Hospitalisation and Mortality after Acute Myocardial Infarction
University of Leeds · Academic
Expired The latest version ended on 31 December 2024. The September 2026 register still lists the agreement, but its term has passed.
- Reference
- DARS-NIC-17649-G0X4B
- Latest version
- v3.2
- Term of latest version
- 13 February 2023 to 31 December 2024
- Start date
- 13 February 2017
- Data controller
- Sole Data Controller
- Commercial purposes
- No
- Sublicensing
- No
- Files released to date
- 9
Why the data was released
Objective for processing
On 1 February 2023, NHS Digital merged with NHS England. NHS England has assumed responsibility for all activities previously undertaken by NHS Digital. The merger was completed by a statute change. Any reference made to NHS Digital within this Data Sharing Agreement is in reference to the merged organisation known as NHS England.
No further data will flow under this version of the agreement.
The objective for processing HES data is to perform research into survival following heart attack in England.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease.
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving Acute Myocardial Infarction (AMI) in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
As a publicly funded organisation, the University of Leeds processes personal data to undertake scientific research which is in the public interest. Therefore, the legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
The justification for processing the data by University of Leeds is Article 6(1)(e) of the GDPR: (processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller).
The justification for the processing of the special category data (health data) by University of Leeds is Article 9(2)(j) of the GDPR: (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.
Over the last decade, there has been a substantial and sustained decline in mortality rates from cardiovascular disease in the UK. Despite this, cardiovascular disease remains the biggest killer in the UK and someone is admitted to an NHS hospital with a heart attack every three minutes. Moreover, improvements in acute myocardial infarction (AMI; heart attacks) survival are likely to be a major cause for the increasing incidence of heart failure (‘transferred morbidity’), which now affects around 900,000 individuals in the UK and accounts for 5% of all emergency hospitalisations. Presently, most patients with cardiovascular disease are elderly and because AMI survival has increased there are more patients living longer with co-morbidities. More frequently, such patients require specialist cardiovascular care in the form of invasive cardiac procedures including high and low voltage and resynchronization pacemakers and coronary revascularisation. Moreover, they frequently re-present to hospital – escalating the burden of admissions with heart failure.
Specifically, the research will aim to quantify the long-term outcomes and hospitalisation rates for survivors of acute myocardial infarction in England.
The objectives of the analysis are:
1. To describe hospitalisation patterns and endpoints (heart failure, cerebrovascular disease, coronary revascularisation, vascular dementia, severe bleeding, acute myocardial infarction, atrial fibrillation, all-cause mortality) for patients hospitalised with non-fatal AMI (i.e., survivors of the index hospital stay) compared to those who have no recorded AMI.
2. To identify factors associated with hospitalisation and endpoints for hospital survivors of index AMI compared to those who have no recorded AMI specifically focusing on geographical variation and the provision of timely percutaneous coronary intervention.
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease for example) following survival from AMI, the research group at the University of Leeds need to ensure they have a clean cohort for analysis to minimise confounding where possible as well as data of the hospitalisations occurring among patients who have no recorded AMI. Detailed justification for the request of this level of data is outlined below.
1. Reasons for requiring hospitalisations amongst patients with no recorded AMI
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease, heart failure and other outcomes) following survival from AMI, the research group needs to compare the number of each hospitalisation type occurring amongst AMI patients to the number of each hospitalisation type which occur in the background population (in this case, the background population is the population of patients admitted to hospital without an AMI in the same study period).
The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the research group will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
The research group will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations.
The research group considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population.
A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population-based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e., all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the research group.
The research group would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
Finally, the research group aim to determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
2. Ensuring derivation of a ‘clean’ cohort
The research group cannot be sure that patients admitted to hospital with AMI in a given period of HES data have not had a previous AMI, or, whether they have had any of the conditions they are considering as outcomes, such as cerebrovascular disease prior to their AMI. Therefore, the research group proposes to derive the cohort for analysis from admissions from 2008/09 to 2016/17, however, has requested for NHS digital to exclude all people from the data who have had a previous AMI or any of the hospitalisations the research group are considering as outcomes. The same filtering of prior conditions will be done by NHS digital for patients who have not had an AMI from 2008/09 onwards. This is a substantial minimisation effort by the research group, as without this step, data from 200-1/02 would have been required as part of the data application. Data from 2008/09 onwards will give sufficient data to look at time trends in hospitalisations and mortality as part of the analysis.
