS agreement NIC-194629-S4F9X).

• Understanding the impact of local area factors on child health outcomes, in order to identify actions which could be taken at a local authority level, which is an internally funded project using HES data to derive local area measures of activity.

• Provision of Hospice Activity, a project jointly funded with Hospice UK, for which HES and CSDS data is used to provide an additional source of data on activity and assess data quality for the hospice sector and end of life care.

• Ethnic differences in lost activity, treatment rates and times to treatment, a project funded by the NHS Race and Health Observatory (NHS RHO) to understand whether ethnic inequalities exist and inform NHS recovery plans. The NHS RHO does not have access to the data nor does NHS RHO determine the purpose or means of processing data under this agreement.

Further details of projects are included in the outputs and benefits sections.

Project Scoping and Responsive Analysis

Across the Nuffield Trust's strategic priorities, analysis may be undertaken of HES for scoping research and responsive analysis.

Scoping analysis:

In developing research questions for each of the projects, there are benefits in carrying out preliminary analysis of HES (and other NHS Digital-supplied) data before projects are formally approved. Such scoping analyses are done on an ad hoc basis where there was a need to test some basic questions that might fundamentally alter how a particular analysis is approached. In some cases, this preliminary work may even persuade the Nuffield Trust to not go ahead with a project. The questions that might be asked in scoping analyses include some of the following:

• Testing whether key outcomes of interest are numerous enough,

• Checking whether coding is consistent across organisations and geographic areas, and over time,

• Determining whether particular statistical methods would be appropriate for the questions being asked,

• Testing whether the Nuffield Trust would have the statistical power to be able to make high quality conclusions.

• Assessing the minimum level of data required for the purpose.

Any such scoping analyses would be approved by the Director or Deputy Director of Research. Approved scoping exercises are recorded in a register - recording aims of the scoping, data fields and years necessary, approval date and person, conclusions of exercise, final status of data used. Where the conclusion is to proceed with a formal research project then data used for the scoping will be transferred to and managed under that project. If the scoping exercise ends with the conclusion that no further work should be done, then the data used will be erased.

For example, the Nuffield Trust have used the data for scoping potential analysis on "lost" planned activity because of Covid-19, and what the implications are for recovery. These outputs informed current work on ethnic differences in elective care. Scoping work has also been undertaken recently on hospital care of people with learning disabilities.

Responsive analysis:

The Nuffield Trust regularly acts to improve the quality of public debate on use of hospital services by publishing quick- turnaround responsive research, which helps focus the debate on evidence. Triggers for this work might include a specific issue suddenly coming to national prominence, or an individual or organisation making an assertion which is easily tested using data already available. As an independent research organisation and registered charity, with independence from party politics overseen by the board of trustees, such interventions are carefully considered to ensure that an evidence- based statement may add value to the overall debate. They are not provided at the request of any individual organisation.

The Nuffield Trust used the data in the chart of the week series (see here - https://www.nuffieldtrust.org.uk/spotlight/chart-of-the-week), for example for tracking length of stay of covid-19 patients during the pandemic. The Nuffield Trust examined the impact on emergency admissions for respiratory conditions during the pandemic, and what the lessons are for the longer term - https://www.nuffieldtrust.org.uk/resource/chart-of-the- week-emergency-admissions-for-asthma-and-copd-during-covid-19.

A key aspect for all the research projects undertaken is ensuring learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with its charitable purposes. All outputs comprise aggregate data only, with small numbers suppressed in line with HES disclosure requirements.

A communications plan is developed for each project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to projects related to that priority. There are also web pages for individual projects.

Outputs from a project typically include (outputs will vary from project to project):

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Data visualisations, for example "Chart of the week"

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust use their extensive communications facilities and networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

For the Quality Watch programme, outputs included:

• Over 200 care quality indicators hosted on the Nuffield Trust website, organised into domains and sectors and updated monthly. More information on the use of monthly data can be found here: https://www.nuffieldtrust.org.uk/qualitywatch/indicators

Since the Nuffield Trust started to receive monthly data at the end of 2020, it has been used for QualityWatch in a number of other ways, including:

• The 2020 annual data visualisation on the impact of the rapid growth in remote care delivery during Covid- 19. This considered the impact on quality of care, and what issues need to be evaluated and monitored in the longer term. As well as releasing the data story and blog, the Nuffield Trust also presented on this issue to the Rewired digital health conference. Having monthly data enabled the analysis to be updated for the conference and reached a wider audience of stakeholders working on delivering digital services to patients.

• The Nuffield Trust also used monthly ECDS data for the 2021 annual data visualisation on the impact of Covid on children's health. The Nuffield Trust published monthly data on emergency department attendances for eating disorders in children and young people compared to adults.

• Analysis of the urgent care pathway within ICSs, and the extent to which performance across the pathway is consistent. This used monthly data in order to derive performance measures at ICS level covering the winter months.

Further, the Nuffield Trust have used monthly data in responsive analysis, such as charts of the week, and the Trust plans to continue to use monthly data to continue to produce similar outputs in future.

Links to outputs:

https://www.nuffieldtrust.org.uk/research/growing-problems-what-has-been-the-impact-of-covid-19-on-health-care-for- children-and-young-people-in-england

https://www.nuffieldtrust.org.uk/news-item/growing-problems-one-year-on-the-state-of-childrens-health-care-and-the-covid-19-backlog

https://www.nuffieldtrust.org.uk/files/2020-12/QWAS/digital-and-remote-care-in-covid-19.html#1 https://www.nuffieldtrust.org.uk/resource/chart-of-the-week-how-long-do-covid-19-patients-spend-in-hospital

https://www.nuffieldtrust.org.uk/resource/how-have-hospital-admissions-for-covid-19-and-flu-changed-in-recent-years

https://www.nuffieldtrust.org.uk/news-item/how-are-strikes-by-health-care-staff-impacting-nhs-waiting-lists

https://www.nuffieldtrust.org.uk/news-item/chain-reaction-understanding-the-causes-of-backlogs-through-urgent-and-emergency-care

Other outputs from recently completed or ongoing projects include:

• Report on Ethnicity coding in the NHS, with an accompanying blog for HSJ and press release. The findings have been presented at workshops run by the NHS Race and Health Observatory and the Cabinet Office Race Disparity Unit (RDU), and also presented to the NHS Analyst X forum. The recommendations have been adopted by the RDU. The report was widely downloaded when it was published, and the Trust had strong engagement through social media, to ensure the Trust reached a wide audience for the findings. The Trust are now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit.

• Report on integrated care across the four UK nations, with accompanying press release. Discussions are planned about the report with key government stakeholders, and the analysis was presented at the Nuffield Trust Summit in March 2022.

• Report on factors impacting obesity in child and the impact of local environmental factors, which was published in October 2022, along with a launch event for public health and other stakeholders. The report has been well received and the Trust has followed up on the findings by discussing the report with key stakeholders and networks working on children’s health. Nuffield presented the Childhood Obesity report findings to several stakeholders, including Directors of Public Health, several of which told us that they had used it as evidence in their strategies and business cases for childhood obesity. This is the first report from the project, with a follow-up report on infant mortality to be published later this year.

• Report on ethnic differences in elective care, with accompanying blog on the Nuffield Trust and NHS Race and Health Observatory website, published in November 2022. The findings have been share widely with the NHS England elective recovery team, a NHS events on elective recovery in England and Wales, and with the Analyst X forum. The report has been widely downloaded and referenced, and disseminated through a range of networks. The Trust will continue to use the findings in the policy work on elective recovery and inequalities in health care.

Impact from recently completed are summarised below.

Ethnicity coding in the health datasets:

This project examined the validity, completeness and consistency of coding in NHS datasets, and how this has changed over time. The report makes recommendations for improving the quality of coding, including calling for updated guidance, as well as recommendations for users of data, to ensure data quality is considered in analysis, and steps taken to address data issues.

In the process of the research, the Trust consulted widely with stakeholders including the National Director for Inequalities at NHS England, Public Health England, the Office of National Statistics, and NHS Digital. The Trust collaborated with NHS Race and Health Observatory (RHO)on the project. The report was widely downloaded when it was published, and actively pursued strong engagement through social media, to ensure that a wide audience for the findings was reached.

The Trust is continuing to follow up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit. Through this follow-up work, the Trust aims to ensure that the issue of poor-quality data on ethnicity is addressed. This is expected to benefit patients by leading to improved coding of data and better information on ethnic health inequalities, to ensure that service improvements can be targeted to improve care for the benefit of patients.

The Trust's recommendations have been adopted by the Race Disparity Unit in their report on addressing Covid-19 inequalities: https://www.gov.uk/government/publications/final-report-onprogress-to-address-covid-19-health- inequalities/appendix-f-prioritisation-and-progress-of-dataquality-recommendations

Growing problems: the impact of Covid-19 on healthcare for children and young people:

This Quality Watch analysis was initially undertaken in 2021 and published in February 2022. As a result of continued concerns about the impact on children’s health and care, and the lack of focus in the national elective recovery programme on children, the Trust updated the analysis in June 2023.

The results were presented to National Council for Children’s Health and Wellbeing, who on the back of it wrote a letter to Steven Barkly the Secretary of State for Health about their concerns of how much priority children were being given in national health policy.

Since the Trust published the second version of Growing Problems the study team have also generally seen an increase in people calling for a greater focus on children’s health, including 24 members of the Academy of Medical Royal Colleges signing up to say there is political deprioritisation of children and young people within the UK Government’s current health agenda https://www.rcpch.ac.uk/news-events/news/our-new-report-securing-our-healthy-futureprevention-better-cure

Medical Generalism Project:

This produced evidence on the impact of the trend towards increased specialisation in medicine on the ability of hospitals to deliver high quality of care, particularly focusing on the care delivered to patients in smaller hospitals. This is an example where analysis of HES data is uniquely able to provide evidence on the quality of care within hospitals, for which aggregate data is not suitable. The results of this study have been recently published by the National Institute for Health Research, and disseminated through events, blogs and social media to reach a wider audience. The study has important lessons for the quality of care for patients, and how this can be improved in small hospitals: these hospitals face specific pressures such as difficulty attracting staff and providing safe levels of specialist care. The Nuffield Trust’s research findings enable service providers and policy makers to ensure that the issues facing smaller hospitals are better understood, resulting in action to benefit patient care

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees with a wide range of experience including perspectives from outside the NHS as well as senior clinicians, NHS managers and academics. It aims to improve the quality of health and social care in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

The data from NHS Digital is vital to the Trust's work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. Data is primarily used in projects relating to quality and equity but has also been used for projects on models of care and legislative reform, for example integrated care policy. The Trust's work aims to serve the public interest, and for some topics, research analysis using patient data is vital.

Under the Health Research Authority's (HRA) GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject". The Nuffield Trust's legitimate interest is carrying out health and care research in the wider public interest of improved policy, services and care for NHS patients. Examples of how the Nuffield Trust process the data to pursue their legitimate interests can be found in the DPIA at https://www.nuffieldtrust.org.uk/about/corporate-policies/schedule-of-current-data-protection-documentation

The Trust relies on Article 9 Condition 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) as the condition for processing "Special" categories of personal data.

5ai. RATIONALE FOR STRATEGIC PRIORITIES AND PROJECTS

The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence applied, and the recommendations made take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable - between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled. The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised. The Trust focuses on policy areas that are relatively neglected, yet of clear significance to the health and care sector.

Current Nuffield Trust projects on quality of care and equity work are described later in this section.

The Nuffield Trust’s other strategic priorities for 2020 to 2025 are Workforce, Primary Care, Small Hospitals, Technology and Digital and Politics, Legislation and Governance.

5aii. USE OF HOSPITAL EPISODE STATISTICS (HES) DATA AND COMMUNITY SERVICES DATASET (CSDS) WITHIN PROGRAMMES AND PROJECTS

The number of concurrent projects using HES data will vary, but typically there are 5-10 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Although the methods for use of HES and CSDS will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken i.e. analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as A&E attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, such as avoidable harm, and patients with particular needs (e.g., frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient follow-up models, digital delivery of care, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis - comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, panel data and cohort studies;

o Use of the above methods in combination for particular research projects;

o Producing visualisations of analysis and results from the above methods.

Monthly HES and Emergency Care Dataset (ECDS) data:

Monthly data ensures that at any point in time the Nuffield Trust have access to the most recent data as soon as becomes available. This is beneficial as the Nuffield Trust have a number of projects in progress to investigate the impact of Covid-19 on the use of health services which has led to rapid changes in health services, and for which monthly data is required.

The data will not be linked with other record level datasets and there will be no attempt to re-identify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g., indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset. Any necessary risk assessment would be carried out as part of establishing the analysis plan during the course of the project planning process. The DPO (who is a member of the Project Planning Group) would advise on any potential issues. Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital will be made.

5aiii. DECISIONS ABOUT PROJECTS AND USE OF NHS DIGITAL DATA

This Agreement permits the Nuffield Trust to use the data for the purposes of projects through the following process:

1. Projects intended will be conceived and planned through an iterative process involving the senior researchers with appropriate input from the Data Protection Officer (DPO). A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. The project lead will ensure:

o Projects have clearly defined objectives and operational plans;

o An analysis plan is prepared for each project, setting out the data requirements and methods;

o In each case, the use of the HES is necessary and proportionate to the purpose of the project and the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

o Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Group (PPG). The PPG consists of the Chief Executive, Directors and Deputy Directors, and representatives from across the organisation including the DPO. PPG provides a forum for discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust. The project template remains valid throughout a project's life, serving as the central control document in the management and delivery of the project. The PPG is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound and

• Draw fully on the expertise within the Trust including making connections to other related work;

The PPG draws on expertise from other organisational committees to ensure that obligations are met in specific areas:

• The Data Strategy Group (responsible for Information Governance)

• The Leadership Team (responsible for Patient / Public Involvement and Equality, diversity and inclusion)

The strategic aims of the Nuffield Trust to inform decisions made to improve health and social care are set out in the following document:

https://www.nuffieldtrust.org.uk/files/2020-08/nutj7957-leaflet-200220-web-pages.pdf

After the project has been approved there will be follow up meeting with the project team and a member of the Leadership Team where details of the project are looked at in more detail including a section of questions to understand the data requirements of the project. A member of the team will be appointed as the IG lead if the size of the project suggests that this is necessary.

3. The team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales: the time frame for undertaking each project will vary according to project resource, extent of the research and data analysis required. This is always considered as part of the Project Planning Group's review and approved based upon the detail of each individual project.

Project Funding: some projects are funded by the Nuffield Trust, but some receive funding from other partners. At the current time, the Trust are receiving funding for projects which use HES data from the National Institute for Health Research (NIHR), the NHS Race and Health Observatory (RHO) and the Health Foundation. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose, as outlined below.

Data Governance: the Nuffield Trust has independently determined the purposes for which it requires and will process data under this Agreement in terms of its priorities and programmes. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the sole data controller named in this Agreement, The Trust cannot be compelled by any third party to process the data for any purpose in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

The Trust do not undertake events for commercial purposes and the data requested as part of this application is not being requested for commercial purposes.

Public Participation: The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence the Trust applies, and the recommendations made, take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust's secure, ISO27001 certified data environment. Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital will be made.

Results may be shared in aggregate form to meet the objectives of the project, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust and will not be shared with any other third parties. The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities. Details will also be available on the Nuffield Trust's website.

Current projects using HES, ECDS or CSDS data include:

• QualityWatch, a programme jointly funded by the Health Foundation which provides independent scrutiny of the quality of care in the NHS and social care, in order to highlight to policy makers and the public about services where quality of care is at risk.

• Prisoner health, a programme funded by the Health Foundation which in this phase will compare the effect of being in prison on use of hospital care, by comparison with a matched case control group (this project is also covered by DARS agreement NIC-195377-M9L8Z). This will inform policy for prisoner health.

• Evaluation of integration pioneers, a long running evaluation funded by NIHR for which the Nuffield Trust is undertaking quantitative analysis as part of a wider project with the London School of Hygiene and Tropical Medicine. The analysis of HES to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data nor does LSHTM determine the purposes or means of processing data under this agreement.

• Rapid Service Evaluation Team funded by NIHR and jointly delivered with University College London (see separate DARS agreement NIC-194629-S4F9X).

• Understanding the impact of local area factors on child health outcomes, in order to identify actions which could be taken at a local authority level, which is an internally funded project using HES data to derive local area measures of activity.

• Provision of Hospice Activity, a project jointly funded with Hospice UK, for which HES and CSDS data is used to provide an additional source of data on activity and assess data quality for the hospice sector and end of life care.

