Hospital level aggregate NDA data for young adults (16-25) with diabetes in England

Diabetes care in the UK is among the best in Europe (eg UK is 4th out of 30 countries in the Euro Diabetes Index). However, good overall performance may mask poor clinical outcomes and patient experience within specific patient groups.

The TOGETHER study is a collaboration between clinicians and researchers at Queen Mary's University London, University College London and Imperial College London. This agreement relates to the national quantitative analysis aspect of the study only, for which Imperial College London is the lead, and therefore the sole data controller who will process data. More specifically, only Imperial College London will define the means and purpose for which the NHS Digital data will be used for this agreement, and will be the only organisation to access this data.

Previous work by Imperial College London has shown that mortality among young adults with diabetes in the UK is significantly worse than in other European countries and rose significantly between 1990 and 2010. This is consistent with other data Imperial College London have published showing that young adults report the worst NHS experience of any age group, and have distinct healthcare needs and priorities compared to other age groups. Recent advances in neuroscience and psychology also suggest that peer influences are likely to be particularly important for assessing risk and guiding

behaviour at this stage of the life-course. This is supported by evidence of successful introduction of a group clinic for young adults following renal transplant, which led to a significant reduction in the incidence of graft loss. However, recent systematic reviews have found that group clinics often had a positive effect on clinical and patient-reported outcomes but identified uncertainty about the mechanisms and pathway through which some groups were effective in improving outcomes.

Study aim

The overall aim of this NIHR-funded study is to explore the scope, feasibility and potential scalability of group clinics for young adults with diabetes and complex health and social care needs.

Study evaluation plan

As part of this study, Imperial College London are conducting a mixed-methods evaluation of the impact of group clinics on young adults' engagement with services, and their confidence and success in managing diabetes.

The main quantitative element of this evaluation includes detailed analyses of individual-level, longitudinal data from controls and participants in the group clinics: these data are collected through questionnaires and from hospital records and are not part of this application.

To complement the individual-level analyses, Imperial College London wish to analyse Hospital level data on young adults (16-25 years) with diabetes in the National Diabetes Audit. The legal basis for processing the NHS Digital data under

GDPR is Article 6 (1)(e) - "processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller" and Article 9(2)(j) - "processing is necessary for archiving purposes in the public

interest, scientific or historical research purposes or statistical purposes".

There are three main objectives in doing these additional analyses.

1. To investigate how the clinic population in this study compares to other services across the country (for example, in terms of age, sex, deprivation, and ethnicity).

2. To investigate how this study compares to other services on the three treatment targets and eight care processes included in the National Diabetes Audit.

3. To study change over time in performance on the three treatment targets and eight care processes in the study sites, and how these changes compare to changes in other services across the country.

Depending on the findings of this evaluation, Imperial College London anticipate applying for funding to perform a randomised controlled trial of group clinics for young adults with diabetes.

Imperial College London requires National Diabetes Audit data for use as part of the NIHR-funded TOGETHER study to study group clinics for young adults with diabetes.

Papoutsi,C, Hargreaves DS, Colligan G, et al. Group clinics for young adults with diabetes in an ethnically diverse, socioeconomically deprived setting (TOGETHER study): protocol for a realist review, co-design and mixed methods, participatory evaluation of a new care model. BMJ Open. 2017;7(6).

Role of Imperial and other organisations

The TOGETHER study is a collaboration between clinicians and researchers at Queen Mary's University London, UCL and Imperial College London. This agreement relates to the national quantitative analysis aspect of the study only, for which Imperial College London is the lead, and therefore the sole data controller who will process data. More specifically, only Imperial College London will define the means and purpose for which the NHS Digital data will be used for this agreement, and will be the only organisation to access this data.

This agreement solely relates to quantitative analyses at Imperial College London. As part of the wider TOGETHER study, some qualitative research is being done at Oxford University, but this does not involve any application to NHS Digital or any other body and Imperial College London is not involved in collection, storage, or analysis of any qualitative data.

No record level data are requested. Hospital level data (aggregated with small numbers unsuppressed) from the National Diabetes Audit will only be accessed by authorised members of the research team, who are substantive employees of Imperial College London.

Data access is strictly controlled by the Big Data and Analytical Unit (BDAU) at Imperial College London, through a robust dataset registration process. No one other than BDAU staff can authorise access to the data.