Details of all hospital attendances (not restricted to specific conditions with known associations with AMI) are required in order to understand the history of the patient and whether past (non-related) attendances have contributed in any way to that AMI attendance, or to any of the study outcomes including heart and non-heart related outcomes for patients in the AMI or non-AMI cohort. Post attendances also supports this (whether AMI contribute to non-heart related attendances).
Additionally, for each individual NHS Digital will provide a vital status indicator (alive/deceased) and, where individuals are deceased, the number of days between the data of admission and the date of death. The date of admission will not be supplied to the University of Leeds making it impossible for the research group to derive the date of death from the data supplied.
The proposed work to study the hospitalization patterns and outcomes for patients with acute myocardial infarction is part of a larger programme of work funded by the British Heart Foundation (Project Grant PG/13/81/30474) in order to fill an important knowledge gap of the long term hospital burden and non-fatal outcomes for patients with AMI using contemporary, large scale and national observational data. The British Heart Foundation Project Grant was entitled: “Cumulative Missed Opportunities for Care after Acute Myocardial Infarction: a linked national cardiovascular registries cohort study to identify preventable deaths“ and has now ended. NHS Digital data in relation to this DSA will not be used for any other programmes of work except those outlined in the specific objectives within this agreement.
DATA MINIMISATION:
Under GDPR, University of Leeds are adhering to the standard of data minimisation ensuring the data is adequate, relevant and limited to what is necessary for the stated purposes of processing. This has been achieved in the following ways:
1) Data is pseudonymised;
2) Data is filtered to exclude patients with AMI prior to 2008/09 and exclude any patients with any of the clinical conditions considered as outcomes (see section 2 ‘Ensuring derivation of a ‘clean’ cohort’ of the Objective for processing part of the application for full details of filtering)
3) Data fields are minimised by aggregating all date fields to Month and Year format
4) Date of death and date of birth are minimised to number of days survived with a mortality flag and age respectively.
University of Leeds are both the Data Controller and Data Processor for this study. There are no further data processors, or third parties, involved in the analysis of the data. Microsoft Limited supply provide Cloud Services for the University of Leeds and are therefore listed as a data processor. They supply support to the system, but do not access data.
The British Heart Foundation (Project Grant PG/13/81/30474) which was used to fund the initial data application has now ended. The cardiovascular epidemiology group are continuing to analyse NHS Digital data for the specific objectives and purpose outlined within this application to ensure the measurable benefits outlined within the application can be realised. Funding to support ongoing research has been obtained from the Wellcome Trust (ref 206470/Z/17/Z).
For the avoidance of doubt, the British Heart Foundation and the Wellcome Trust will not influence the results or dissemination of the research conducted, and the British Heart Foundation and the Wellcome Trust will have no role in the design, analysis or interpretation of the research.
Data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Processing activities
No further data will be disseminated under this version of the agreement.
Pseudonymised record level data was sent by NHS Digital to the University of Leeds under v0 of this agreement, for the purposes outlined within this application. There were no subsequent flows of data.
Processing and cleaning of data are conducted in SQL Server and statistical analyses are conducted in specialist statistical software including Stata and R. Only aggregated data with small number suppression applied in line with the HES Analysis guide are moved outside of Microsoft Azure for presentation and publication purposes. The data will only be accessible to authorized individuals in the Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds via password restricted computer access.
The data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project. All individuals accessing these data have undertaken the advanced IT security training provided by the University of Leeds and will do so on an annual basis for the duration of the project.
The data will be geocoded based on Lower Super Output Area (LSOA) to obtain information on higher aggregated geographical units (Clinical Commissioning Groups).
No further linkage to the data will occur. Only summarised and aggregated data will be disseminated in the form of academic presentations, peer-reviewed journals and lay summaries (in line with the HES analysis guide)
The data will not be used for commercial purposes, provided in record level form to any third party or used for any direct marketing. There will be no requirement or attempt to re-identify any individuals within the data.
The study was funded by the British Heart Foundation. The British Heart Foundation did not influence the results or dissemination of the research conducted, and the British Heart Foundation had no role in the design, analysis or interpretation of the research.
Microsoft Limited supply Cloud Services for the University of Leeds and are therefore listed as a data processor. They supply support to the system, but do not access data. Therefore, any access to the data held under this agreement would be considered a breach of the agreement.
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.