• Ethnic differences in lost activity, treatment rates and times to treatment, a project funded by the NHS Race and Health Observatory (NHS RHO) to understand whether ethnic inequalities exist and inform NHS recovery plans. The NHS RHO does not have access to the data nor does NHS RHO determine the purpose or means of processing data under this agreement.

Further details of projects are included in the outputs and benefits sections.

Project Scoping and Responsive Analysis

Across the Nuffield Trust's strategic priorities, analysis may be undertaken of HES for scoping research and responsive analysis.

Scoping analysis:

In developing research questions for each of the projects, there are benefits in carrying out preliminary analysis of HES (and other NHS Digital-supplied) data before projects are formally approved. Such scoping analyses are done on an ad hoc basis where there was a need to test some basic questions that might fundamentally alter how a particular analysis is approached. In some cases, this preliminary work may even persuade the Nuffield Trust to not go ahead with a project. The questions that might be asked in scoping analyses include some of the following:

• Testing whether key outcomes of interest are numerous enough,

• Checking whether coding is consistent across organisations and geographic areas, and over time,

• Determining whether particular statistical methods would be appropriate for the questions being asked,

• Testing whether the Nuffield Trust would have the statistical power to be able to make high quality conclusions.

• Assessing the minimum level of data required for the purpose.

Any such scoping analyses would be approved by the Director or Deputy Director of Research. Approved scoping exercises are recorded in a register - recording aims of the scoping, data fields and years necessary, approval date and person, conclusions of exercise, final status of data used. Where the conclusion is to proceed with a formal research project then data used for the scoping will be transferred to and managed under that project. If the scoping exercise ends with the conclusion that no further work should be done, then the data used will be erased.

For example, the Nuffield Trust have used the data for scoping potential analysis on "lost" planned activity because of Covid-19, and what the implications are for recovery. These outputs informed current work on ethnic differences in elective care. Scoping work has also been undertaken recently on hospital care of people with learning disabilities.

Responsive analysis:

The Nuffield Trust regularly acts to improve the quality of public debate on use of hospital services by publishing quick- turnaround responsive research, which helps focus the debate on evidence. Triggers for this work might include a specific issue suddenly coming to national prominence, or an individual or organisation making an assertion which is easily tested using data already available. As an independent research organisation and registered charity, with independence from party politics overseen by the board of trustees, such interventions are carefully considered to ensure that an evidence- based statement may add value to the overall debate. They are not provided at the request of any individual organisation.

The Nuffield Trust used the data in the chart of the week series (see here - https://www.nuffieldtrust.org.uk/spotlight/chart-of-the-week), for example for tracking length of stay of covid-19 patients during the pandemic. The Nuffield Trust examined the impact on emergency admissions for respiratory conditions during the pandemic, and what the lessons are for the longer term - https://www.nuffieldtrust.org.uk/resource/chart-of-the- week-emergency-admissions-for-asthma-and-copd-during-covid-19.

A key aspect for all the research projects undertaken is ensuring learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with its charitable purposes. All outputs comprise aggregate data only, with small numbers suppressed in line with HES disclosure requirements.

A communications plan is developed for each project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to projects related to that priority. There are also web pages for individual projects.

Outputs from a project typically include (outputs will vary from project to project):

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Data visualisations, for example "Chart of the week"

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust use their extensive communications facilities and networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

For the Quality Watch programme, outputs included:

• Over 200 care quality indicators hosted on the Nuffield Trust website, organised into domains and sectors and updated monthly. More information on the use of monthly data can be found here: https://www.nuffieldtrust.org.uk/qualitywatch/indicators

Since the Nuffield Trust started to receive monthly data at the end of 2020, it has been used for QualityWatch in a number of other ways, including:

• The 2020 annual data visualisation on the impact of the rapid growth in remote care delivery during Covid-19. This considered the impact on quality of care, and what issues need to be evaluated and monitored in the longer term. As well as releasing the data story and blog, the Nuffield Trust also presented on this issue to the Rewired digital health conference. Having monthly data enabled the analysis to be updated for the conference and reached a wider audience of stakeholders working on delivering digital services to patients.

• The Nuffield Trust also used monthly ECDS data for the 2021 annual data visualisation on the impact of Covid on children’s health. The Nuffield Trust published monthly data on emergency department attendances for eating disorders in children and young people compared to adults.

The Nuffield Trust are using monthly data to update analysis for the ongoing work on ethnicity and elective treatment (project in progress).

Further, the Nuffield Trust have used monthly data in responsive analysis, such as charts of the week, and the Trust plans to continue to use monthly data to continue to produce similar outputs in future.

Links to outputs:

https://www.nuffieldtrust.org.uk/research/growing-problems-what-has-been-the-impact-of-covid-19-on-health-care-for-children-and-young-people-in-england

https://www.nuffieldtrust.org.uk/research/growing-problems-what-has-been-the-impact-of-covid-19-on-health-care-for-children-and-young-people-in-england

https://www.nuffieldtrust.org.uk/files/2020-12/QWAS/digital-and-remote-care-in-covid-19.html#1

https://www.nuffieldtrust.org.uk/resource/chart-of-the-week-how-long-do-covid-19-patients-spend-in-hospital

Other outputs from recently completed or ongoing projects include:

• Report on Ethnicity coding in the NHS, with an accompanying blog for HSJ and press release. The findings have been presented at workshops run by the NHS Race and Health Observatory and the Cabinet Office Race Disparity Unit (RDU), and also presented to the NHS Analyst X forum. The recommendations have been adopted by the RDU. The report was widely downloaded when it was published, and the Trust had strong engagement through social media, to ensure the Trust reached a wide audience for the findings. The Trust are now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit.

• Report on integrated care across the four UK nations, with accompanying press release. Discussions are planned about the report with key government stakeholders, and the analysis was presented at the Nuffield Trust Summit in March 2022.

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been published by NIHR and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. The findings have been presented at conferences, including a presentation on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

The following main outputs are planned for other current projects:

• Reports on factors impacting child health outcomes (June 2022)

• Report on ethnicity differences in elective care (August 2022)

Impact from recently completed projects during 2021/2022 is summarised below.

Ethnicity coding in the health datasets:

This project examined the validity, completeness and consistency of coding in NHS datasets, and how this has changed over time. The report makes recommendations for improving the quality of coding, including calling for updated guidance, as well as recommendations for users of data, to ensure data quality is considered in analysis, and steps taken to address data issues. In the process of the research, the Trust consulted widely with stakeholders including the National Director for Inequalities at NHS England, Public Health England, the Office of National Statistics, and NHS Digital. The Trust collaborated with NHS Race and Health Observatory (RHO) on the project. The report was widely downloaded when it was published, and actively pursued strong engagement through social media, to ensure that a wide audience for the findings was reached. The Trust is now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit. Through this follow-up work, the Trust aims to ensure that the issue of poor-quality data on ethnicity is addressed. This is expected to benefit patients by leading to improved coding of data and better information on ethnic health inequalities, to ensure that service improvements can be targeted to improve care for the benefit of patients. The Trust's recommendations have been adopted by the Race Disparity Unit in their report on addressing Covid-19 inequalities: https://www.gov.uk/government/publications/final-report-on-progress-to-address-covid-19-health-inequalities/appendix-f-prioritisation-and-progress-of-data-quality-recommendations

Medical Generalism Project:

This produced evidence on the impact of the trend towards increased specialisation in medicine on the ability of hospitals to deliver high quality of care, particularly focusing on the care delivered to patients in smaller hospitals. This is an example where analysis of HES data is uniquely able to provide evidence on the quality of care within hospitals, for which aggregate data is not suitable. The results of this study have been recently published by the National Institute for Health Research, and disseminated through events, blogs and social media to reach a wider audience. The study has important lessons for the quality of care for patients, and how this can be improved in small hospitals: these hospitals face specific pressures such as difficulty attracting staff and providing safe levels of specialist care. The Nuffield Trust’s research findings enable service providers and policy makers to ensure that the issues facing smaller hospitals are better understood, resulting in action to benefit patient care

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees with a wide range of experience including perspectives from outside the NHS as well as senior clinicians, NHS managers and academics. It aims to improve the quality of health and social care in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

The data from NHS Digital is vital to the Trust's work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. Data is primarily used in projects relating to quality and equity but has also been used for projects on models of care and legislative reform, for example integrated care policy. The Trust's work aims to serve the public interest, and for some topics, research analysis using patient data is vital.

Under the Health Research Authority's (HRA) GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject". The Nuffield Trust's legitimate interest is carrying out health and care research in the wider public interest of improved policy, services and care for NHS patients. Examples of how the Nuffield Trust process the data to pursue their legitimate interests can be found in the DPIA at https://www.nuffieldtrust.org.uk/about/corporate-policies/schedule-of-current-data-protection-documentation

The Trust relies on Article 9 Condition 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) as the condition for processing "Special" categories of personal data.

5ai. RATIONALE FOR STRATEGIC PRIORITIES AND PROJECTS

The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence applied, and the recommendations made take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable - between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled. The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised. The Trust focuses on policy areas that are relatively neglected, yet of clear significance to the health and care sector.

Current Nuffield Trust projects on quality of care and equity work are described later in this section.

The Nuffield Trust’s other strategic priorities for 2020 to 2025 are Workforce, Primary Care, Small Hospitals, Technology and Digital and Politics, Legislation and Governance.

5aii. USE OF HOSPITAL EPISODE STATISTICS (HES) DATA AND COMMUNITY SERVICES DATASET (CSDS) WITHIN PROGRAMMES AND PROJECTS

The number of concurrent projects using HES data will vary, but typically there are 5-10 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Although the methods for use of HES and CSDS will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken i.e. analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as A&E attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, such as avoidable harm, and patients with particular needs (e.g., frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient follow-up models, digital delivery of care, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis - comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, panel data and cohort studies;

o Use of the above methods in combination for particular research projects;

o Producing visualisations of analysis and results from the above methods.

Monthly HES and Emergency Care Dataset (ECDS) data:

Monthly data ensures that at any point in time the Nuffield Trust have access to the most recent data as soon as becomes available. This is beneficial as the Nuffield Trust have a number of projects in progress to investigate the impact of Covid-19 on the use of health services which has led to rapid changes in health services, and for which monthly data is required.

The data will not be linked with other record level datasets and there will be no attempt to re-identify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g., indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset. Any necessary risk assessment would be carried out as part of establishing the analysis plan during the course of the project planning process. The DPO (who is a member of the Project Planning Group) would advise on any potential issues. Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital will be made.

5aiii. DECISIONS ABOUT PROJECTS AND USE OF NHS DIGITAL DATA

This Agreement permits the Nuffield Trust to use the data for the purposes of projects through the following process:

1. Projects intended will be conceived and planned through an iterative process involving the senior researchers with appropriate input from the Data Protection Officer (DPO). A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. The project lead will ensure:

o Projects have clearly defined objectives and operational plans;

o An analysis plan is prepared for each project, setting out the data requirements and methods;

o In each case, the use of the HES is necessary and proportionate to the purpose of the project and the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

o Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Group (PPG). The PPG consists of the Chief Executive, Directors and Deputy Directors, and representatives from across the organisation including the DPO. PPG provides a forum for discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust. The project template remains valid throughout a project's life, serving as the central control document in the management and delivery of the project. The PPG is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound and

• Draw fully on the expertise within the Trust including making connections to other related work;

The PPG draws on expertise from other organisational committees to ensure that obligations are met in specific areas:

• The Data Strategy Group (responsible for Information Governance)

• The Leadership Team (responsible for Patient / Public Involvement and Equality, diversity and inclusion)

The strategic aims of the Nuffield Trust to inform decisions made to improve health and social care are set out in the following document:

https://www.nuffieldtrust.org.uk/files/2020-08/nutj7957-leaflet-200220-web-pages.pdf

After the project has been approved there will be follow up meeting with the project team and a member of the Leadership Team where details of the project are looked at in more detail including a section of questions to understand the data requirements of the project. A member of the team will be appointed as the IG lead if the size of the project suggests that this is necessary.

3. The team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales: the time frame for undertaking each project will vary according to project resource, extent of the research and data analysis required. This is always considered as part of the Project Planning Group's review and approved based upon the detail of each individual project.

Project Funding: some projects are funded by the Nuffield Trust, but some receive funding from other partners. At the current time, the Trust are receiving funding for projects which use HES data from the National Institute for Health Research (NIHR), the NHS Race and Health Observatory (RHO) and the Health Foundation. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose, as outlined below.

Data Governance: the Nuffield Trust has independently determined the purposes for which it requires and will process data under this Agreement in terms of its priorities and programmes. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the sole data controller named in this Agreement, The Trust cannot be compelled by any third party to process the data for any purpose in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

The Trust do not undertake events for commercial purposes and the data requested as part of this application is not being requested for commercial purposes.

Public Participation: The Nuffield Trust draws on patient and public stakeholders at an individual project level, ensuring that the research questions, the evidence the Trust applies, and the recommendations made, take proper account of the experience and needs of the people who use services. At a strategic level, the Trust are establishing a partnership with two or more organisations which represent the voice of patients and the public, to provided challenge and support as the Trust set their medium to long term work programme, ensuring that they are considering issues which are truly relevant to those who use services.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust's secure, ISO27001 certified data environment. Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital will be made.

Results may be shared in aggregate form to meet the objectives of the project, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust and will not be shared with any other third parties. The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities. Details will also be available on the Nuffield Trust's website.

Current projects using HES, ECDS or CSDS data include:

• QualityWatch, a programme jointly funded by the Health Foundation which provides independent scrutiny of the quality of care in the NHS and social care, in order to highlight to policy makers and the public about services where quality of care is at risk.

• Prisoner health, a programme funded by the Health Foundation which in this phase will compare the effect of being in prison on use of hospital care, by comparison with a matched case control group (this project is also covered by DARS agreement NIC-195377-M9L8Z). This will inform policy for prisoner health.

• Evaluation of integration pioneers, a long running evaluation funded by NIHR for which the Nuffield Trust is undertaking quantitative analysis as part of a wider project with the London School of Hygiene and Tropical Medicine. The analysis of HES to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data nor does LSHTM determine the purposes or means of processing data under this agreement.

• Rapid Service Evaluation Team funded by NIHR and jointly delivered with University College London (see separate DARS agreement NIC-194629-S4F9X).

• Understanding the impact of local area factors on child health outcomes, in order to identify actions which could be taken at a local authority level, which is an internally funded project using HES data to derive local area measures of activity.

• Provision of Hospice Activity, a project jointly funded with Hospice UK, for which HES and CSDS data is used to provide an additional source of data on activity and assess data quality for the hospice sector and end of life care.

• Ethnic differences in lost activity, treatment rates and times to treatment, a project funded by the NHS Race and Health Observatory (NHS RHO) to understand whether ethnic inequalities exist and inform NHS recovery plans. The NHS RHO does not have access to the data nor does NHS RHO determine the purpose or means of processing data under this agreement.

Further details of projects are included in the outputs and benefits sections.

Project Scoping and Responsive Analysis

Across the Nuffield Trust's strategic priorities, analysis may be undertaken of HES for scoping research and responsive analysis.

Scoping analysis:

In developing research questions for each of the projects, there are benefits in carrying out preliminary analysis of HES (and other NHS Digital-supplied) data before projects are formally approved. Such scoping analyses are done on an ad hoc basis where there was a need to test some basic questions that might fundamentally alter how a particular analysis is approached. In some cases, this preliminary work may even persuade the Nuffield Trust to not go ahead with a project. The questions that might be asked in scoping analyses include some of the following:

• Testing whether key outcomes of interest are numerous enough,

• Checking whether coding is consistent across organisations and geographic areas, and over time,

• Determining whether particular statistical methods would be appropriate for the questions being asked,

• Testing whether the Nuffield Trust would have the statistical power to be able to make high quality conclusions.

• Assessing the minimum level of data required for the purpose.

Any such scoping analyses would be approved by the Director or Deputy Director of Research. Approved scoping exercises are recorded in a register - recording aims of the scoping, data fields and years necessary, approval date and person, conclusions of exercise, final status of data used. Where the conclusion is to proceed with a formal research project then data used for the scoping will be transferred to and managed under that project. If the scoping exercise ends with the conclusion that no further work should be done, then the data used will be erased.

For example, the Nuffield Trust have used the data for scoping potential analysis on "lost" planned activity because of Covid-19, and what the implications are for recovery. These outputs informed current work on ethnic differences in elective care. Scoping work has also been undertaken recently on hospital care of people with learning disabilities.