Access to the data will be only for the purpose outlined in this Data Sharing Agreement, all staff are bound to the policies, procedures and equivalent controls of the BDAU Secure Environment (SE) and Imperial College London, as substantive employees of the College.

The National Diabetes Audit (NDA) is part of the National Clinical Audit Programme.

NHS England have commissioned NHS Digital to deliver the NDA Programme for a further 3 years covering 2017-2020

NDA is held by NHS Digital under controllership and no additional data is received and disseminated.

Many units have a relatively small number of patients aged 16-25 and diabetes complications such as eye and kidney problems are relatively uncommon among younger patients. In combination, these two factors mean that small number suppression would effectively exclude a significant number of services from the analyses and would prevent meeting the objective of these analyses, which is to place the local findings in a national context.

The raw data provided by NHS Digital will be analysed solely in the BDAU SE. Any further analysis done outside the BDAU SE (usually for visualisation purposes for output) will be done using data that has been aggregated with small numbers suppressed in line with the HES Analysis Guide. The data will be analysed to examine how variation in scheme design affects performance on the incentivised dimensions.

Summary of data processing for desired outputs

Step 1: Data will be received, stored and analysed within the Imperial Big Data and Analytical Unit (BDAU)

Step 2: Analyses of aggregate, Hospital level data will be performed. No individual level data will be requested or analysed.

Step 3: Three sets of analyses will be performed on Hospital level socio-demographic characteristics, audit performance, and change in audit performance over time (corresponding to the three objectives above).

- Descriptive analyses of how the sociodemographic profile of the clinic population in Imperial College London's study sites compares to other services across the country (with regard to sex, age, deprivation, ethnicity)

- Comparison of Hospital level audit performance for young adults on the three treatment targets and eight care processes in Imperial College London study sites compared to other services across the country. This will involve descriptive analyses of mean values/proportion of patients receiving each care process.

- Analyses of change in Hospital level audit performance in Imperial College London study sites between 2016/17 and 2017/18 on the three treatment targets and eight care processes, followed by comparison of changes in Imperial College London study sites to changes in the whole population. Differences will be assessed using t tests and chi-squared tests.

Please see the published study protocol for further details of Hospital level analyses and how these will link to other strands of the study evaluation (individual-level analyses of study participants and controls, health economic analyses and qualitative work). This is available to the public as an open access article in British Medical Journal (BMJ) Open.

There will be no linkage with other record level data and Imperial College London will make no attempt to re-identify any individual in the data provided.

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by Personnel (as defined within the Data Sharing Framework Contract ie: employees, agents and contractors of the Data Recipient who may have access to that data).

The use of any cloud based solution for data storage is not permitted under this agreement. Any changes must be reflected through an amendment and subsequent approval of the agreement.

Specific outputs of Hospital level analyses of National Diabetes Audit (NDA) data

1. Final report to NIHR (due December 2019)

2. Peer-reviewed journal article of quantitative evaluation of group clinics (planned submission date December 2019/January 2020)

3. Since Spring 2017, Imperial College London have submitted regular updates about progress in applying for NDA data to NIHR, and continue to do so.

Outputs from this project to date include

1. Papoutsi,C, Hargreaves DS, Colligan G, et al. Group clinics for young adults with diabetes in an ethnically diverse, socioeconomically deprived setting (TOGETHER study): protocol for a realist review, co-design and mixed methods, participatory evaluation of a new care model. BMJ Open. 2017;7(6).

2. A realist review of group clinics for young people with long-term conditions is currently under review.

3. This work has been presented at 3 conferences (focusing on diabetes and/or health services research).

4. Preliminary findings were presented to project partners, commissioners and national policy makers at a celebration event in London in November 2018.

Overall dissemination plans and outputs for the TOGETHER study

Findings from this analyses of NDA data will be an important component of the final project findings and be included in all major outputs from this project.

Imperial College London will publish briefing statements, summaries and academic papers to disseminate the findings of this study to a range groups with roles in service use, service delivery, policy making and health service design. These will include:

Local networks:

• Institution-based: e.g. Imperial College London, Queeny Mary University of London,

• Newham CCG have agreed to support dissemination activities, via their Patient and Public Engagement Officer, and their existing network of Diabetes Champions and community neighbourhoods

• Local Transforming Services Together programme - Imperial College London will regularly update this commissioning group which is signed up to by four inner London CCGs and has prioritised YAs.