Data will only be accessed and processed by substantive employees of the University of Leeds and will not be accessed or processed by any other third parties not mentioned in this agreement.
Expected output
The Cardiovascular Epidemiology research group dissemination strategy has remained the same as outlined within the original agreement with NHS Digital with the exception of the target dates.
Since the last request for an extension to the DSA, The Cardiovascular Epidemiology research group have made significant progress in the analyses of this extremely large extract of HES data for our advanced analytical project as outlined within the original application.
The Cardiovascular Epidemiology research group now have two draft publications undergoing final internal review and refinements with our project collaborators. The analytical methods required to ensure the highest scientific rigour is achieved with these data were more advanced than initially thought - challenging methods such as risk-set matching requiring extensive effort to achieve in data of these size were required; and as such The Cardiovascular Epidemiology research group are requiring further extension to ensure The Cardiovascular Epidemiology research group can meet the intended outputs and benefits of the project.
The expected outputs are as follows:
1) Peer-reviewed publications
Objective 1: Hospitalisation and mortality after acute myocardial infarction.
Due to the significant amount of information derived as part of this large scale project - the first objective will now be met through two complimentary but distinct publications assessing 1) the absolute and relative risks of each individual hospitalisation outcome following AMI and 2) assessing the disease trajectories and sequences of multiple hospitalisations following AMI. These publications will each achieve the originally planned overarching objective to determine the long term hospitalisations and outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity. These two papers have been progressing in parallel - and the decision to split the work was based upon the significant and distinct methods required for each aspect of this analyses.
Publication 1: Health outcomes after myocardial infarction: an entire population study of adults hospitalised in England. Progress: Full draft paper undergoing collaborator review and finalising of methods, presentation and wording. Anticipated submission for initial peer review: December 2022. This paper includes a thorough epidemiological assessment of the absolute and relative risk of post AMI hospitalisations for each individual outcomes compared with the non-AMI population.
Publication 2: Post myocardial infarction disease trajectories of 145 million hospitalised episodes. Progress: Full draft paper undergoing collaborator review and finalising of methods, presentation and wording. Anticipated submission for initial peer review: January 2023. This paper includes a thorough assessment of the trajectories of hospitalisation following AMI, assessing in detail the possible pathways of disease progression between the multiple outcomes to present a complete picture of post-AMI hospitalisations compared with non-AMI patients.
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: July 2023-December 2023.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners. We have allocated additional researcher time in order to complete this objective with new timelines. Progress: Draft coding scripts for the analyses of post AMI hospitalisations by geographical area have been generated and preliminary analyses is underway.
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of percutaneous coronary intervention (PCI). Anticipated submission date: December 2023. This paper will specifically look at the association between receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2) Wider academic dissemination of the research findings will be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2023/2024) and the European Society of Cardiology Congress (August 2023/2024). We have opted to not share preliminary findings from this work at conferences at an earlier stage as originally planned. The size and complexity of this study has meant that we do not wish to have the outputs in the public domain until it has undergone scientific peer-review to the highest standards (i.e. through external peer-review processes) - which has delayed our original dissemination strategy.
3) Lay summaries of the research findings will be generated and disseminated to the following key stakeholders:
- The British Cardiovascular Society (BCS),
- British Heart Foundation (BHF), TakeHeart,
- NHS commissioners and clinicians/health professionals involved in managing heart attack.
The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community.
In addition, the group website (Cardiovascular Epidemiology - https://lida.leeds.ac.uk/research-projects/566-2/) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project at appropriate stages.
Research referenced in NICE guidelines:
1) Clinical risk scoring for acute myocardial infarction referenced in NICE Clinical Guideline 94 (Gale CP et al, Heart, 2008; 95(3):221-227)
2) Atrial fibrillation research referenced in Atrial fibrillation: management, NICE clinical guideline 180 (Cowan, Long and Gale et al. Heart (2013): heartjnl-2012)
3) Pre-hospital ECG research referenced in European Resuscitation Council Guidelines for Resuscitation 2015 (Quinn, and Gale, et al. Heart (2014): heartjnl-2013).
Widespread media coverage and beyond:
1) Research by Dondo and Hall et al (Dondo T, Hall M, et al. Eur Heart J Acute Cardiovasc Care. 2016) has received widespread media coverage including radio and TV broadcasts as well as broadsheets and tabloids. This research also resulted in being invited to present at the Westminster health forum and forms the critical evidence for the non-ST elevation myocardial infarction NICE Implementation Collaborative.