Responsive analysis:

The Nuffield Trust regularly acts to improve the quality of public debate on use of hospital services by publishing quick- turnaround responsive research, which helps focus the debate on evidence. Triggers for this work might include a specific issue suddenly coming to national prominence, or an individual or organisation making an assertion which is easily tested using data already available. As an independent research organisation and registered charity, with independence from party politics overseen by the board of trustees, such interventions are carefully considered to ensure that an evidence- based statement may add value to the overall debate. They are not provided at the request of any individual organisation.

The Nuffield Trust used the data in the chart of the week series (see here - https://www.nuffieldtrust.org.uk/spotlight/chart-of-the-week), for example for tracking length of stay of covid-19 patients during the pandemic. The Nuffield Trust examined the impact on emergency admissions for respiratory conditions during the pandemic, and what the lessons are for the longer term - https://www.nuffieldtrust.org.uk/resource/chart-of-the- week-emergency-admissions-for-asthma-and-copd-during-covid-19.

A key aspect for all the research projects undertaken is ensuring learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with its charitable purposes. All outputs comprise aggregate data only, with small numbers suppressed in line with HES disclosure requirements.

A communications plan is developed for each project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to projects related to that priority. There are also web pages for individual projects.

Outputs from a project typically include (outputs will vary from project to project):

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Data visualisations, for example "Chart of the week"

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust use their extensive communications facilities and networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

For the Quality Watch programme, outputs included:

• Over 200 care quality indicators hosted on the Nuffield Trust website, organised into domains and sectors and updated monthly. More information on the use of monthly data can be found here: https://www.nuffieldtrust.org.uk/qualitywatch/indicators

Since the Nuffield Trust started to receive monthly data at the end of 2020, it has been used for QualityWatch in a number of other ways, including:

• The 2020 annual data visualisation on the impact of the rapid growth in remote care delivery during Covid-19. This considered the impact on quality of care, and what issues need to be evaluated and monitored in the longer term. As well as releasing the data story and blog, the Nuffield Trust also presented on this issue to the Rewired digital health conference. Having monthly data enabled the analysis to be updated for the conference and reached a wider audience of stakeholders working on delivering digital services to patients.

• The Nuffield Trust also used monthly ECDS data for the 2021 annual data visualisation on the impact of Covid on children’s health. The Nuffield Trust published monthly data on emergency department attendances for eating disorders in children and young people compared to adults.

The Nuffield Trust are using monthly data to update analysis for the ongoing work on ethnicity and elective treatment (project in progress).

Further, the Nuffield Trust have used monthly data in responsive analysis, such as charts of the week, and the Trust plans to continue to use monthly data to continue to produce similar outputs in future.

Links to outputs:

https://www.nuffieldtrust.org.uk/research/growing-problems-what-has-been-the-impact-of-covid-19-on-health-care-for-children-and-young-people-in-england

https://www.nuffieldtrust.org.uk/research/growing-problems-what-has-been-the-impact-of-covid-19-on-health-care-for-children-and-young-people-in-england

https://www.nuffieldtrust.org.uk/files/2020-12/QWAS/digital-and-remote-care-in-covid-19.html#1

https://www.nuffieldtrust.org.uk/resource/chart-of-the-week-how-long-do-covid-19-patients-spend-in-hospital

Other outputs from recently completed or ongoing projects include:

• Report on Ethnicity coding in the NHS, with an accompanying blog for HSJ and press release. The findings have been presented at workshops run by the NHS Race and Health Observatory and the Cabinet Office Race Disparity Unit (RDU), and also presented to the NHS Analyst X forum. The recommendations have been adopted by the RDU. The report was widely downloaded when it was published, and the Trust had strong engagement through social media, to ensure the Trust reached a wide audience for the findings. The Trust are now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit.

• Report on integrated care across the four UK nations, with accompanying press release. Discussions are planned about the report with key government stakeholders, and the analysis was presented at the Nuffield Trust Summit in March 2022.

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been published by NIHR and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. The findings have been presented at conferences, including a presentation on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

The following main outputs are planned for other current projects:

• Reports on factors impacting child health outcomes (June 2022)

• Report on ethnicity differences in elective care (August 2022)

Impact from recently completed projects during 2021/2022 is summarised below.

Ethnicity coding in the health datasets:

This project examined the validity, completeness and consistency of coding in NHS datasets, and how this has changed over time. The report makes recommendations for improving the quality of coding, including calling for updated guidance, as well as recommendations for users of data, to ensure data quality is considered in analysis, and steps taken to address data issues. In the process of the research, the Trust consulted widely with stakeholders including the National Director for Inequalities at NHS England, Public Health England, the Office of National Statistics, and NHS Digital. The Trust collaborated with NHS Race and Health Observatory (RHO) on the project. The report was widely downloaded when it was published, and actively pursued strong engagement through social media, to ensure that a wide audience for the findings was reached. The Trust is now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop held by the Cabinet Office Race Disparity Unit. Through this follow-up work, the Trust aims to ensure that the issue of poor-quality data on ethnicity is addressed. This is expected to benefit patients by leading to improved coding of data and better information on ethnic health inequalities, to ensure that service improvements can be targeted to improve care for the benefit of patients. The Trust's recommendations have been adopted by the Race Disparity Unit in their report on addressing Covid-19 inequalities: https://www.gov.uk/government/publications/final-report-on-progress-to-address-covid-19-health-inequalities/appendix-f-prioritisation-and-progress-of-data-quality-recommendations

Medical Generalism Project:

This produced evidence on the impact of the trend towards increased specialisation in medicine on the ability of hospitals to deliver high quality of care, particularly focusing on the care delivered to patients in smaller hospitals. This is an example where analysis of HES data is uniquely able to provide evidence on the quality of care within hospitals, for which aggregate data is not suitable. The results of this study have been recently published by the National Institute for Health Research, and disseminated through events, blogs and social media to reach a wider audience. The study has important lessons for the quality of care for patients, and how this can be improved in small hospitals: these hospitals face specific pressures such as difficulty attracting staff and providing safe levels of specialist care. The Nuffield Trust’s research findings enable service providers and policy makers to ensure that the issues facing smaller hospitals are better understood, resulting in action to benefit patient care

June 2021 - Monthly data dissemination are requested to gain access to more timely data (especially related to COVID-19 research) from 1/7/21 to 30/6/22. Justification:

Since the Nuffield Trust received monthly data at the end of 2020, the Trust have used this for Quality Watch annual data visualisation on the impact of the rapid growth in remote care delivery during Covid-19. This considered the long-term impact on quality of care, and what issues need to be evaluated and monitored in the longer term. As well as releasing the data story and blog, the Nuffield Trust also presented on this issue to the Rewired digital health conference. Having monthly data enabled updates to the analysis for the conference.

The Trust do not undertake events for commercial purposes and the data requested as part of this application is not being requested for commercial purposes.

The Nuffield Trust used the data in the chart of the week series (see here - https://www.nuffieldtrust.org.uk/spotlight/chart-of-the-week), for example for tracking length of stay of covid-19 patients during the pandemic. The Nuffield Trust examined the impact on emergency admissions for respiratory conditions during the pandemic, and what the lessons are for the longer term - https://www.nuffieldtrust.org.uk/resource/chart-of-the-week-emergency-admissions-for-asthma-and-copd-during-covid-19. The monthly data has enabled the Nuffield Trust to track cancelled operations for Quality Watch, data which contributes to the analysis of how the NHS is managing during the pandemic, and what the impacts are on health care for non-Covid conditions.

The Nuffield Trust have used the data for scoping potential analysis on “lost” planned activity because of Covid-19, and what the implications are for recovery. These outputs fall within the Quality and equity programme, and are relevant to Politics, legislation, and governance work, where the Nuffield Trust consider issues which are a priority for politicians and government.

The need for monthly data also supports RSET projects led by the Nuffield Trust. See here - https://www.nuffieldtrust.org.uk/rset-the-rapid-service-evaluation-team.

The Nuffield Trust have a number of projects in progress (covered under existing data sharing agreements) to investigate the impact of Covid-19 on the use of health services which are dependent on or would benefit from access to more timely (i.e. monthly rather than quarterly) HES data. These include, for example, an analysis of outpatient attendances with a view to identify variations in activity by trust and specialty over the early pandemic period as part of evaluation work for the NIHR-funded Rapid Service Evaluation Team. The Trust has started a project to understand the profile of patients discharged before and during the pandemic, and their subsequent use of services in the community.

____________________________________________________________________________________________________

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees including a number of senior NHS clinicians, managers and academics. The Nuffield Trust aims to improve the quality of health care to improve the health of people in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

Under the HRA’s GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) “processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject”. The Nuffield Trust’s legitimate interest is carrying out healthcare research in the wider public interest of improved healthcare outcomes for NHS patients.

The Trust relies on Article 9 Condition 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) as the condition for processing “Special” categories of personal data.

All data supplied is pseudonymised. All outputs contain only aggregate data, with small numbers suppressed to agreed thresholds in line with the HES (or appropriate) analysis guide.

The focus of the Nuffield Trust activities for 2020-2025 is on six priority areas:

1. Workforce,

2. Technology and digital,

3. Primary care,

4. Small hospitals,

5. Quality and equity,

6. Politics, legislation, and governance.

The quality and equity area includes care for specific population groups, for example children and young people, as well as care for underserved populations, for example prisoners, and inequalities in health and care service delivery.

The work of the Trust is organised into a number of programmes which address these priorities. The programmes are broadly aligned to priorities, with a desire to develop a number of the programmes to cut across one or more of the strategic priorities to maximise reach and impact. For example, the Quality programme is relevant to both Quality of Care and Primary Care priorities, and the Workforce programme impacts on both Workforce and New Models of Care. Each programme has a sponsor at Director level, as well as a programme lead, who is typically a Senior Researcher or Senior Policy Fellow. The data from NHS Digital is vital to the Trust’s work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. The use within programmes and projects is outlined in more detail in the following sections.

____________________________________________________________________________________________________

5a.i. RATIONALE FOR STRATEGIC PRIORITIES AND PROGRAMMES

Politics, legislation and governance:

The NHS is introducing new models of care and a different way to work with councils through Integrated Care Systems. All of this is being attempted at a time of historic financial constraint, with record trust deficits and an intense search for efficiencies. In addition, Britain's departure from the EU could mean major changes and deep uncertainty for health and social care staffing, regulation and workforce. The Nuffield Trust’s focus is on improving the quality of policy-making by providing evidence-based analysis, asking insightful questions and providing a challenging view. Programmes in this area cover:

• Integrated care systems and delivery – how the health system is changing, for example, moving to integrated care.

• Topical Issues – for example, issues such as Brexit, which impact on the health system.

• Funding and Sustainability – impact of spending constraints and how funding is allocated to health and care organisations

Workforce:

The NHS workforce is under extreme pressure, and it is not just a matter of numbers. Some areas are undersubscribed, but others have an over-supply. New technologies and care models require new types of staff, and training for this highly skilled workforce has long lead times. Getting all this right has never been more critical, especially following Covid-19 and with the uncertainty around migration following the country’s departure from the EU. The development of new models of care is often driven by changes in workforce and vice versa. The workforce programme addresses how the NHS manages workforce pressures and develops a sustainable workforce. The Nuffield Trust ensures their research and analysis informs the Government’s strategy for the future health care workforce.

Quality and equity:

Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable – between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled. The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised. Quality of care programmes cover:

• Quality – Drawing on its other work programmes, the Nuffield Trust uses expertise in measurement and analysis of quality of care, to provide independent scrutiny, and undertakes research to improve the evidence on quality of care.

• Evaluation – The Nuffield Trust also considers how effective policies intended to improve quality have been, and what can be learnt, in order to influence future decision makers, locally and nationally. This includes evaluating service changes and innovations in the delivery of care.

• Children and young people – including analysis of changes in service use, and the wider influences on health of children and young people.

• Prisoner health – to understand hospital use by prisoners.

• Learning disabilities – this work is currently being scoped.

Models of health care delivery (including primary care, digital and technology and smaller hospitals):

NHS England’s Long Term Plan, and previously the Five Year Forward View, outlines a vision for how the traditional boundaries between primary care, community services and hospitals will be dissolved over the next five years and beyond. Breaking down the traditional boundaries between different parts of the NHS and social care offers the prospect of reshaping services around the needs of individuals and reducing reliance on hospitals. This will require a move away from single institutions towards networks of care. In this world the capacity of care delivered outside hospitals will need to be boosted through reforms to general practice, while technology and new types of staff enable expertise and information to be shared. The Nuffield Trust will build on the considerable experience and reputation it has in conducting evaluative work of new models of care, with a number of research projects in place with Royal colleges and specialist societies. The Trust will provide briefings and analysis that help health leaders choose and implement changes, and bring them together to share ideas. Programmes in this area cover:

• Acute Medical Models – models of hospital care and the optimal configuration of services to deliver inpatient services.

• Primary Care – future of primary care and role of primary care in a changing health system.

• Digital – the impact of new technology on delivery of health care, and opportunities to improve care using technology.

The work programmes within the strategic priorities were developed in 2020 and are expected to remain aligned to the strategic priorities above until at least 2025 when they may be reviewed.

____________________________________________________________________________________________________

5a.ii. USE OF HES AND CSDS DATA WITHIN PROGRAMMES AND PROJECTS

Each programme is delivered through individual projects. Projects vary in length and complexity from major research studies which could span a number of years, to shorter term projects resulting in a single output, for example a briefing or presentation. Some Nuffield Trust projects are qualitative, involve analysis of published data, or data from surveys, and so do not require use of HES or other NHS Digital data sources. However, many others are reliant on HES and other NHS Digital-supplied data sets.

To undertake such projects the Nuffield Trust has determined that it requires HES and the CSDS which is now available and is a very welcome addition to the existing hospital based HES datasets. Although the methods for use of HES and other patient datasets will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken i.e. analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as A&E attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, for example as part of the harm project (see below), and patients with particular needs (eg frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient advice models, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis – comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, including interrupted time series, panel data and cohort studies;

• Use of the above methods in combination for particular research projects;

• Producing visualisations of analysis and results from the above methods.

The data will not be linked with other record level datasets and there will be no attempt to reidentify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g. indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset. Any necessary risk assessment would be carried out as part of establishing the analysis plan during the course of the project planning process. The DPO (who is a member of the Project Planning Group) would advise on any potential issues. Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital will be made.

The number of concurrent projects using HES data will vary, but typically there are 5-6 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings. Projects are grouped within programmes, but frequently cut across other programmes. As the Nuffield Trust is a small organisation, it seeks to maximise impact by undertaking work which can inform more than one programme or strategic priority. Undertaking projects which support multiple programmes enables the Trust to deliver greater benefits to the health and care system.

_______________________________________________________________________________________

5a.iii. DECISIONS ABOUT PROJECTS AND USE OF NHS DIGITAL DATA

This Agreement permits the Nuffield Trust to use the data for the purposes of projects through the following process:

1. Projects intended to meet the programme’s aims will be conceived and planned through an iterative process involving the Programme Director and Programme Lead with appropriate input from the Data Protection Officer (DPO). A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. A Legitimate Interest Assessment (LIA) will be completed internally for each research project and signed off by the DPO. The Programme Director and Programme Lead will ensure:

• Projects have a clearly defined objectives and operational plans;

• The aims of projects align with at least one of the programme’s aims;

• An analysis plan is prepared for each project, setting out the data requirements and methods;

• In each case, the use of the HES is necessary and proportionate to the purpose of the project and the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

• Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Committee (PPC). The PPC approve or recommend the approval of projects in line with the internally approved schedule of management authority and responsibility. The PPC is chaired by the Director of Communications and consists of the Chief Executive, Director of Research, Director of Policy, Senior Fellow, Senior Policy Analyst and other representatives from Research, Policy and Communications, the PPC provides a forum for the discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust. This excel template serves a wider purpose than just planning and remains valid throughout a project’s life, serving as the central control document in the management and delivery of the project. The PPC is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound; and

• Draw fully on the expertise within the Trust including making connections to other related work.

3. The team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales

The time frame for undertaking each project will vary according to project resource, extent of the research and data analysis required. This is always considered as part of the Project Planning Committee's review and approved based upon the detail of each individual project.

Project Funding

Some projects under the programme will be funded by the Nuffield Trust but funding is also being sought from other partners including National Research and Charitable organisations, such as the Nuffield Foundation, NIHR and the Health Foundation but may not be limited to these organisations. Funders will take both the forms of partners in collaborative working, as well as commissioners only, that is, the Nuffield Trust will be sole data controller for the analytical work. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose. The Nuffield Trust will not be reliant on securing funding from external partners to complete this research. However, the Nuffield Trust will need to recognise the contribution of any external partner in their outputs.