National organisations

• One of the researchers attends the National Children and Young Adult Working groups chaired by the National Leads for Children and Teenagers and Young Adults and will report to them regularly on the project.

• The study team will engage the Association of British Clinical Diabetologists and Diabetes UK in the national scoping exercise in this study and will plan to disseminate and scale up the study via their key contacts.

• The study team will engage the Young Adult and Adolescent Steering Group of the Royal College of Physicians, as well as other relevant committees and patient groups at the Royal Colleges of Paediatrics and Child Health, and General Practice.

•The study team will also engage with Improvement Science London toshare results across wider academic partnerships and cross-sector groups e.g. NHS England, Greater London Authority

• This study has been adopted onto the North Thames Cancer Research Network portfolio and hope to use their existing structures to disseminate output to their members and lay committees. Imperial College London also have the support of the North Thames CLAHRC (Collaborations for Leadership in Applied Health Research and Care).

International

• Social networks: Imperial College London will use twitter groups of people living with diabetes for the journal Diabetic Medicine, to disseminate the findings.

• Academic: Imperial College London plan to present and publish their work to an international audience interested in diabetes, pregnancy, young adults with long term conditions, new models of care, e.g. via the NIHR Health Services and Delivery Research journal, conference

If this trial were to prove successful impact of group clinics on diabetes outcomes, then Imperial College London could expect to see significant improvement in diabetes outcomes for young adults in the NHS within the next 5-10 years.

Analyses of Hospital level National Diabetes Audit data will be an important part of the overall findings - both as key research findings in themselves, and also in supporting interpretation of findings from the qualitative work and analyses of clinical records of study participants (not covered by this application).

The expected measurable benefits of the study overall are listed below

• Assessment of current needs of service users and stakeholders and review of evidence and current use of group clinics and non-traditional care models in diabetes will be performed. The outcome of this realist review and scoping process will, (i) identify baseline needs of local users and stakeholders and, (ii) identify the potential need for new care models in the wider UK population of people living with diabetes.

• The outcome of the co-designed group clinic model will be evaluated against the complex health and social care needs of the participating young adults with diabetes in Newham. Qualitative assessment using individual semi-structured interviews of all stakeholders will derive outcomes on

(i) acceptability/feasibility of delivering group clinics including impact on staff offering them, training needs, resource implications,

(ii) active ingredients/key mechanisms in the intervention,

(iii) contextual factors relevant to successful implementation at macro/meso/micro level,

(iv) sustainability of successful care delivery through group clinics. Quantitative assessment, using national benchmarking and audit data, will be performed from the new care model and compared to current care at Newham, two comparable external units, and national audit data (NDA).

• Mechanisms by which the co-designed group clinics work will be investigated by analysis of differences in patient experience and behaviour before implementation, during iterative revision, and after implementation of the new care models.

• Economic evaluation will identify outcomes related to, (i) costs of this new care model from the health care provider perspective, (ii) comparative costs between new and standard pathways, including direct and substitution costs, (iii) incremental cost-effectiveness, (iv) cost/benefit from the perspective of patient users.

• Outcomes from the qualitative analysis and economic evaluation will have wider impact via guiding commissioners and policy members in the future development of group-based diabetes care. Information from local and national scoping will guide the scaling up and generalisation of this new care model to other geographic locations and long-term conditions.

• The outcomes from this study will also provide a platform for a future cluster-randomised controlled study of integrated group clinics in diabetes, their effect on quantitative clinical outcomes, cost effectiveness and generalisability across different locations.

Specific points about the benefit of findings from NDA data

- Previous work has suggested that group clinics can reduce costs and can improve patient experience and engagement with care. Imperial College London know from their engagement work with commissioners and others that many commissioners and professionals are interested in introducing group clinic models for young people with diabetes and other long-term conditions. However, one important barrier to doing so is concerns that clinical outcomes may deteriorate where such clinics are introduced, particularly for patients that do not attend the groups.

- Ethics protocols for the study prevent individual-level analyses of clinical records on patients who decline participation in the TOGETHER study. Analyses of Hospital level data from NDA will allow Imperial College London to investigate whether any clinic-level change is seen after introduction of group clinic models. Data on all Hospitals will allow Imperial College London to use a sophisticated difference-in-difference analyses rather than simple before-and-after analyses.

Not stated in the register.