2) Research by Hall and Gale et al (JAMA 2016; Aug 30. doi: 10.1001/jama.2016.10766) as well as Wu, Hall and Gale et al (Eur Heart J Acute Cardiovasc Care 2016; Aug 29. pii: 2048872616661693) forms the evidence for the guidelines in practice NIC project (https://www.guidelinesinpractice.co.uk/nic-projects).
3) The geographic variation in AMI treatment by Dondo, Hall and Gale et al (BMJ Open 2016; 6 (7): e011600) has received widespread media coverage and has led to a successful Department of Health / NHS England business case to develop the work further into a feedback quality improvement programme for patients, hospitals and commissioners (work currently ongoing).
All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.
Expected measurable benefits
This study will quantify the burden of hospitalisations and long-term outcomes for patients who are admitted to NHS hospitals and surviving acute myocardial infarction (AMI) in England. The research outputs as described will be disseminated widely to University of Leeds' established informal networks including the academic community, clinicians, patients and the public as well as NHS commissioners via the formal networks discussed above (outputs).
Dissemination of the factors which could lead to increased hospitalisations and mortality to clinicians (via academic publications, presentations at clinical conferences and dissemination via the British Cardiac Society and British Heart Foundation) is envisaged to be a driver for improved patient care and has far reaching clinical and social benefits as outlined below.
Quantifying the burden of hospitalisations and long-term outcomes for patients surviving AMI in England will for the first time, on a national scale, provide NHS commissioners with the necessary evidence to plan effectively for service provision and resource allocation for the large proportion of patients who now survive their AMI. Although improvements in treatment have resulted in improved survival rates for patients with AMI – the long-term health burden of patients following their AMI is not yet known – and this is what the research group propose to determine. In addition, the findings from this study can be used to inform new endpoints for future clinical trials, to ensure that not only the mortality or short-term cardiovascular outcomes of AMI patients are considered, but also longer-term cardiovascular and non-cardiovascular outcomes in developing and testing new treatments in future.
In addition, by quantifying the impact of lack of adherence to guideline recommended care on re-hospitalisation and mortality or the impact of delayed PCI treatment on re-hospitalisation and mortality, the Cardiovascular Epidemiology research group would provide the scientific supporting evidence to clinicians to strive for improved adherence to guidelines, which therefore has the potential to improve outcomes for patients.
Increased patient awareness of the impact of AMI on future hospitalisation and mortality, through dissemination to the public as described, could not only lead to patients modifying their own health behaviours to minimise their own risks of future hospitalisation, but patient and public who are informed by this knowledge can also provide strong motivation for clinicians and commissioners to improve patient care and care pathways.
Knowledge of geographical variation in hospitalisation and mortality from AMI will identify key areas of inequality in the NHS, dissemination of this knowledge to NHS commissioners will enable them to act upon such inequalities to ultimately drive-up standards and provide direct and measurable benefit to patients and the NHS. This work forms part of the wider research conducted by the University of Leeds' research team (Cardiovascular Epidemiology), and will therefore contribute to a growing body of evidence regarding the quality of care and outcomes for patients surviving AMI, whilst adding important new insights into the long term health burden for the increasing number of AMI survivors which focusses not only on mortality, but importantly, also on re-hospitalisation for a range of cardiovascular and non-cardiovascular conditions.
Several members of the Cardiovascular Epidemiology research group have experience with Public and Patient Involvement, meeting patients to discuss a range of different research proposal and allowing influence and participation in the research agenda as well as access to disseminate research findings, whilst some members of the research group hold very close involvement with the British Heart Foundation, especially their press office and policy group who therefore act as a powerful conduit for change and knowledge dissemination.
The research the Cardiovascular Epidemiology research group propose here is of direct and critical importance to the NHS and the Department of Health. Although heart attacks remain the biggest killer worldwide, survival rates are improving. As such, patients are living longer with their cardiovascular disease, and there are an estimated 7 million people living with cardiovascular disease in the UK. The Cardiovascular Epidemiology research group propose to look at the components of the disease process including the wide range of outcomes following acute myocardial infarctions so that the commissioners, hospitals and NHS England can make evidence informed policy decisions about the need for cardiovascular care, as well as where, when and in whom.