Data Governance

The Nuffield Trust has independently determined the purposes for which it requires and will process data under this Agreement in terms of its priorities and programmes. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the primary data controller and the sole data controller named in this Agreement, The Trust cannot be compelled by any third party to process the data for any purpose in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust’s secure, ISO27001 certified data environment. Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital will be made.

Results may be shared in aggregate form in accordance with the Nuffield Trust's Research Governance framework, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust, and will not be shared with any other third-parties. The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities and programme(s). Details will also be available on the Nuffield Trust’s website.

____________________________________________________________________________________________________

Project Scoping and Responsive Analysis

Across many of the Nuffield Trust’s strategic priorities, analysis may be undertaken of HES for scoping research and responsive analysis. Scoping analysis:

In developing research questions for each of the projects, there would be benefits in occasionally carrying out a preliminary use of HES (and other NHS Digital-supplied) data before projects are formally approved. Such scoping analyses would be done on an ad hoc basis where there was a need to test some basic questions that might fundamentally alter how a particular analysis is approached. In some cases this preliminary work may even persuade the Nuffield Trust to not go ahead with a project. The questions that might be asked in scoping analyses might include some of the following:

• Testing whether key outcomes of interest are numerous enough,

• Checking whether coding is consistent across organisations and geographic areas, and over time,

• Determining whether particular statistical methods would be appropriate for the questions being asked,

• Testing whether the Nuffield Trust would have the statistical power to be able to make high quality conclusions.

• Assessing the minimum level of data required for the purpose.

Any such scoping analyses would be approved by the Director or Deputy Director of Research. Approved scoping exercises would be recorded in a register – recording aims of the scoping, data fields and years necessary, approval date and person, conclusions of exercise, final status of data used. Where the conclusion is to proceed with a formal research project then data used for the scoping will be transferred to and managed under that project. If the scoping exercise ends with the conclusion that no further work should be done, then the data used will be erased.

Responsive analysis:

The Nuffield Trust regularly acts to improve the quality of public debate on use of hospital services by publishing quick-turnaround responsive research, which helps focus the debate on evidence. Triggers for this work might include a specific issue suddenly coming to national prominence, or an individual or organisation making an assertion which is easily tested using data already available. As an independent research organisation and registered charity, with independence from party politics overseen by the board of trustees, such interventions are carefully considered to ensure that an evidence-based statement may add value to the overall debate. They are not provided at the request of any individual organisation.

A key aspect for all the research projects undertaken is ensuring learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with their charitable purposes.

A communications plan is developed for each programme and project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to programmes and projects related to that priority.

Outputs from a project typically include (outputs will vary from project to project):

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Data visualisations, for example “Chart of the week”

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust will use their extensive communications facilities and networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

In a typical year outputs for the Nuffield Trust as a whole include:

• 19 reports and briefings

• 66 blogs and explainers and 19 charts of the week

• 865,222 visitors to our website

• 53.7 K followers on Twitter, and increase of 4.1K on the previous year

• 7,872 unique mentions across print and broadcast media

• 17 mentions in parliament, 9 written submissions and 5 oral evidence sessions

• 71 speaking engagements

• 7 events (3 virtual roundtables, 2 in-person workshops and a conference)

• 18 Nuffield Trust authored articles in journals and external publications

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES (or appropriate) Analysis Guide.

For example, the Comprehensive Geriatric Assessment project, the outputs delivered or planned have so far included:

• A project report to the National Institute of Health Research that is waiting to be published (https://www.journalslibrary.nihr.ac.uk/programmes/hsdr/12500302/#/)

• A peer reviewed publication in the Lancet presenting the methodology and validation of the hospital frailty risk score (https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30668-8/fulltext)

• A toolkit in excel which is available at NHS hospital trust and local authority level, for local needs assessment and benchmarking (https://www.nuffieldtrust.org.uk/research/comprehensive-geriatric-assessment-needs-assessment-tool), that the Nuffield Trust has promoted through its twitter feed (https://twitter.com/NuffieldTrust/status/1034089904735768576, https://twitter.com/NuffieldTrust/status/1033356320915824640) and it has also been disseminated by the British Geriatrics Society (https://www.bgs.org.uk/resources/hospital-wide-comprehensive-geriatric-assessment)

• A guest blog from Professor Simon Conroy on the Hospital Frailty Risk Score discussing the advantages of being able to identify older people at risk in hospitals, and how it could make a real difference (https://www.nuffieldtrust.org.uk/news-item/the-hospital-frailty-risk-score)

• A BGS event for clinicians on frail older people which covered the work from the programme: http://www.acutemedicine.org.uk/wp-content/uploads/2018/04/BGS_Urgentcare_2018_v1-2.pdf

• European Geriatric Medicine Society Conference in October with two posters to present findings: http://www.eugms.org/2018.html

• Two further papers for peer reviewed journals are planned covering specific aspects of the findings.

The project was a finalist in the ONS Research Excellence Awards 2018 (https://www.ons.gov.uk/aboutus/whatwedo/statistics/requestingstatistics/onsresearchexcellenceaward) which was a further opportunity to disseminate the findings and methods within the wider research community.

For the Quality Watch programme, outputs included:

• Over three hundred healthcare quality indicators on a dedicated website (now part of the Nuffield Trust website).

• 14 “Focus on” reports, two briefings, and several data blogs.

• Four annual reports that reviewed the state of care quality in the NHS in England and how it had changed over time and three annual data stories

• The Nuffield Trust provided a platform for internal and external expert commentary. In the last 12 months there have been 11 indicator updates covering a range of topics, 12 ‘latest data’ posts, covering monthly NHS combined performance summary data; the annual data story, this year on quality of remote care, and additional blogs

• Visits to the QualityWatch pages continue to increase, and our audience has diversified to reach a more general public, in addition to organisations and people with a specific interest in health.

Previous outputs have also included:

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been published by NIHR and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. The findings have been presented at conferences, including a presentation on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• Evaluations of new services for patients outside of hospital: Outputs from the project included interim and final reports for funders, Nuffield Trust blogs and other publications. For example, the Nuffield Trust produced reports for Age UK at different stages of the project, a Nuffield Trust report and blog, and a comment article for the Health Service Journal. A seminar on findings from these evaluations took place in 2019.

• Harms and Quality of care measures from routine data: Outputs from the project include a project report for NIHR and papers which have been published in peer reviewed journals.

• Prison Health: Outputs from the project include Nuffield trust reports, blogs and briefings, conference presentations and papers for peer reviewed journals.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

• Organisation of Primary Care: Outputs include a range of external publications and reports, and papers in peer reviewed journals.

The aim is to improve healthcare by influencing policy makers and those involved in delivering services, by improving information available about health services. Standing independent both of the NHS and of politics, the Trust produce incisive commentary and research with three main aims:

• To influence policy and practice through generating and synthesising information on health and social care to facilitate both better policy and better practice.

• To challenge and support those involved in planning and delivering healthcare to think more creatively and innovatively about how to adapt and redesign services to meet changing patient needs;

• To provide information on the evidence, statistics, facts and research which politicians and policymakers use in their interventions in the health and care system in the UK.

As such, the impact is often through influencing the decisions of others, rather than directly changing services. In some cases the analysis contributes to an improved understanding of the challenges facing the NHS, rather than making recommendations for specific service changes. Impact from recently completed projects is summarised below.

Ethnicity coding in the health datasets: This project examined the validity, completeness and consistency of coding in NHS datasets, and how this has changed over time. The report makes recommendations for improving the quality of coding, including calling for updated guidance, as well as recommendations for users of data, to ensure data quality is considered in analysis, and steps take to address data issues. In the process of the research we consulted widely with stakeholders including the National Director for Inequalities at NHS England, Public Health England, the Office of National Statistics, and NHS Digital. We collaborated with NHS Race and Health Observatory (RHO) on the project. The report was widely downloaded when it was published, and we had strong engagement through social media, to ensure we reached a wide audience for our findings. We are now following up on the recommendations, including contributing to a workshop about the recommendations held by the NHS RHO, and a further workshop being held by the Cabinet Office Race Disparity Unit. Through this follow-up work we aim to ensure that the issue of poor quality data on ethnicity is addressed. This will lead to improved coding of data and better information on ethnic health inequalities, to ensure that service improvements can be targeted to improve care for the benefit of patients.

Prisoner health: Our work on prisoner health resulted in the first in-depth analysis of hospital use by prisoners in England, and identified problems for prisoners in accessing services, quality concerns relating to long-term conditions, and poor quality information about pregnancy and birth for prisoners. As well as raising the profile of health issues for prisoners, there are examples where our work has influenced policies impacting prisoners or action being taken. For example, our work on prisoner health was used within the SAGE review of Covid-19 vaccination for prisoners. Further, our recommendation that the prison services should collect and publish data on pregnant women in prisons has been accepted by the Ministry of Justice report ‘Review of operational policy on pregnancy, Mother and Baby Units and maternal separation’: they committed to addressing the recommendation, and said they would publish quarterly figures on the number of pregnant women in prison as well as the yearly number of births. Improving the recording of information about pregnancy in prisons will enable prison health services to plan for the needs of this vulnerable group of prisoners and improve outcomes for mothers and babies.

QualityWatch (QW): The Nuffield Trust use HES data for several QW indicators, for example hip replacements and cancelled operations. This is part of the quality and equity work, and contributes to developing the evidence base and understanding of quality of care and inequalities in health. QW is widely used by policy makers, and senior leaders in the NHS, as well as by a general audience. For example, indicators from QW featured in the Chief Medical Officers last annual report. The programme was also cited in discussions we have had with the NHS England and Improvement Director of Strategy, as providing valuable insights into the state of care, which is informing NHS England and Improvements strategy for addressing the backlog in care caused by Covid. The work of QW to make evidence on the impact of Covid on quality of care makes a valuable contribution to public understanding of the issues facing the NHS, and their impact on patients.

Service evaluations: The Nuffield Trust have used HES data to track emergency admissions in integrated care 'pioneer' areas over time. This is part of the politics, legislation and governance programme, and the work is part of a large scale of evaluation which will contribute to learning and on effectiveness of integrated care. Publications from the project include analysis of differences in emergency admissions between pioneer and non-pioneer areas. A dashboard including the indicators used has also been published, and we have discussed the contents of the dashboard with NHS England and Improvement, which has informed their tracking of progress of integrated care. This project is ongoing.

Service delivery: the Medical Generalism project produce evidence on the impact of the trend towards increased specialisation in medicine on the ability of hospitals to deliver high quality of care, particularly focusing on the care delivered to patients in smaller hospitals. This is an example where analysis of HES data is uniquely able to provide evidence on the quality of care within hospitals, for which aggregate data is not suitable. The results of this study have been recently published by the National Institute for Health Research, and disseminated through events, blogs and social media to reach a wider audience. The study has important lessons for the quality of care for patients, and how this can be improved within the context of the pressures on the health system. The evidence we have generated from the study emphasises the opportunity to importance of continuity of care in providing high-quality of care patients.

Special Measures for Quality (SMQ): This RSET project has examined the impact of this form of regulation on improvements in quality, and has important lessons for policy makers about the future of regulation. The team have carried out a considerable amount of dissemination and engagement with key stakeholders (Department of Health and Social Care, Care Quality Commission, NHS Improvement/England). The latest output (since July 2020) are a rapid literature review published in the International Journal of Health Policy and Management (IJHPM) please see https://www.ijhpm.com/article_3943.html), and a Q&A blogpost on the impact of improvement interventions in NHS trusts (please see - https://www.nuffieldtrust.org.uk/news-item/special-measures-for-quality-a-q-a-on-the-impact-of-improvement-interventions-in-nhs-trusts). The team are currently working on further papers for submission.

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees including a number of senior NHS clinicians, managers and academics. The Nuffield Trust aims to improve the quality of health care to improve the health of people in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

Under the HRA’s GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) “processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject”. The Nuffield Trust’s legitimate interest is carrying out healthcare research in the wider public interest of improved healthcare outcomes for NHS patients.

The Trust relies on Article 9 Condition 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) as the condition for processing “Special” categories of personal data.

The Nuffield Trust has determined that no moral or ethical issues are raised by its processing of HES or other patient data sets (such as emergency or community care data). All data supplied is pseudonymised. All outputs contain only aggregate data, with small numbers suppressed to agreed thresholds in line with the HES (or appropriate) analysis guide.

The Nuffield Trust focuses its activities on six priority areas:

• Health & social care finance and reform

• NHS Workforce

• Older people and complex care

• Quality of care

• New models of health care delivery

• Children and young people

These priorities were set in 2015, with the exception of “Children and Young People” which was added in 2017, and are anticipated to remain valid until at least 2020.

The work of the Trust is organised into a number of programmes which address these priorities. The programmes are broadly aligned to priorities as shown below, with a desire to develop a number of the programmes to cut across one or more of the strategic priorities to maximise reach and impact. For example, the Quality programme is relevant to both Quality of Care and Primary Care priorities, and the Workforce programme impacts on both Workforce and New Models of Care. Each programme has a sponsor at Director level, as well as a programme lead, who is typically a Senior Researcher or Senior Policy Fellow.

The data from NHS Digital is vital to the Trust’s work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. The use within programmes and projects is outlined in more detail in the following sections.

5a.i. Rationale for strategic priorities and programmes

Health & social care finance and reform: The NHS is introducing new models of care and a different way to work with councils through Sustainability and Transformation Partnerships. All of this is being attempted at a time of historic financial constraint, with record trust deficits and an intense search for efficiencies. In addition, Britain's departure from the EU could mean major changes and deep uncertainty for health and social care staffing, regulation and workforce.

The Nuffield Trust’s focus is on improving the quality of policy-making by providing evidence-based analysis, asking insightful questions and providing a challenging view. Programmes in this area cover:

• Commissioning and System Delivery – how the health system is changing, for example, moving to integrated care.

• Topical Issues – for example, issues such as Brexit, which impact on the health system.

• Funding and Sustainability – impact of spending constraints and how funding is allocated to health and care organisations

Workforce: The NHS workforce is under extreme pressure, and it is not just a matter of numbers. Some areas are undersubscribed, but others have an over-supply. New technologies and care models require new types of staff, and training for this highly skilled workforce has long lead times. Getting all of this right has never been more critical, especially with the uncertainty around migration following the country's vote to leave the EU. The development of new models of care is often driven by changes in workforce and vice versa. Being able to make the links between the two areas will be very important.

The workforce programme addresses how the NHS manages workforce pressures and develops a sustainable workforce. The Nuffield Trust ensures their research and analysis informs the Government’s strategy for the future health care workforce.

Older people and complex care: Older people are among the most intense users of health and social care services and opportunities exist for improving the care offered to them. In particular, the needs of older people with multiple health problems and complex conditions are recognised as being a key driver of health service design utilisation and a sentinel marker of the quality of care.

The older people and complex care programme examines models for delivering care to older people and people with complex needs, given the growing number of people with multiple conditions.

Quality of care: Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable – between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled.

The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised.

Quality of care programmes cover:

• Quality – Drawing on its other work programmes, the Nuffield Trust uses expertise in measurement and analysis of quality of care, to provide independent scrutiny, and undertakes research to improve the evidence on quality of care.

• Evaluation – The Nuffield Trust also considers how effective policies intended to improve quality have been, and what can be learnt, in order to influence future decision makers, locally and nationally. This includes evaluating service changes and innovations in the delivery of care.

New models of health care delivery: NHS England’s Long Term Plan, and previously the Five Year Forward View, outlines a vision for how the traditional boundaries between primary care, community services and hospitals will be dissolved over the next five years and beyond. Breaking down the traditional boundaries between different parts of the NHS and social care offers the prospect of reshaping services around the needs of individuals and reducing reliance on hospitals. This will require a move away from single institutions towards networks of care. In this world the capacity of care delivered outside hospitals will need to be boosted through reforms to general practice, while technology and new types of staff enable expertise and information to be shared.

The Nuffield Trust will build on the considerable experience and reputation it has in conducting evaluative work of new models of care, with a number of research projects in place with Royal colleges and specialist societies. The Trust will also provide briefings and analysis that help health leaders choose and implement changes, and bring them together to share ideas.

New models of health care delivery programmes in this area cover:

• Acute Medical Models – models of hospital care and the optimal configuration of services to deliver inpatient services.

• Primary Care – future of primary care and role of primary care in a changing health system.

• Digital – the impact of new technology on delivery of health care, and opportunities to improve care using technology.

Children and young people: The health and wellbeing of children and young people depend on the efforts and commitment of their parents and families, their schools and local communities and the decisions and actions of public service providers and policymakers. This creates a moral imperative to safeguard and promote their interests. There are nearly 20 million people aged 0-24 years old living in the UK, almost a third of the population. There have been long term improvements in health outcomes and quality indicators for children and young people, however, more recently those improvements have slowed or even reversed and internationally he UK compares less well than it might wish. Child health has changed over the last 45 years. Mortality data shows an epidemiological transition away from acute infectious illness towards chronic long-term conditions, yet the way health care services are provided is still heavily hospital focused and reactive. Change has been slow to come due to a long term lack of policy focus on most of the services for children and young people

The Children and Young People work programme will develop the evidence base on how problems and challenges could be addressed by policy and decision makers at a national and local level and/or by individual teams and professionals working with children and young people. The work will have a particular focus on what health care services and systems can do, but will also include how the different parts of the wider context for children and young people interact with each other to address the issues. The Trust also aims to help build networks between different organisations and people who can shape the direction of health care services, health systems and other services.

The work programmes within the strategic priorities were developed in 2018 and are expected to remain aligned to the strategic priorities above until at least 2020 when they may be reviewed.

5a.ii. Use of HES and CSDS data within programmes and projects

Each programme is delivered through individual projects. Projects vary in length and complexity from major research studies which could span a number of years, to shorter term projects resulting in a single output, for example a briefing or presentation. Some Nuffield Trust projects are qualitative, involve analysis of published data, or data from surveys, and so do not require use of HES or other NHS Digital data sources. However, many others are reliant on HES and other NHS Digital-supplied data sets.

To undertake such projects the Nuffield Trust has determined that it requires HES data including the Emergency Care Data Set which will replace the HES A&E datasetand the Community Services Data Set which is now available and is a very welcome addition to the existing hospital based HES datasets.

Although the methods for use of HES and other patient datasets will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken (in addition to analysis of relevant comparator organisations and areas), for example analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as emergency department attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, for example as part of the harm project (see below), and patients with particular needs (eg frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient advice models, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact on outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis – comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, including interrupted time series, panel data and cohort studies;

• Use of the above methods in combination for particular research projects;

• Producing visualisations of analysis and results from the above methods.

The data will not be linked with other record level datasets and there will be no attempt to reidentify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g. indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset.

Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

The number of concurrent projects using HES data will vary, but typically there are 5-6 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Projects are grouped within programmes, but frequently cut across other programmes. As the Nuffield Trust is a small organisation, it seeks to maximise impact by undertaking work which can inform more than one programme or strategic priority. Undertaking projects which support multiple programmes enables the Trust to deliver greater benefits to the health and care system.

5a.iii. How decisions are made about projects and use of HES and CSDS data

This Agreement permits the Nuffield Trust to use the data for the purposes of projects undertaken within the work programmes described above, and which are conceived, planned, approved and initiated through the following process:

1. Projects intended to meet the programme’s aims will be conceived and planned through an iterative process involving the Programme Director and Programme Lead with appropriate input from the Data Protection Officer (DPO). The Programme Director and Programme Lead will ensure that:

• Projects have a clearly defined objectives and operational plans;

• The aims of projects align with at least one of the programme’s aims (as stated above)- all projects must clearly and logically fall within the scope of having one or more of the aims listed and achieving those aims through one or more of the methods listed above ;

• An analysis plan is prepared for each project, setting out the data requirements and methods;

• In each case, the use of the HES data is necessary and proportionate to the purpose of the project and that the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

• Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. A Legitimate Interest Assessment (LIA) will be completed internally for each research project and signed off by the Data Protection Officer (DPO).

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Committee (PPC). This excel template serves a wider purpose than just planning and remains valid throughout a project’s life, serving as the central control document in the management and delivery of the project. The PPC is chaired by the Director of Communications and consists of the Chief Executive, Director of Research, Director of Policy, Senior Fellow, Senior Policy Analyst and other representatives from Research, Policy and Communications. It provides a forum for the discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust.

The committee is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound; and

• Draw fully on the expertise within the Trust including making connections to other related work.

The Committee will approve or recommend the approval of projects in line with the internally approved schedule of management authority and responsibility.

3. The individual or team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales

The time frame for undertaking each project will vary according to project resource, extent of the research and data

analysis required. This is always considered as part of the Project Planning Committee's review and approved based upon

the detail of each individual project.

Project Funding

Some projects under the programme will be funded by the Nuffield Trust but funding is also being sought from other partners including National Research and Charitable organisations, such as the Nuffield Foundation, NIHR and the Health Foundation but may not be limited to these organisations. Funders will take both the forms of partners in collaborative working, as well as commissioners only, that is, the Nuffield Trust will be sole data controller for the analytical work. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose. The Nuffield Trust will not be reliant on securing funding from external partners to complete this research. However, the Nuffield Trust will need to recognise the contribution of any external partner in their outputs.

Data Governance

The Nuffield Trust has independently determined the purposes for which it requires and will process the data under this Agreement in terms of its priorities and programmes which will use the data. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the primary data controller and the sole data controller named in this Agreement. In certain projects which involve collaboration with individuals or organisations outside of the Nuffield Trust, it may be the case that there is joint controllership for the specific project or aspects of it but in all cases, the Nuffield Trust will retain sole autonomy for determining if and how the data under this Agreement will be processed and the Trust cannot be compelled by any third party to process the data for any purpose of in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel (defined as employees, agents and contractors of the Trust) all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust’s secure, ISO27001 certified data environment.

Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

Results may be shared in aggregate form in accordance with the Nuffield Trust's Research Governance framework, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust, and will not be shared with any other third-parties.

The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities and programme(s). Details will also be available on the Nuffield Trust’s website.

5a.iv. Examples of projects

The following examples of projects which have used HES data illustrate the range of work already undertaken in recent years, and upcoming projects. These are provided as examples of work the Nuffield Trust has undertaken or is or will be undertaking to meet the objectives of its programmes in support of its strategic priorities. The examples are not intended to form a comprehensive list of projects permitted under this Agreement.

Integrated care pioneers evaluation

Integrated Care 'Pioneers' are models of care aimed at reducing the impact of boundaries between health and social care providers. The evaluation of the pioneers has wider lessons for the current policy for integrated care systems.

• Nuffield Trust programmes: Commissioning and System Delivery; Evaluation; New models of care

• Overview: The Nuffield Trust is leading on one work package to develop and monitor a set of system level indicators, as part of a wider project with the DH Policy Innovation Research Unit based at the London School of Hygiene and Tropical Medicine. The analysis of HES data to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data.

• Data minimisation approach: This project involves looking at time trends across a range of health and social care indicators. The Pioneers were introduced in 2013 but to have a good understanding of trends prior to this, HES data are used from 2004/05 with follow up until 2019/20, when that data is available. The indicators are presented in aggregated form in a dashboard with local authority district as the lowest geographical unit available. For the majority of indicators all ages are included but where possible indicators are restricted to a particular age group. Only variables relevant for each indicator were used.

• Duration: The project runs from 2016 to 2021

Medical Generalism

The rising numbers of older and more complex patients is one of the most pressing problems facing the NHS. Although they receive the most resource-intensive care, their problems are less likely to be accurately diagnosed and have more adverse outcomes than other age groups. The current models of hospital care, which are heavily based around specialists delivering disease-specific care, serve these patients poorly, as it is often fragmented and poorly co-ordinated. A revival of medical generalism has been suggested to provide better and more cost- effective care. The reality, however, is that there is a paucity of evidence on which to base new models of medical generalism.

• Nuffield trust programmes: Older people and complex care; Workforce

• Overview: The overarching aim of this NIHR funded research project was to identify the models of medical generalism used in smaller hospitals and explore their strengths and weaknesses from patient, professional and service perspectives. The Nuffield Trust used HES data to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals.

• Data minimisation approach: Hospital Episode Statistics data (year range 2007/08–2017/18) was used to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals. The final sample included 69 smaller NHS Trusts providing acute medical care in England, although some analyses used only 68 hospitals as a result of merges during the data period.

A data set based on ‘Index episodes of care’ for emergency admissions across the selected generalist medicine specialities identified in HES inpatient data 2012/13 for the smaller hospitals cohort was created so that five years of prior patient history for cancer patients as well as three years of subsequent history could be included. Cases with a specific diagnosis indicating specialist care or where patients had been transferred out of hospital were excluded. A data set was created covering 1.9 million episodes in the selected smaller hospitals.

Following development of the case mix classification, analysis focused on different data collection years dependent on the nature of the research question. For instance, the degree of alignment between patient case mix and medical generalist/skills mix in smaller hospitals was assessed was based on analysis of the 2015/16 data set. Population level analysis of admission patterns used data from 2017/18.

• Duration: The main research phase of the project was from 2016 to 2018.

This analysis was undertaken under a previous Data Sharing Agreement with NHS Digital in conjunction with University College London (UCL) which undertook parts of the analysis as a data processor under contract to the Nuffield Trust. Nuffield Trust defined the analysis and manner in which the data should be processed as well as the tools used (i.e. the software). The Trust also owns the outputs. The UCL Department of Applied Health research team carried out the above work on the Nuffield Trust site under the same Information Governance arrangements as Nuffield Trust staff - i.e. a signed confidentiality agreement as evidence that they had read and understood the Trust's Information Security Management System (ISMS), completed equivalent training and worked in the same environment. No data provided by NHS Digital left the Nuffield Trust site and the data was only processed on Nuffield Trust servers.

Evaluations of new services for patients outside of hospital

The Nuffield Trust has undertaken a number of evaluations of programmes to reduce admissions or readmissions to hospital, through better support for patients at home.

• Nuffield Trust programmes: Older people and complex care; New models of care; Evaluations

• Overview: This is a programme of work undertaken over a number of years. For example, the Nuffield Trust undertook a project funded by Age UK of services to support often vulnerable older people at high risk of hospital admission, which assessed the impact of this service on future emergency admissions. Previous projects were undertaken for a range of funders including the British Red Cross, Royal Voluntary Services and the Cabinet Office.

• Data minimisation approach: These projects took place using data linkage to HES data for specific cohorts of patients receiving innovative services. The Nuffield Trust used HES data from two years prior to the date each individual started receiving the service to allow for calculation of risk of emergency admission scores. To implement these models, the Nuffield Trust used a selected list of fields from inpatient, outpatient and A&E datasets (including admission method, diagnosis codes, procedure codes, A&E investigation codes, outpatient attended/did not attend). The Nuffield Trust compared each group of service recipients with a selected matched control group – matching one control person to each service recipient. The Nuffield Trust selected controls from a larger pool of possible control individuals. This larger pool of possible control individuals were selected to reflect the intervention cohort – eg they were the same ages (for example 55+ or 60+), and lived in similar areas (as defined by ONS and other analysis). Risk scores were calculated for all possible controls – involving again two prior years of HES data. For the evaluation outcomes, the intervention and matched control groups were followed up for a limited number of months (maximum 16, but more usually 6-9) on a limited number of activity measures.

• Duration: This was undertaken from 2013 to 2019

Comprehensive Geriatric Assessment (CGA)

The CGA was a multidisciplinary process designed to assess a frail older person’s medical conditions, mental health, functional capacity and social circumstances. However information is lacking on the types of patients that might benefit the most. The aim of this work is to describe existing models of care and to validate tools to deliver CGA on a hospital wide basis.

• Nuffield Trust programmes: Older people and complex care; Quality of Care; Evaluations

• Overview: This was a collaborative project undertaken as an NIHR funded project undertaken with the Universities of Leicester and Newcastle. The project incorporated linked HES, ONS and clinical data, managed under a specific Data Sharing Agreement (DARS-NIC-383324-D6B8T).

• Data minimisation approach: Three clinical cohorts were linked to HES as part of this project. The recruitment dates for these ranged from 2006 to 2012 and therefore, HES was required from 2004/05 to 2016/17 to allow a period g

A key aspect for all the research projects undertaken is ensuring that learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with their charitable purposes.

A communications plan is developed for each programme and project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to programmes and projects related to that priority.

Outputs from a project could include:

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust will use their extensive communications facilities & networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

The combination of outputs will vary from project to project.

For example, the Comprehensive Geriatric Assessment project, the outputs delivered or planned have so far included:

• A project report to the National Institute of Health Research that is waiting to be published (https://www.journalslibrary.nihr.ac.uk/programmes/hsdr/12500302/#/)

• A peer reviewed publication in the Lancet presenting the methodology and validation of the hospital frailty risk score (https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30668-8/fulltext)

• A toolkit in excel which is available at NHS hospital trust and local authority level, for local needs assessment and benchmarking (https://www.nuffieldtrust.org.uk/research/comprehensive-geriatric-assessment-needs-assessment-tool), that the Nuffield Trust has promoted through its twitter feed (https://twitter.com/NuffieldTrust/status/1034089904735768576, https://twitter.com/NuffieldTrust/status/1033356320915824640) and it has also been disseminated by the British Geriatrics Society (https://www.bgs.org.uk/resources/hospital-wide-comprehensive-geriatric-assessment)

• A guest blog from Professor Simon Conroy on the Hospital Frailty Risk Score discussing the advantages of being able to identify older people at risk in hospitals, and how it could make a real difference (https://www.nuffieldtrust.org.uk/news-item/the-hospital-frailty-risk-score)

• A BGS event for clinicians on frail older people which covered the work from the programme: http://www.acutemedicine.org.uk/wp-content/uploads/2018/04/BGS_Urgentcare_2018_v1-2.pdf

• European Geriatric Medicine Society Conference in October with two posters to present findings: http://www.eugms.org/2018.html

• Two further papers for peer reviewed journals are planned covering specific aspects of the findings.

• The project was a finalist in the ONS Research Excellence Awards 2018 (https://www.ons.gov.uk/aboutus/whatwedo/statistics/requestingstatistics/onsresearchexcellenceaward) which was a further opportunity to disseminate the findings and methods within the wider research community.

For the Quality Watch programme, outputs included:

• Over three hundred healthcare quality indicators on a dedicated website (now part of the Nuffield Trust website).

• 14 “Focus on” reports, two briefings, and several data blogs.

• Four annual reports that reviewed the state of care quality in the NHS in England and how it had changed over time.

• The Nuffield Trust provided a platform for internal and external expert commentary, with 135 editorial items (79 blogs (34 by external authors); 17 ‘latest data’ posts, covering monthly NHS combined performance summary data; 13 ‘indicator update’ posts, detailing stories emerging from ongoing data updates on the site; 26 news stories (mostly comprising press releases)).

• The Nuffield Trust also held a large number of events (QW conferences: October 2013, October 2014, November 2015; All Parliamentary Health Group events, February 2014 and December 2014; Social care event at Nuffield, May 2014; Allied Health Professionals event at Nuffield, November 2014; Public health roundtable at Nuffield, June 2016; Children and Young People roundtable at Nuffield, June 2017).

Previous outputs have also included:

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been submitted to NIHR for review and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. A conference presentation has been accepted on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• Evaluations of new services for patients outside of hospital: Outputs from the project included interim and final reports for funders, Nuffield Trust blogs and other publications. For example, the Nuffield Trust produced reports for Age UK at different stages of the project, a Nuffield Trust report and blog, and a comment article for the Health Service Journal. A seminar on findings from these evaluations is planned for later in 2019.

• Comprehensive Geriatric Assessment (CGA) Outputs from the project are described in detail above. Further work on peer reviewed papers from the project is ongoing.

• Harms and Quality of care measures from routine data: Outputs from the project include a project report for NIHR and papers which have been submitted to peer reviewed journals.

• Prison Health: Outputs from the project include Nuffield trust reports, blogs and briefings, conference presentations and papers for peer reviewed journals.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

• Organisation of Primary Care: Outputs include a range of external publications and reports, and papers in peer reviewed journals.

In the past year outputs for the Nuffield Trust as a whole have included:

• 25 reports (in all cases with complementary blogs, charts or infographics)

• 12 briefings and explainers

• 90 blogs and long reads

• 10 charts and infographics (in addition to those in reports, blogs and briefings)

• 44 press releases

• Approximately 1200 Nuffield Trust tweets (The Trust has 45000 followers)

• 20 citations for NT staff in external peer reviewed journals

• 70 speaking engagements

• 10 corporate events

• 426,147 web site visits, averaging 1,674 per day across 610,413 sessions

• 100+ updated QualityWatch indicators and around 300 tweets

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES (or appropriate) Analysis Guide.