Whilst exact estimates of the numbers of people in the UK who have survived AMI are difficult to estimate – a report in 2015 estimated the prevalence of AMI in the UK to be around 915,000 people (Bhatnagar P et al. Heart 2015;101:1182–9. https://doi.org/10.1136/heartjnl-2015-307516). Knowledge of the hospitalisation patterns and endpoints following AMI will lead to better service provision planning which can improve the quality of health care for more than 900,000 people in the UK who have suffered AMI, but also for people who may experience AMI in future as well as their families and carers.
Benefits reported so far
The sheer size of these data (145 million hospitalisation records) and complexity of analyses have meant the originally targeted timeframes were not met, and as such, the benefits outlined above have not yet been realised.
The Cardiovascular Epidemiology research group have made significant progress with the analyses of these data since the last DSA extension, and continue to make this work a primary focus of the research group. We now have two publications at a final draft stage, currently undergoing collaborator review and final refinements before submission to journals according to the updated timelines listed in the outputs section of this application.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the results now emerging to answer the proposed research questions will not be lost.
Datasets on the latest version
Legal basis for provision: Health and Social Care Act 2012 – s261(2)(a)
| Dataset | Type of data | Sensitivity | Frequency | Confidential data |
|---|---|---|---|---|
| Hospital Episode Statistics Admitted Patient Care (HES APC) | Anonymised - ICO Code Compliant | Non-Sensitive | One-Off | Does not include the flow of confidential data |
Files released
Files released counts only files released externally by DARS. Access granted in NHS England's own systems, such as its Secure Data Environment, is not included.
Patient opt-outs were not applied to any of the 9 files released under this agreement, across every version. About opt-outs
No files recorded as released under the latest version. 9 were released under earlier versions, shown in the version history.
Version history
The register lists each renewal of this agreement as a separate row. This site has 4 versions.
DARS-NIC-17649-G0X4B-v3.2 13 February 2023 to 31 December 2024
- Title
- Hospitalisation and Mortality after Acute Myocardial Infarction
- Commercial
- No
- Sublicensing
- No
- Datasets
- 1
- Files released
- 0
Datasets: Hospital Episode Statistics Admitted Patient Care (HES APC)
What changed from DARS-NIC-17649-G0X4B-v2.4
Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.
| Field | Was | Became |
|---|---|---|
| Start date | 2023-02-13 | |
| End date | 2024-12-31 | |
| Hospital Episode Statistics Admitted Patient Care (HES APC): legal basis | Health and Social Care Act 2012 – s261(2)(a) |
Objective for processing
On 1 February 2023, NHS Digital merged with NHS England. NHS England has assumed responsibility for all activities previously undertaken by NHS Digital. The merger was completed by a statute change. Any reference made to NHS Digital within this Data Sharing Agreement is in reference to the merged organisation known as NHS England.
No further data will flow under this version of the agreement.
[3 paragraphs unchanged]
The University of Leeds is an independent corporation established by Royal Charter.
[40 words unchanged]
quantify the burden of hospitalisations and long term outcomes for patients surviving
AMI
Acute Myocardial Infarction (AMI)
in England, thereby providing NHS commissioners, clinicians and patients with the necessary
[20 words unchanged]
of health care, for the large number of patients who have AMI.
[1 paragraph unchanged]
The justification for processing the data by University of Leeds is Article 6(1)(e) of the GDPR: (processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller).
The justification for the processing of the special category data (health data) by University of Leeds is Article 9(2)(j) of the GDPR: (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.
[29 paragraphs unchanged]
Processing activities
No further data will be disseminated under this version of the agreement.
[1 paragraph unchanged]
No new data will be disseminated under this agreement.
[7 paragraphs unchanged]
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.
Data will only be accessed and processed by substantive employees of the University of Leeds and will not be accessed or processed by any other third parties not mentioned in this agreement.
Expected output
[1 paragraph unchanged]
There are several reasons which have factored into delays in producing the planned outputs according to the originally planned timeframes which include:
Since the last request for an extension to the DSA, The Cardiovascular Epidemiology research group have made significant progress in the analyses of this extremely large extract of HES data for our advanced analytical project as outlined within the original application.
1) The application took a significant amount of time to complete which meant there were other projects competing for research staff time at the time of data receipt, and subsequently, there was a period of maternity leave for the primary researcher for the project.