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees including a number of senior NHS clinicians, managers and academics. The Nuffield Trust aims to improve the quality of health care to improve the health of people in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

Under the HRA’s GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) “processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject”. The Nuffield Trust’s legitimate interest is carrying out healthcare research in the wider public interest of improved healthcare outcomes for NHS patients.

The Trust relies on Article 9 Condition 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) as the condition for processing “Special” categories of personal data.

The Nuffield Trust has determined that no moral or ethical issues are raised by its processing of HES or other patient data sets (such as emergency or community care data). All data supplied is pseudonymised. All outputs contain only aggregate data, with small numbers suppressed to agreed thresholds in line with the HES (or appropriate) analysis guide.

The Nuffield Trust focuses its activities on six priority areas:

• Health & social care finance and reform

• NHS Workforce

• Older people and complex care

• Quality of care

• New models of health care delivery

• Children and young people

These priorities were set in 2015, with the exception of “Children and Young People” which was added in 2017, and are anticipated to remain valid until at least 2020.

The work of the Trust is organised into a number of programmes which address these priorities. The programmes are broadly aligned to priorities as shown below, with a desire to develop a number of the programmes to cut across one or more of the strategic priorities to maximise reach and impact. For example, the Quality programme is relevant to both Quality of Care and Primary Care priorities, and the Workforce programme impacts on both Workforce and New Models of Care. Each programme has a sponsor at Director level, as well as a programme lead, who is typically a Senior Researcher or Senior Policy Fellow.

The data from NHS Digital is vital to the Trust’s work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. The use within programmes and projects is outlined in more detail in the following sections.

5a.i. Rationale for strategic priorities and programmes

Health & social care finance and reform: The NHS is introducing new models of care and a different way to work with councils through Sustainability and Transformation Partnerships. All of this is being attempted at a time of historic financial constraint, with record trust deficits and an intense search for efficiencies. In addition, Britain's departure from the EU could mean major changes and deep uncertainty for health and social care staffing, regulation and workforce.

The Nuffield Trust’s focus is on improving the quality of policy-making by providing evidence-based analysis, asking insightful questions and providing a challenging view. Programmes in this area cover:

• Commissioning and System Delivery – how the health system is changing, for example, moving to integrated care.

• Topical Issues – for example, issues such as Brexit, which impact on the health system.

• Funding and Sustainability – impact of spending constraints and how funding is allocated to health and care organisations

Workforce: The NHS workforce is under extreme pressure, and it is not just a matter of numbers. Some areas are undersubscribed, but others have an over-supply. New technologies and care models require new types of staff, and training for this highly skilled workforce has long lead times. Getting all of this right has never been more critical, especially with the uncertainty around migration following the country's vote to leave the EU. The development of new models of care is often driven by changes in workforce and vice versa. Being able to make the links between the two areas will be very important.

The workforce programme addresses how the NHS manages workforce pressures and develops a sustainable workforce. The Nuffield Trust ensures their research and analysis informs the Government’s strategy for the future health care workforce.

Older people and complex care: Older people are among the most intense users of health and social care services and opportunities exist for improving the care offered to them. In particular, the needs of older people with multiple health problems and complex conditions are recognised as being a key driver of health service design utilisation and a sentinel marker of the quality of care.

The older people and complex care programme examines models for delivering care to older people and people with complex needs, given the growing number of people with multiple conditions.

Quality of care: Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable – between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled.

The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised.

Quality of care programmes cover:

• Quality – Drawing on its other work programmes, the Nuffield Trust uses expertise in measurement and analysis of quality of care, to provide independent scrutiny, and undertakes research to improve the evidence on quality of care.

• Evaluation – The Nuffield Trust also considers how effective policies intended to improve quality have been, and what can be learnt, in order to influence future decision makers, locally and nationally. This includes evaluating service changes and innovations in the delivery of care.

New models of health care delivery: NHS England’s Long Term Plan, and previously the Five Year Forward View, outlines a vision for how the traditional boundaries between primary care, community services and hospitals will be dissolved over the next five years and beyond. Breaking down the traditional boundaries between different parts of the NHS and social care offers the prospect of reshaping services around the needs of individuals and reducing reliance on hospitals. This will require a move away from single institutions towards networks of care. In this world the capacity of care delivered outside hospitals will need to be boosted through reforms to general practice, while technology and new types of staff enable expertise and information to be shared.

The Nuffield Trust will build on the considerable experience and reputation it has in conducting evaluative work of new models of care, with a number of research projects in place with Royal colleges and specialist societies. The Trust will also provide briefings and analysis that help health leaders choose and implement changes, and bring them together to share ideas.

New models of health care delivery programmes in this area cover:

• Acute Medical Models – models of hospital care and the optimal configuration of services to deliver inpatient services.

• Primary Care – future of primary care and role of primary care in a changing health system.

• Digital – the impact of new technology on delivery of health care, and opportunities to improve care using technology.

Children and young people: The health and wellbeing of children and young people depend on the efforts and commitment of their parents and families, their schools and local communities and the decisions and actions of public service providers and policymakers. This creates a moral imperative to safeguard and promote their interests. There are nearly 20 million people aged 0-24 years old living in the UK, almost a third of the population. There have been long term improvements in health outcomes and quality indicators for children and young people, however, more recently those improvements have slowed or even reversed and internationally he UK compares less well than it might wish. Child health has changed over the last 45 years. Mortality data shows an epidemiological transition away from acute infectious illness towards chronic long-term conditions, yet the way health care services are provided is still heavily hospital focused and reactive. Change has been slow to come due to a long term lack of policy focus on most of the services for children and young people

The Children and Young People work programme will develop the evidence base on how problems and challenges could be addressed by policy and decision makers at a national and local level and/or by individual teams and professionals working with children and young people. The work will have a particular focus on what health care services and systems can do, but will also include how the different parts of the wider context for children and young people interact with each other to address the issues. The Trust also aims to help build networks between different organisations and people who can shape the direction of health care services, health systems and other services.

The work programmes within the strategic priorities were developed in 2018 and are expected to remain aligned to the strategic priorities above until at least 2020 when they may be reviewed.

5a.ii. Use of HES and CSDS data within programmes and projects

Each programme is delivered through individual projects. Projects vary in length and complexity from major research studies which could span a number of years, to shorter term projects resulting in a single output, for example a briefing or presentation. Some Nuffield Trust projects are qualitative, involve analysis of published data, or data from surveys, and so do not require use of HES or other NHS Digital data sources. However, many others are reliant on HES and other NHS Digital-supplied data sets.

To undertake such projects the Nuffield Trust has determined that it requires HES data including the Emergency Care Data Set which will replace the HES A&E datasetand the Community Services Data Set which is now available and is a very welcome addition to the existing hospital based HES datasets.

Although the methods for use of HES and other patient datasets will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken (in addition to analysis of relevant comparator organisations and areas), for example analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as emergency department attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, for example as part of the harm project (see below), and patients with particular needs (eg frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient advice models, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact on outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis – comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, including interrupted time series, panel data and cohort studies;

• Use of the above methods in combination for particular research projects;

• Producing visualisations of analysis and results from the above methods.

The data will not be linked with other record level datasets and there will be no attempt to reidentify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g. indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset.

Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

The number of concurrent projects using HES data will vary, but typically there are 5-6 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Projects are grouped within programmes, but frequently cut across other programmes. As the Nuffield Trust is a small organisation, it seeks to maximise impact by undertaking work which can inform more than one programme or strategic priority. Undertaking projects which support multiple programmes enables the Trust to deliver greater benefits to the health and care system.

5a.iii. How decisions are made about projects and use of HES and CSDS data

This Agreement permits the Nuffield Trust to use the data for the purposes of projects undertaken within the work programmes described above, and which are conceived, planned, approved and initiated through the following process:

1. Projects intended to meet the programme’s aims will be conceived and planned through an iterative process involving the Programme Director and Programme Lead with appropriate input from the Data Protection Officer (DPO). The Programme Director and Programme Lead will ensure that:

• Projects have a clearly defined objectives and operational plans;

• The aims of projects align with at least one of the programme’s aims (as stated above)- all projects must clearly and logically fall within the scope of having one or more of the aims listed and achieving those aims through one or more of the methods listed above ;

• An analysis plan is prepared for each project, setting out the data requirements and methods;

• In each case, the use of the HES data is necessary and proportionate to the purpose of the project and that the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

• Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. A Legitimate Interest Assessment (LIA) will be completed internally for each research project and signed off by the Data Protection Officer (DPO).

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Committee (PPC). This excel template serves a wider purpose than just planning and remains valid throughout a project’s life, serving as the central control document in the management and delivery of the project. The PPC is chaired by the Director of Communications and consists of the Chief Executive, Director of Research, Director of Policy, Senior Fellow, Senior Policy Analyst and other representatives from Research, Policy and Communications. It provides a forum for the discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust.

The committee is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound; and

• Draw fully on the expertise within the Trust including making connections to other related work.

The Committee will approve or recommend the approval of projects in line with the internally approved schedule of management authority and responsibility.

3. The individual or team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales

The time frame for undertaking each project will vary according to project resource, extent of the research and data

analysis required. This is always considered as part of the Project Planning Committee's review and approved based upon

the detail of each individual project.

Project Funding

Some projects under the programme will be funded by the Nuffield Trust but funding is also being sought from other partners including National Research and Charitable organisations, such as the Nuffield Foundation, NIHR and the Health Foundation but may not be limited to these organisations. Funders will take both the forms of partners in collaborative working, as well as commissioners only, that is, the Nuffield Trust will be sole data controller for the analytical work. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose. The Nuffield Trust will not be reliant on securing funding from external partners to complete this research. However, the Nuffield Trust will need to recognise the contribution of any external partner in their outputs.

Data Governance

The Nuffield Trust has independently determined the purposes for which it requires and will process the data under this Agreement in terms of its priorities and programmes which will use the data. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the primary data controller and the sole data controller named in this Agreement. In certain projects which involve collaboration with individuals or organisations outside of the Nuffield Trust, it may be the case that there is joint controllership for the specific project or aspects of it but in all cases, the Nuffield Trust will retain sole autonomy for determining if and how the data under this Agreement will be processed and the Trust cannot be compelled by any third party to process the data for any purpose of in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel (defined as employees, agents and contractors of the Trust) all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust’s secure, ISO27001 certified data environment.

Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

Results may be shared in aggregate form in accordance with the Nuffield Trust's Research Governance framework, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust, and will not be shared with any other third-parties.

The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities and programme(s). Details will also be available on the Nuffield Trust’s website.

5a.iv. Examples of projects

The following examples of projects which have used HES data illustrate the range of work already undertaken in recent years, and upcoming projects. These are provided as examples of work the Nuffield Trust has undertaken or is or will be undertaking to meet the objectives of its programmes in support of its strategic priorities. The examples are not intended to form a comprehensive list of projects permitted under this Agreement.

Integrated care pioneers evaluation

Integrated Care 'Pioneers' are models of care aimed at reducing the impact of boundaries between health and social care providers. The evaluation of the pioneers has wider lessons for the current policy for integrated care systems.

• Nuffield Trust programmes: Commissioning and System Delivery; Evaluation; New models of care

• Overview: The Nuffield Trust is leading on one work package to develop and monitor a set of system level indicators, as part of a wider project with the DH Policy Innovation Research Unit based at the London School of Hygiene and Tropical Medicine. The analysis of HES data to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data.

• Data minimisation approach: This project involves looking at time trends across a range of health and social care indicators. The Pioneers were introduced in 2013 but to have a good understanding of trends prior to this, HES data are used from 2004/05 with follow up until 2019/20, when that data is available. The indicators are presented in aggregated form in a dashboard with local authority district as the lowest geographical unit available. For the majority of indicators all ages are included but where possible indicators are restricted to a particular age group. Only variables relevant for each indicator were used.

• Duration: The project runs from 2016 to 2021

Medical Generalism

The rising numbers of older and more complex patients is one of the most pressing problems facing the NHS. Although they receive the most resource-intensive care, their problems are less likely to be accurately diagnosed and have more adverse outcomes than other age groups. The current models of hospital care, which are heavily based around specialists delivering disease-specific care, serve these patients poorly, as it is often fragmented and poorly co-ordinated. A revival of medical generalism has been suggested to provide better and more cost- effective care. The reality, however, is that there is a paucity of evidence on which to base new models of medical generalism.

• Nuffield trust programmes: Older people and complex care; Workforce

• Overview: The overarching aim of this NIHR funded research project was to identify the models of medical generalism used in smaller hospitals and explore their strengths and weaknesses from patient, professional and service perspectives. The Nuffield Trust used HES data to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals.

• Data minimisation approach: Hospital Episode Statistics data (year range 2007/08–2017/18) was used to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals. The final sample included 69 smaller NHS Trusts providing acute medical care in England, although some analyses used only 68 hospitals as a result of merges during the data period.

A data set based on ‘Index episodes of care’ for emergency admissions across the selected generalist medicine specialities identified in HES inpatient data 2012/13 for the smaller hospitals cohort was created so that five years of prior patient history for cancer patients as well as three years of subsequent history could be included. Cases with a specific diagnosis indicating specialist care or where patients had been transferred out of hospital were excluded. A data set was created covering 1.9 million episodes in the selected smaller hospitals.

Following development of the case mix classification, analysis focused on different data collection years dependent on the nature of the research question. For instance, the degree of alignment between patient case mix and medical generalist/skills mix in smaller hospitals was assessed was based on analysis of the 2015/16 data set. Population level analysis of admission patterns used data from 2017/18.

• Duration: The main research phase of the project was from 2016 to 2018.

This analysis was undertaken under a previous Data Sharing Agreement with NHS Digital in conjunction with University College London (UCL) which undertook parts of the analysis as a data processor under contract to the Nuffield Trust. Nuffield Trust defined the analysis and manner in which the data should be processed as well as the tools used (i.e. the software). The Trust also owns the outputs. The UCL Department of Applied Health research team carried out the above work on the Nuffield Trust site under the same Information Governance arrangements as Nuffield Trust staff - i.e. a signed confidentiality agreement as evidence that they had read and understood the Trust's Information Security Management System (ISMS), completed equivalent training and worked in the same environment. No data provided by NHS Digital left the Nuffield Trust site and the data was only processed on Nuffield Trust servers.

Evaluations of new services for patients outside of hospital

The Nuffield Trust has undertaken a number of evaluations of programmes to reduce admissions or readmissions to hospital, through better support for patients at home.

• Nuffield Trust programmes: Older people and complex care; New models of care; Evaluations

• Overview: This is a programme of work undertaken over a number of years. For example, the Nuffield Trust undertook a project funded by Age UK of services to support often vulnerable older people at high risk of hospital admission, which assessed the impact of this service on future emergency admissions. Previous projects were undertaken for a range of funders including the British Red Cross, Royal Voluntary Services and the Cabinet Office.

• Data minimisation approach: These projects took place using data linkage to HES data for specific cohorts of patients receiving innovative services. The Nuffield Trust used HES data from two years prior to the date each individual started receiving the service to allow for calculation of risk of emergency admission scores. To implement these models, the Nuffield Trust used a selected list of fields from inpatient, outpatient and A&E datasets (including admission method, diagnosis codes, procedure codes, A&E investigation codes, outpatient attended/did not attend). The Nuffield Trust compared each group of service recipients with a selected matched control group – matching one control person to each service recipient. The Nuffield Trust selected controls from a larger pool of possible control individuals. This larger pool of possible control individuals were selected to reflect the intervention cohort – eg they were the same ages (for example 55+ or 60+), and lived in similar areas (as defined by ONS and other analysis). Risk scores were calculated for all possible controls – involving again two prior years of HES data. For the evaluation outcomes, the intervention and matched control groups were followed up for a limited number of months (maximum 16, but more usually 6-9) on a limited number of activity measures.

• Duration: This was undertaken from 2013 to 2019

Comprehensive Geriatric Assessment (CGA)

The CGA was a multidisciplinary process designed to assess a frail older person’s medical conditions, mental health, functional capacity and social circumstances. However information is lacking on the types of patients that might benefit the most. The aim of this work is to describe existing models of care and to validate tools to deliver CGA on a hospital wide basis.

• Nuffield Trust programmes: Older people and complex care; Quality of Care; Evaluations

• Overview: This was a collaborative project undertaken as an NIHR funded project undertaken with the Universities of Leicester and Newcastle. The project incorporated linked HES, ONS and clinical data, managed under a specific Data Sharing Agreement (DARS-NIC-383324-D6B8T).