The Cardiovascular Epidemiology research group now have two draft publications undergoing final internal review and refinements with our project collaborators. The analytical methods required to ensure the highest scientific rigour is achieved with these data were more advanced than initially thought - challenging methods such as risk-set matching requiring extensive effort to achieve in data of these size were required; and as such The Cardiovascular Epidemiology research group are requiring further extension to ensure The Cardiovascular Epidemiology research group can meet the intended outputs and benefits of the project.
2) The data extract is incredibly large (145 million hospitalisation episodes), which has meant loading, formatting, processing and cleaning of the data have taken a significantly longer period of time to complete than originally anticipated. The University of Leeds now have a research ready dataset, with an extensive set of preliminary analyses and are on target for the following extended target dates for the specific outputs, as outlined in the original application.
3) Further delays have occurred due to the Covid 19 pandemic. The primary researcher for the project was placed on full time furlough from April 2020 to September 2020. Full time work on the research recommenced as of October 2020, however the primary researcher has since left temporarily on maternity leave. Timelines have been further updated to reflect these additional delays.
[2 paragraphs unchanged]
Paper
Objective
1: Hospitalisation and mortality after acute myocardial infarction.
Anticipated submission for initial peer review: October 2021 – February 2022.
This paper will quantify the hospitalisations and long term outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity.
Due to the significant amount of information derived as part of this large scale project - the first objective will now be met through two complimentary but distinct publications assessing 1) the absolute and relative risks of each individual hospitalisation outcome following AMI and 2) assessing the disease trajectories and sequences of multiple hospitalisations following AMI. These publications will each achieve the originally planned overarching objective to determine the long term hospitalisations and outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity. These two papers have been progressing in parallel - and the decision to split the work was based upon the significant and distinct methods required for each aspect of this analyses.
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August 2022.
Publication 1: Health outcomes after myocardial infarction: an entire population study of adults hospitalised in England. Progress: Full draft paper undergoing collaborator review and finalising of methods, presentation and wording. Anticipated submission for initial peer review: December 2022. This paper includes a thorough epidemiological assessment of the absolute and relative risk of post AMI hospitalisations for each individual outcomes compared with the non-AMI population.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners.
Publication 2: Post myocardial infarction disease trajectories of 145 million hospitalised episodes. Progress: Full draft paper undergoing collaborator review and finalising of methods, presentation and wording. Anticipated submission for initial peer review: January 2023. This paper includes a thorough assessment of the trajectories of hospitalisation following AMI, assessing in detail the possible pathways of disease progression between the multiple outcomes to present a complete picture of post-AMI hospitalisations compared with non-AMI patients.
Paper
3: Hospitalisation
2: Geographical variation in hospitalisation
and mortality for patients surviving acute myocardial
infarction according to receipt of percutaneous coronary intervention (PCI).
infarction.
Anticipated submission date:
August 2022.
July 2023-December 2023.
This paper will specifically look at the association between receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners. We have allocated additional researcher time in order to complete this objective with new timelines. Progress: Draft coding scripts for the analyses of post AMI hospitalisations by geographical area have been generated and preliminary analyses is underway.
2) Wider academic dissemination of the research findings will also be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2021/2022) and the European Society of Cardiology Congress (August 2021/2022).
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of percutaneous coronary intervention (PCI). Anticipated submission date: December 2023. This paper will specifically look at the association between receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2) Wider academic dissemination of the research findings will be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2023/2024) and the European Society of Cardiology Congress (August 2023/2024). We have opted to not share preliminary findings from this work at conferences at an earlier stage as originally planned. The size and complexity of this study has meant that we do not wish to have the outputs in the public domain until it has undergone scientific peer-review to the highest standards (i.e. through external peer-review processes) - which has delayed our original dissemination strategy.
[6 paragraphs unchanged]
Research referenced in NICE guidelines:
1) Clinical risk scoring for acute myocardial infarction referenced in NICE Clinical Guideline 94 (Gale CP et al, Heart, 2008; 95(3):221-227)
2) Atrial fibrillation research referenced in Atrial fibrillation: management, NICE clinical guideline 180 (Cowan, Long and Gale et al. Heart (2013): heartjnl-2012)
3) Pre-hospital ECG research referenced in European Resuscitation Council Guidelines for Resuscitation 2015 (Quinn, and Gale, et al. Heart (2014): heartjnl-2013).