• Data minimisation approach: Three clinical

A key aspect for all the research projects undertaken is ensuring that learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with their charitable purposes.

A communications plan is developed for each programme and project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to programmes and projects related to that priority.

Outputs from a project could include:

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust will use their extensive communications facilities & networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

The combination of outputs will vary from project to project.

For example, the Comprehensive Geriatric Assessment project, the outputs delivered or planned have so far included:

• A project report to the National Institute of Health Research that is waiting to be published (https://www.journalslibrary.nihr.ac.uk/programmes/hsdr/12500302/#/)

• A peer reviewed publication in the Lancet presenting the methodology and validation of the hospital frailty risk score (https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30668-8/fulltext)

• A toolkit in excel which is available at NHS hospital trust and local authority level, for local needs assessment and benchmarking (https://www.nuffieldtrust.org.uk/research/comprehensive-geriatric-assessment-needs-assessment-tool), that the Nuffield Trust has promoted through its twitter feed (https://twitter.com/NuffieldTrust/status/1034089904735768576, https://twitter.com/NuffieldTrust/status/1033356320915824640) and it has also been disseminated by the British Geriatrics Society (https://www.bgs.org.uk/resources/hospital-wide-comprehensive-geriatric-assessment)

• A guest blog from Professor Simon Conroy on the Hospital Frailty Risk Score discussing the advantages of being able to identify older people at risk in hospitals, and how it could make a real difference (https://www.nuffieldtrust.org.uk/news-item/the-hospital-frailty-risk-score)

• A BGS event for clinicians on frail older people which covered the work from the programme: http://www.acutemedicine.org.uk/wp-content/uploads/2018/04/BGS_Urgentcare_2018_v1-2.pdf

• European Geriatric Medicine Society Conference in October with two posters to present findings: http://www.eugms.org/2018.html

• Two further papers for peer reviewed journals are planned covering specific aspects of the findings.

• The project was a finalist in the ONS Research Excellence Awards 2018 (https://www.ons.gov.uk/aboutus/whatwedo/statistics/requestingstatistics/onsresearchexcellenceaward) which was a further opportunity to disseminate the findings and methods within the wider research community.

For the Quality Watch programme, outputs included:

• Over three hundred healthcare quality indicators on a dedicated website (now part of the Nuffield Trust website).

• 14 “Focus on” reports, two briefings, and several data blogs.

• Four annual reports that reviewed the state of care quality in the NHS in England and how it had changed over time.

• The Nuffield Trust provided a platform for internal and external expert commentary, with 135 editorial items (79 blogs (34 by external authors); 17 ‘latest data’ posts, covering monthly NHS combined performance summary data; 13 ‘indicator update’ posts, detailing stories emerging from ongoing data updates on the site; 26 news stories (mostly comprising press releases)).

• The Nuffield Trust also held a large number of events (QW conferences: October 2013, October 2014, November 2015; All Parliamentary Health Group events, February 2014 and December 2014; Social care event at Nuffield, May 2014; Allied Health Professionals event at Nuffield, November 2014; Public health roundtable at Nuffield, June 2016; Children and Young People roundtable at Nuffield, June 2017).

Previous outputs have also included:

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been submitted to NIHR for review and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. A conference presentation has been accepted on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• Evaluations of new services for patients outside of hospital: Outputs from the project included interim and final reports for funders, Nuffield Trust blogs and other publications. For example, the Nuffield Trust produced reports for Age UK at different stages of the project, a Nuffield Trust report and blog, and a comment article for the Health Service Journal. A seminar on findings from these evaluations is planned for later in 2019.

• Comprehensive Geriatric Assessment (CGA) Outputs from the project are described in detail above. Further work on peer reviewed papers from the project is ongoing.

• Harms and Quality of care measures from routine data: Outputs from the project include a project report for NIHR and papers which have been submitted to peer reviewed journals.

• Prison Health: Outputs from the project include Nuffield trust reports, blogs and briefings, conference presentations and papers for peer reviewed journals.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

• Organisation of Primary Care: Outputs include a range of external publications and reports, and papers in peer reviewed journals.

In the past year outputs for the Nuffield Trust as a whole have included:

• 25 reports (in all cases with complementary blogs, charts or infographics)

• 12 briefings and explainers

• 90 blogs and long reads

• 10 charts and infographics (in addition to those in reports, blogs and briefings)

• 44 press releases

• Approximately 1200 Nuffield Trust tweets (The Trust has 45000 followers)

• 20 citations for NT staff in external peer reviewed journals

• 70 speaking engagements

• 10 corporate events

• 426,147 web site visits, averaging 1,674 per day across 610,413 sessions

• 100+ updated QualityWatch indicators and around 300 tweets

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES (or appropriate) Analysis Guide.

The Nuffield Trust for Research and Policy Studies in Health Services (The Nuffield Trust) is an independent health research charity overseen by a board of Trustees including a number of senior NHS clinicians, managers and academics. The Nuffield Trust aims to improve the quality of health care to improve the health of people in the UK by providing evidence-based research and policy analysis and informing and generating debate. It provides a trusted and respected voice at a time of unprecedented challenge to the NHS and social care system.

Under the HRA’s GDPR Operational Guidance the Nuffield Trust therefore relies on Article 6 (1) (f) “processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject”. The Nuffield Trust’s legitimate interest is carrying out healthcare research in the wider public interest of improved healthcare outcomes for NHS patients.

The Trust relies on Article 9 Condition 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) as the condition for processing “Special” categories of personal data.

The Nuffield Trust has determined that no moral or ethical issues are raised by its processing of HES or other patient data sets (such as emergency or community care data). All data supplied is pseudonymised. All outputs contain only aggregate data, with small numbers suppressed to agreed thresholds in line with the HES (or appropriate) analysis guide.

The Nuffield Trust focuses its activities on six priority areas:

• Health & social care finance and reform

• NHS Workforce

• Older people and complex care

• Quality of care

• New models of health care delivery

• Children and young people

These priorities were set in 2015, with the exception of “Children and Young People” which was added in 2017, and are anticipated to remain valid until at least 2020.

The work of the Trust is organised into a number of programmes which address these priorities. The programmes are broadly aligned to priorities as shown below, with a desire to develop a number of the programmes to cut across one or more of the strategic priorities to maximise reach and impact. For example, the Quality programme is relevant to both Quality of Care and Primary Care priorities, and the Workforce programme impacts on both Workforce and New Models of Care. Each programme has a sponsor at Director level, as well as a programme lead, who is typically a Senior Researcher or Senior Policy Fellow.

The data from NHS Digital is vital to the Trust’s work because it is an essential source of information on patient activity and outcomes, which allows comparisons across different parts of the NHS and over time. The use within programmes and projects is outlined in more detail in the following sections.

5a.i. Rationale for strategic priorities and programmes

Health & social care finance and reform: The NHS is introducing new models of care and a different way to work with councils through Sustainability and Transformation Partnerships. All of this is being attempted at a time of historic financial constraint, with record trust deficits and an intense search for efficiencies. In addition, Britain's departure from the EU could mean major changes and deep uncertainty for health and social care staffing, regulation and workforce.

The Nuffield Trust’s focus is on improving the quality of policy-making by providing evidence-based analysis, asking insightful questions and providing a challenging view. Programmes in this area cover:

• Commissioning and System Delivery – how the health system is changing, for example, moving to integrated care.

• Topical Issues – for example, issues such as Brexit, which impact on the health system.

• Funding and Sustainability – impact of spending constraints and how funding is allocated to health and care organisations

Workforce: The NHS workforce is under extreme pressure, and it is not just a matter of numbers. Some areas are undersubscribed, but others have an over-supply. New technologies and care models require new types of staff, and training for this highly skilled workforce has long lead times. Getting all of this right has never been more critical, especially with the uncertainty around migration following the country's vote to leave the EU. The development of new models of care is often driven by changes in workforce and vice versa. Being able to make the links between the two areas will be very important.

The workforce programme addresses how the NHS manages workforce pressures and develops a sustainable workforce. The Nuffield Trust ensures their research and analysis informs the Government’s strategy for the future health care workforce.

Older people and complex care: Older people are among the most intense users of health and social care services and opportunities exist for improving the care offered to them. In particular, the needs of older people with multiple health problems and complex conditions are recognised as being a key driver of health service design utilisation and a sentinel marker of the quality of care.

The older people and complex care programme examines models for delivering care to older people and people with complex needs, given the growing number of people with multiple conditions.

Quality of care: Patients and the public expect to receive high quality, safe care, where and when they need it. Despite this, it is known that the quality of care is variable – between organisations, different conditions, and different patient groups. It is also known that the UK lags behind other similar countries in treatment of common diseases, and while some aspects of care have improved over time, for many aspects of care improvements in quality have stalled.

The level of funding for health and social care will influence what can be achieved, but regardless of this, the Nuffield Trust needs to understand how the quality of care is changing, to generate evidence on what can be done to improve quality, and to ensure that improvements which will make the most difference to patients and the public are prioritised.

Quality of care programmes cover:

• Quality – Drawing on its other work programmes, the Nuffield Trust uses expertise in measurement and analysis of quality of care, to provide independent scrutiny, and undertakes research to improve the evidence on quality of care.

• Evaluation – The Nuffield Trust also considers how effective policies intended to improve quality have been, and what can be learnt, in order to influence future decision makers, locally and nationally. This includes evaluating service changes and innovations in the delivery of care.

New models of health care delivery: NHS England’s Long Term Plan, and previously the Five Year Forward View, outlines a vision for how the traditional boundaries between primary care, community services and hospitals will be dissolved over the next five years and beyond. Breaking down the traditional boundaries between different parts of the NHS and social care offers the prospect of reshaping services around the needs of individuals and reducing reliance on hospitals. This will require a move away from single institutions towards networks of care. In this world the capacity of care delivered outside hospitals will need to be boosted through reforms to general practice, while technology and new types of staff enable expertise and information to be shared.

The Nuffield Trust will build on the considerable experience and reputation it has in conducting evaluative work of new models of care, with a number of research projects in place with Royal colleges and specialist societies. The Trust will also provide briefings and analysis that help health leaders choose and implement changes, and bring them together to share ideas.

New models of health care delivery programmes in this area cover:

• Acute Medical Models – models of hospital care and the optimal configuration of services to deliver inpatient services.

• Primary Care – future of primary care and role of primary care in a changing health system.

• Digital – the impact of new technology on delivery of health care, and opportunities to improve care using technology.

Children and young people: The health and wellbeing of children and young people depend on the efforts and commitment of their parents and families, their schools and local communities and the decisions and actions of public service providers and policymakers. This creates a moral imperative to safeguard and promote their interests. There are nearly 20 million people aged 0-24 years old living in the UK, almost a third of the population. There have been long term improvements in health outcomes and quality indicators for children and young people, however, more recently those improvements have slowed or even reversed and internationally he UK compares less well than it might wish. Child health has changed over the last 45 years. Mortality data shows an epidemiological transition away from acute infectious illness towards chronic long-term conditions, yet the way health care services are provided is still heavily hospital focused and reactive. Change has been slow to come due to a long term lack of policy focus on most of the services for children and young people

The Children and Young People work programme will develop the evidence base on how problems and challenges could be addressed by policy and decision makers at a national and local level and/or by individual teams and professionals working with children and young people. The work will have a particular focus on what health care services and systems can do, but will also include how the different parts of the wider context for children and young people interact with each other to address the issues. The Trust also aims to help build networks between different organisations and people who can shape the direction of health care services, health systems and other services.

The work programmes within the strategic priorities were developed in 2018 and are expected to remain aligned to the strategic priorities above until at least 2020 when they may be reviewed.

5a.ii. Use of HES and CSDS data within programmes and projects

Each programme is delivered through individual projects. Projects vary in length and complexity from major research studies which could span a number of years, to shorter term projects resulting in a single output, for example a briefing or presentation. Some Nuffield Trust projects are qualitative, involve analysis of published data, or data from surveys, and so do not require use of HES or other NHS Digital data sources. However, many others are reliant on HES and other NHS Digital-supplied data sets.

To undertake such projects the Nuffield Trust has determined that it requires HES data including the Emergency Care Data Set which will replace the HES A&E datasetand the Community Services Data Set which is now available and is a very welcome addition to the existing hospital based HES datasets.

Although the methods for use of HES and other patient datasets will vary from project to project, there are a number of common ways in which the data is used. This Agreement permits use of the data by the following methods:

• Assessing data quality, completeness, relevance and volumes of data prior to and during undertaking research analysis;

• Analysis to provide contextual information about NHS organisations or areas where research projects are being undertaken (in addition to analysis of relevant comparator organisations and areas), for example analysis of volumes of emergency admissions by specialty;

• Descriptive analysis of NHS activity and calculating age-sex standardised activity rates, for demographic or other patient cohorts, NHS organisations or administrative areas relevant to understanding NHS and government policies, and identifying gaps in policy;

• Using health care activity data to track changes in events such as emergency department attendances, admissions and re-admissions, time on caseload (for community patients). These are important (though imperfect) proxies for health outcomes and tracking trends in these events over time enables analysis of the impact of changes in health services;

• Analysis of health care resource use through applying NHS tariff or reference cost data to activity data derived from patient utilisation of services, and analysis of measures of capacity including bed occupancy from utilisation data;

• Undertaking detailed analyses of particular health events to identify particular issues with quality of care, for example as part of the harm project (see below), and patients with particular needs (eg frailty). This includes developing indicators of quality of care, covering access, effectiveness, continuity, coordination, safety or outcome;

• Examining variation between hospitals, patient groups or areas in use of services to identify populations where there are gaps in care, and also areas delivering high quality care from which the NHS can learn more widely using multivariate methods including standardisation, regression modelling and risk analysis;

• Analysis to determine the impact of specific service delivery models, such as the introduction of new pathways of care, or care settings, including acute admission units, same day emergency care, outpatient advice models, primary care networks and other models relevant to current NHS or government policy;

• Development and application of risk prediction models by methods including multivariate regression, cluster analysis, decision trees and machine learning for analysis to identify cohorts of patients with similar needs and to analyse variations between hospitals, patient groups or areas and for measuring the impact of service delivery models;

• Analysis to understand how wider health system and other factors impact on outcomes and activity, including differences between urban and rural service delivery or needs, impact of deprivation and variation in socio-demographic characteristics of the population and local factors such as education and social care provision and quality;

• Analysis to inform international comparisons of health data and quality, including replicating quality measures used in other health systems;

• Making evaluations of healthcare innovations more robust by using matched case-control analysis – comparing outcomes or trends in a service being evaluated with similar patients elsewhere;

• Evaluations of health care innovations using methods including time-series analysis, including interrupted time series, panel data and cohort studies;

• Use of the above methods in combination for particular research projects;

• Producing visualisations of analysis and results from the above methods.

The data will not be linked with other record level datasets and there will be no attempt to reidentify individuals from the data. The data may, if required, be linked with national datasets in the public domain (e.g. indices of social deprivation) subject to a risk assessment that the linkage will not increase the risk of reidentification of individuals within the dataset.

Should the Nuffield Trust wish to undertake a project involving a specific cohort of patients for which a data linkage is required, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

The number of concurrent projects using HES data will vary, but typically there are 5-6 projects in progress which use HES data at any one time, with perhaps 2-3 being completed in a calendar year. There may also be additional projects for which the analysis phase is complete, but work is ongoing on peer review publications and dissemination of the findings.

Projects are grouped within programmes, but frequently cut across other programmes. As the Nuffield Trust is a small organisation, it seeks to maximise impact by undertaking work which can inform more than one programme or strategic priority. Undertaking projects which support multiple programmes enables the Trust to deliver greater benefits to the health and care system.

5a.iii. How decisions are made about projects and use of HES and CSDS data

This Agreement permits the Nuffield Trust to use the data for the purposes of projects undertaken within the work programmes described above, and which are conceived, planned, approved and initiated through the following process:

1. Projects intended to meet the programme’s aims will be conceived and planned through an iterative process involving the Programme Director and Programme Lead with appropriate input from the Data Protection Officer (DPO). The Programme Director and Programme Lead will ensure that:

• Projects have a clearly defined objectives and operational plans;

• The aims of projects align with at least one of the programme’s aims (as stated above)- all projects must clearly and logically fall within the scope of having one or more of the aims listed and achieving those aims through one or more of the methods listed above ;

• An analysis plan is prepared for each project, setting out the data requirements and methods;

• In each case, the use of the HES data is necessary and proportionate to the purpose of the project and that the minimum amount of data necessary is used - this will include consideration of the necessity for use of each individual HES dataset; the number of years of data; the sizes of any cohorts or control cohorts derived from the data, and the inclusion and exclusion criteria (such as presence of specific diagnostic or procedure codes);

• Appropriate safeguards are in place to protect confidentiality; minimise risks of re-identification and use of excessive data beyond necessity.

A Data Protection Impact Assessment (DPIA) is completed at strategic level and covers all associated projects. A Legitimate Interest Assessment (LIA) will be completed internally for each research project and signed off by the Data Protection Officer (DPO).

2. A project management template will be completed and submitted to the Nuffield Trust Project Planning Committee (PPC). This excel template serves a wider purpose than just planning and remains valid throughout a project’s life, serving as the central control document in the management and delivery of the project. The PPC is chaired by the Director of Communications and consists of the Chief Executive, Director of Research, Director of Policy, Senior Fellow, Senior Policy Analyst and other representatives from Research, Policy and Communications. It provides a forum for the discussion, in depth and expert assessment and approval of project ideas, drawing on senior level expertise and knowledge across the Trust.

The committee is responsible for receiving assurance that all projects:

• Align to the strategic aims of the Trust;

• Are methodologically sound; and

• Draw fully on the expertise within the Trust including making connections to other related work.

The Committee will approve or recommend the approval of projects in line with the internally approved schedule of management authority and responsibility.

3. The individual or team within Nuffield Trust which will carry out the project will define and be bound by the analysis plan detailing what data is permitted for use in the project and how it shall be processed.

Project Timescales

The time frame for undertaking each project will vary according to project resource, extent of the research and data

analysis required. This is always considered as part of the Project Planning Committee's review and approved based upon

the detail of each individual project.

Project Funding

Some projects under the programme will be funded by the Nuffield Trust but funding is also being sought from other partners including National Research and Charitable organisations, such as the Nuffield Foundation, NIHR and the Health Foundation but may not be limited to these organisations. Funders will take both the forms of partners in collaborative working, as well as commissioners only, that is, the Nuffield Trust will be sole data controller for the analytical work. Though the Nuffield Trust may be commissioned by another organisation to undertake a project involving the processing of data under this Agreement, the Nuffield Trust will retain sole discretion for determining if and how the data would be used for any purpose. The Nuffield Trust will not be reliant on securing funding from external partners to complete this research. However, the Nuffield Trust will need to recognise the contribution of any external partner in their outputs.

Data Governance

The Nuffield Trust has independently determined the purposes for which it requires and will process the data under this Agreement in terms of its priorities and programmes which will use the data. The Nuffield Trust has sole autonomy for determining if and how the data will be used for projects in support of those priorities and programmes. As such, the Nuffield Trust is the primary data controller and the sole data controller named in this Agreement. In certain projects which involve collaboration with individuals or organisations outside of the Nuffield Trust, it may be the case that there is joint controllership for the specific project or aspects of it but in all cases, the Nuffield Trust will retain sole autonomy for determining if and how the data under this Agreement will be processed and the Trust cannot be compelled by any third party to process the data for any purpose of in any way. The data will only ever be used for purposes that directly support the priorities of the Nuffield Trust as described in this Agreement.

Under this Agreement, NHS Digital data will only be accessed by Nuffield Trust personnel (defined as employees, agents and contractors of the Trust) all of whom have been appropriately trained in data security and confidentiality. On occasion, the Nuffield Trust may invite individuals with significant or unique expertise to join the research team and contribute to data analysis. These individuals will either be seconded into the Nuffield Trust or will have an honorary contract with the Trust for the purpose and duration of a specific project or task within a project and as such will be considered agents of the Trust. These individuals would be subject to the same information governance framework as the Nuffield Trust employees and would be required to meet the level required to access the Nuffield Trust’s secure, ISO27001 certified data environment.

Should the Nuffield Trust wish to utilise an external organisation as its data processor, a separate application to NHS Digital and, subject to approval, a separate Data Sharing Agreement permitting the processing will be required.

Results may be shared in aggregate form in accordance with the Nuffield Trust's Research Governance framework, with small numbers suppressed. The data accessed through this Agreement will be managed by the Nuffield Trust, and will not be shared with any other third-parties.

The Nuffield Trust will produce an annual report for NHS Digital which will detail the outputs from all active and finished projects, which have been delivered during the year, and the planned outputs from new projects. The report will reference the associated strategic priorities and programme(s). Details will also be available on the Nuffield Trust’s website.

5a.iv. Examples of projects

The following examples of projects which have used HES data illustrate the range of work already undertaken in recent years, and upcoming projects. These are provided as examples of work the Nuffield Trust has undertaken or is or will be undertaking to meet the objectives of its programmes in support of its strategic priorities. The examples are not intended to form a comprehensive list of projects permitted under this Agreement.

Integrated care pioneers evaluation

Integrated Care 'Pioneers' are models of care aimed at reducing the impact of boundaries between health and social care providers. The evaluation of the pioneers has wider lessons for the current policy for integrated care systems.

• Nuffield Trust programmes: Commissioning and System Delivery; Evaluation; New models of care

• Overview: The Nuffield Trust is leading on one work package to develop and monitor a set of system level indicators, as part of a wider project with the DH Policy Innovation Research Unit based at the London School of Hygiene and Tropical Medicine. The analysis of HES data to develop indicators was solely the responsibility of the Nuffield Trust. The LSHTM does not have access to the data.

• Data minimisation approach: This project involves looking at time trends across a range of health and social care indicators. The Pioneers were introduced in 2013 but to have a good understanding of trends prior to this, HES data are used from 2004/05 with follow up until 2019/20, when that data is available. The indicators are presented in aggregated form in a dashboard with local authority district as the lowest geographical unit available. For the majority of indicators all ages are included but where possible indicators are restricted to a particular age group. Only variables relevant for each indicator were used.

• Duration: The project runs from 2016 to 2021

Medical Generalism

The rising numbers of older and more complex patients is one of the most pressing problems facing the NHS. Although they receive the most resource-intensive care, their problems are less likely to be accurately diagnosed and have more adverse outcomes than other age groups. The current models of hospital care, which are heavily based around specialists delivering disease-specific care, serve these patients poorly, as it is often fragmented and poorly co-ordinated. A revival of medical generalism has been suggested to provide better and more cost- effective care. The reality, however, is that there is a paucity of evidence on which to base new models of medical generalism.

• Nuffield trust programmes: Older people and complex care; Workforce

• Overview: The overarching aim of this NIHR funded research project was to identify the models of medical generalism used in smaller hospitals and explore their strengths and weaknesses from patient, professional and service perspectives. The Nuffield Trust used HES data to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals.

• Data minimisation approach: Hospital Episode Statistics data (year range 2007/08–2017/18) was used to create a classification of patients that might benefit from general medical care and, based on this classification, provide a descriptive analysis of the workloads of smaller hospitals. The final sample included 69 smaller NHS Trusts providing acute medical care in England, although some analyses used only 68 hospitals as a result of merges during the data period.

A data set based on ‘Index episodes of care’ for emergency admissions across the selected generalist medicine specialities identified in HES inpatient data 2012/13 for the smaller hospitals cohort was created so that five years of prior patient history for cancer patients as well as three years of subsequent history could be included. Cases with a specific diagnosis indicating specialist care or where patients had been transferred out of hospital were excluded. A data set was created covering 1.9 million episodes in the selected smaller hospitals.

Following development of the case mix classification, analysis focused on different data collection years dependent on the nature of the research question. For instance, the degree of alignment between patient case mix and medical generalist/skills mix in smaller hospitals was assessed was based on analysis of the 2015/16 data set. Population level analysis of admission patterns used data from 2017/18.

• Duration: The main research phase of the project was from 2016 to 2018.

This analysis was undertaken under a previous Data Sharing Agreement with NHS Digital in conjunction with University College London (UCL) which undertook parts of the analysis as a data processor under contract to the Nuffield Trust. Nuffield Trust defined the analysis and manner in which the data should be processed as well as the tools used (i.e. the software). The Trust also owns the outputs. The UCL Department of Applied Health research team carried out the above work on the Nuffield Trust site under the same Information Governance arrangements as Nuffield Trust staff - i.e. a signed confidentiality agreement as evidence that they had read and understood the Trust's Information Security Management System (ISMS), completed equivalent training and worked in the same environment. No data provided by NHS Digital left the Nuffield Trust site and the data was only processed on Nuffield Trust servers.

Evaluations of new services for patients outside of hospital

The Nuffield Trust has undertaken a number of evaluations of programmes to reduce admissions or readmissions to hospital, through better support for patients at home.

• Nuffield Trust programmes: Older people and complex care; New models of care; Evaluations

• Overview: This is a programme of work undertaken over a number of years. For example, the Nuffield Trust undertook a project funded by Age UK of services to support often vulnerable older people at high risk of hospital admission, which assessed the impact of this service on future emergency admissions. Previous projects were undertaken for a range of funders including the British Red Cross, Royal Voluntary Services and the Cabinet Office.

• Data minimisation approach: These projects took place using data linkage to HES data for specific cohorts of patients receiving innovative services. The Nuffield Trust used HES data from two years prior to the date each individual started receiving the service to allow for calculation of risk of emergency admission scores. To implement these models, the Nuffield Trust used a selected list of fields from inpatient, outpatient and A&E datasets (including admission method, diagnosis codes, procedure codes, A&E investigation codes, outpatient attended/did not attend). The Nuffield Trust compared each group of service recipients with a selected matched control group – matching one control person to each service recipient. The Nuffield Trust selected controls from a larger pool of possible control individuals. This larger pool of possible control individuals were selected to reflect the intervention cohort – eg they were the same ages (for example 55+ or 60+), and lived in similar areas (as defined by ONS and other analysis). Risk scores were calculated for all possible controls – involving again two prior years of HES data. For the evaluation outcomes, the intervention and matched control groups were followed up for a limited number of months (maximum 16, but more usually 6-9) on a limited number of activity measures.

• Duration: This was undertaken from 2013 to 2019

Comprehensive Geriatric Assessment (CGA)

The CGA was a multidisciplinary process designed to assess a frail older person’s medical conditions, mental health, functional capacity and social circumstances. However information is lacking on the types of patients that might benefit the most. The aim of this work is to describe existing models of care and to validate tools to deliver CGA on a hospital wide basis.

• Nuffield Trust programmes: Older people and complex care; Quality of Care; Evaluations

• Overview: This was a collaborative project undertaken as an NIHR funded project undertaken with the Universities of Leicester and Newcastle. The project incorporated linked HES, ONS and clinical data, managed under a specific Data Sharing Agreement (DARS-NIC-383324-D6B8T).

• Data minimisation approach: Three clinical cohorts were linked to HES as part of this project. The recruitment dates for these ranged from 2006 to 2012 and therefore, HES was required from 2004/05 to 2016/17 to allow a period g

A key aspect for all the research projects undertaken is ensuring that learning and research findings are disseminated widely, using press and television media, social media, conferences and practitioner networks. The Nuffield Trust aims to maximise the impact of its work, to ensure the greatest benefit to the health and care system, in line with their charitable purposes.

A communications plan is developed for each programme and project, based on the most effective way of securing impact for that project. Each strategic priority has a dedicated web page on the Nuffield Trust website, which provides an overview of why the topic is important, the overall approach, and links to programmes and projects related to that priority.

Outputs from a project could include:

• Nuffield Trust reports or briefings

• Blogs commenting on the findings

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

The Nuffield Trust will use their extensive communications facilities & networks for dissemination (including professionals in the fields of media relations, public affairs, digital communications and event management), working with their partner communications teams, to maximise the impact of findings.

The combination of outputs will vary from project to project.

For example, the Comprehensive Geriatric Assessment project, the outputs delivered or planned have so far included:

• A project report to the National Institute of Health Research that is waiting to be published (https://www.journalslibrary.nihr.ac.uk/programmes/hsdr/12500302/#/)

• A peer reviewed publication in the Lancet presenting the methodology and validation of the hospital frailty risk score (https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30668-8/fulltext)

• A toolkit in excel which is available at NHS hospital trust and local authority level, for local needs assessment and benchmarking (https://www.nuffieldtrust.org.uk/research/comprehensive-geriatric-assessment-needs-assessment-tool), that the Nuffield Trust has promoted through its twitter feed (https://twitter.com/NuffieldTrust/status/1034089904735768576, https://twitter.com/NuffieldTrust/status/1033356320915824640) and it has also been disseminated by the British Geriatrics Society (https://www.bgs.org.uk/resources/hospital-wide-comprehensive-geriatric-assessment)

• A guest blog from Professor Simon Conroy on the Hospital Frailty Risk Score discussing the advantages of being able to identify older people at risk in hospitals, and how it could make a real difference (https://www.nuffieldtrust.org.uk/news-item/the-hospital-frailty-risk-score)

• A BGS event for clinicians on frail older people which covered the work from the programme: http://www.acutemedicine.org.uk/wp-content/uploads/2018/04/BGS_Urgentcare_2018_v1-2.pdf

• European Geriatric Medicine Society Conference in October with two posters to present findings: http://www.eugms.org/2018.html

• Two further papers for peer reviewed journals are planned covering specific aspects of the findings.

• The project was a finalist in the ONS Research Excellence Awards 2018 (https://www.ons.gov.uk/aboutus/whatwedo/statistics/requestingstatistics/onsresearchexcellenceaward) which was a further opportunity to disseminate the findings and methods within the wider research community.

For the Quality Watch programme, outputs included:

• Over three hundred healthcare quality indicators on a dedicated website (now part of the Nuffield Trust website).

• 14 “Focus on” reports, two briefings, and several data blogs.

• Four annual reports that reviewed the state of care quality in the NHS in England and how it had changed over time.

• The Nuffield Trust provided a platform for internal and external expert commentary, with 135 editorial items (79 blogs (34 by external authors); 17 ‘latest data’ posts, covering monthly NHS combined performance summary data; 13 ‘indicator update’ posts, detailing stories emerging from ongoing data updates on the site; 26 news stories (mostly comprising press releases)).

• The Nuffield Trust also held a large number of events (QW conferences: October 2013, October 2014, November 2015; All Parliamentary Health Group events, February 2014 and December 2014; Social care event at Nuffield, May 2014; Allied Health Professionals event at Nuffield, November 2014; Public health roundtable at Nuffield, June 2016; Children and Young People roundtable at Nuffield, June 2017).

Previous outputs have also included:

• Integrated Care Pioneers: Outputs from the HES data analysis elements of the project are a system level dashboard to monitor indicators of integration in pioneer and non-pioneer areas, peer reviewed publications of this analysis, Nuffield Trust blogs/briefings on analysis challenges and research reports for the Department of Health.

• Medical Generalism: The project report has been submitted to NIHR for review and work is in progress on preparing papers for peer reviewed journals and other dissemination routes. A conference presentation has been accepted on the method used to develop patient pathways using HES data, at the Health Services Research UK conference in July 2019.

• Evaluations of new services for patients outside of hospital: Outputs from the project included interim and final reports for funders, Nuffield Trust blogs and other publications. For example, the Nuffield Trust produced reports for Age UK at different stages of the project, a Nuffield Trust report and blog, and a comment article for the Health Service Journal. A seminar on findings from these evaluations is planned for later in 2019.

• Comprehensive Geriatric Assessment (CGA) Outputs from the project are described in detail above. Further work on peer reviewed papers from the project is ongoing.

• Harms and Quality of care measures from routine data: Outputs from the project include a project report for NIHR and papers which have been submitted to peer reviewed journals.

• Prison Health: Outputs from the project include Nuffield trust reports, blogs and briefings, conference presentations and papers for peer reviewed journals.

• London Quality Standards (LQS): Internal reports to funders were produced, along with Nuffield Trust publications and blogs. Work is ongoing on papers for peer reviewed journals.

• Organisation of Primary Care: Outputs include a range of external publications and reports, and papers in peer reviewed journals.

In the past year outputs for the Nuffield Trust as a whole have included:

• 25 reports (in all cases with complementary blogs, charts or infographics)

• 12 briefings and explainers

• 90 blogs and long reads

• 10 charts and infographics (in addition to those in reports, blogs and briefings)

• 44 press releases

• Approximately 1200 Nuffield Trust tweets (The Trust has 45000 followers)

• 20 citations for NT staff in external peer reviewed journals

• 70 speaking engagements

• 10 corporate events

• 426,147 web site visits, averaging 1,674 per day across 610,413 sessions

• 100+ updated QualityWatch indicators and around 300 tweets

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES (or appropriate) Analysis Guide.

Yielded Benefits is not a requirement for new applications.