Widespread media coverage and beyond:
1) Research by Dondo and Hall et al (Dondo T, Hall M, et al. Eur Heart J Acute Cardiovasc Care. 2016) has received widespread media coverage including radio and TV broadcasts as well as broadsheets and tabloids. This research also resulted in being invited to present at the Westminster health forum and forms the critical evidence for the non-ST elevation myocardial infarction NICE Implementation Collaborative.
2) Research by Hall and Gale et al (JAMA 2016; Aug 30. doi: 10.1001/jama.2016.10766) as well as Wu, Hall and Gale et al (Eur Heart J Acute Cardiovasc Care 2016; Aug 29. pii: 2048872616661693) forms the evidence for the guidelines in practice NIC project (https://www.guidelinesinpractice.co.uk/nic-projects).
3) The geographic variation in AMI treatment by Dondo, Hall and Gale et al (BMJ Open 2016; 6 (7): e011600) has received widespread media coverage and has led to a successful Department of Health / NHS England business case to develop the work further into a feedback quality improvement programme for patients, hospitals and commissioners (work currently ongoing).
[1 paragraph unchanged]
Expected measurable benefits
[6 paragraphs unchanged]
The Cardiovascular Epidemiology research group has expertise in health services research which directly impacts on patients, policies and healthcare professionals through research papers, the media and conferences. The research group has an excellent track record for translating research to clinical impact, which specific examples listed below:
Research referenced in NICE guidelines:
1) Clinical risk scoring for acute myocardial infarction referenced in NICE Clinical Guideline 94 (Gale CP et al, Heart, 2008; 95(3):221-227)
2) Atrial fibrillation research referenced in Atrial fibrillation: management, NICE clinical guideline 180 (Cowan, Long and Gale et al. Heart (2013): heartjnl-2012)
3) Pre-hospital ECG research referenced in European Resuscitation Council Guidelines for Resuscitation 2015 (Quinn, and Gale, et al. Heart (2014): heartjnl-2013).
Widespread media coverage and beyond:
1) Research by Dondo and Hall et al (Dondo T, Hall M, et al. Eur Heart J Acute Cardiovasc Care. 2016) has received widespread media coverage including radio and TV broadcasts as well as broadsheets and tabloids. This research also resulted in being invited to present at the Westminster health forum and forms the critical evidence for the non-ST elevation myocardial infarction NICE Implementation Collaborative.
2) Research by Hall and Gale et al (JAMA 2016; Aug 30. doi: 10.1001/jama.2016.10766) as well as Wu, Hall and Gale et al (Eur Heart J Acute Cardiovasc Care 2016; Aug 29. pii: 2048872616661693) forms the evidence for the guidelines in practice NIC project (https://www.guidelinesinpractice.co.uk/nic-projects).
3) The geographic variation in AMI treatment by Dondo, Hall and Gale et al (BMJ Open 2016; 6 (7): e011600) has received widespread media coverage and has led to a successful Department of Health / NHS England business case to develop the work further into a feedback quality improvement programme for patients, hospitals and commissioners (work currently ongoing).
[3 paragraphs unchanged]
Benefits reported
The sheer size of these data (145 million
hospitalization
hospitalisation
records) and complexity of analyses have meant the originally targeted timeframes were not met, and as such, the benefits outlined above have not yet been realised.
A research ready dataset is now held, with an extensive set of preliminary analyses and is on target for the planned outputs as outlined within this agreement.
The Cardiovascular Epidemiology research group have made significant progress with the analyses of these data since the last DSA extension, and continue to make this work a primary focus of the research group. We now have two publications at a final draft stage, currently undergoing collaborator review and final refinements before submission to journals according to the updated timelines listed in the outputs section of this application.
As a result of the coronavirus pandemic, the primary researcher was placed on furlough from April-September 2020, meaning that the research halted for several months. It is for this reason that there is no update to the yielded benefits section since the previous version of this application was submitted in early 2020.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the results now emerging to answer the proposed research questions will not be lost.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the preliminary results now emerging to answer the proposed research questions will not be lost.
DARS-NIC-17649-G0X4B-v2.4 3 September 2020 to 12 February 2023
- Title
- Hospitalisation and Mortality after Acute Myocardial Infarction
- Commercial
- No
- Sublicensing
- No
- Datasets
- 1
- Files released
- 0
Datasets: Hospital Episode Statistics Admitted Patient Care (HES APC)
What changed from DARS-NIC-17649-G0X4B-v1.5
Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.
| Field | Was | Became |
|---|---|---|
| Start date | 2020-09-03 | |
| Hospital Episode Statistics Admitted Patient Care (HES APC): legal basis | Health and Social Care Act 2012 - s261 - 'Other dissemination of information' |
Objective for processing
[2 paragraphs unchanged]
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving AMI in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
As a publicly funded organisation, the University of Leeds processes personal data to undertake scientific research which is in the public interest. Therefore, the legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
[1 paragraph unchanged]
Specifically, the research will aim to quantify the
long term
long-term
outcomes and hospitalisation rates for survivors of acute myocardial infarction in England.
[1 paragraph unchanged]
1. To describe hospitalisation patterns and endpoints (heart failure, cerebrovascular disease, coronary
[6 words unchanged]
myocardial infarction, atrial fibrillation, all-cause mortality) for patients hospitalised with non-fatal AMI
(i.e.
(i.e.,
survivors of the index hospital stay) compared to those who have no recorded
AMI .
AMI.
[7 paragraphs unchanged]
A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a
population based
population-based
cohort study to a case-control study, through which it is not possible
[17 words unchanged]
a sample of ‘controls’ which are representative across a range of hospitalisations
(i.e.
(i.e.,
all the study outcomes) in England could not be guaranteed. Without a
[38 words unchanged]
suitable for studying multiple outcome measures as proposed by the research group.
[3 paragraphs unchanged]
The research group cannot be sure that patients admitted to hospital with
[40 words unchanged]
proposes to derive the cohort for analysis from admissions from 2008/09 to
2016/17
2016/17,
however, has requested for NHS digital to exclude all people from the
[79 words unchanged]
at time trends in hospitalisations and mortality as part of the analysis.
[3 paragraphs unchanged]
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving AMI in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
The legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
[6 paragraphs unchanged]
University of Leeds are both the Data Controller and Data Processor for this study. There are no further data
processors
processors,
or third
parties
parties,
involved in the analysis of the data.
Microsoft Limited supply provide Cloud Services for the University of Leeds and are therefore listed as a data processor. They supply support to the system, but do not access data.
The British Heart Foundation (Project Grant PG/13/81/30474) which was used to fund the initial data application has now ended. The cardiovascular epidemiology group are continuing to analyse NHS Digital data for the specific objectives and purpose outlined within this application to ensure the measurable benefits outlined within the application can be realised. Funding to support ongoing research has been obtained from the Wellcome Trust (ref 206470/Z/17/Z).
For the avoidance of doubt, the British Heart Foundation and the Wellcome Trust will not influence the results or dissemination of the research conducted, and the British Heart Foundation and the Wellcome Trust will have no role in the design, analysis or interpretation of the research.
Data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Processing activities
Pseudonymised record level data was sent by NHS Digital to the University of Leeds under v0 of this agreement, for the purposes outlined within this application.
There were no subsequent flows of data.
Apart from this single transfer of data – there will be no other flows of data to or from NHS Digital as part of v1 of this agreement.
No new data will be disseminated under this agreement.
Data will be stored on
Processing and cleaning of data are conducted in SQL Server and statistical analyses are conducted in specialist statistical software including Stata and R. Only aggregated data with small number suppression applied in line with
the
University
HES Analysis guide are moved outside
of
Leeds' Secure Electronic Environment
Microsoft Azure
for
Data (SEED) system.
presentation and publication purposes.
The data will only be accessible to authorized individuals in the Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of
Leeds.
Leeds via password restricted computer access.
The data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
All individuals accessing these data have undertaken the advanced IT security training provided by the University of Leeds and will do so on an annual basis for the duration of the project.
[4 paragraphs unchanged]
Iron Mountain and University of York provide off-site secure backup facilities. Iron Mountain and University of York have no access to NHS Digital Data under this proposal.
Microsoft Limited supply Cloud Services for the University of Leeds and are therefore listed as a data processor. They supply support to the system, but do not access data. Therefore, any access to the data held under this agreement would be considered a breach of the agreement.
Data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Expected output
[1 paragraph unchanged]
The University of Leeds are now applying for an extension of the data sharing agreement to allow target dissemination plans (with dates now shifted as per new estimated target dates).
[2 paragraphs unchanged]
2) The data extract is incredibly large (145 million hospitalisation episodes), which
[39 words unchanged]
analyses and are on target for the following extended target dates for
our
the
specific outputs, as outlined in the original application.
1). Peer-reviewed publications
3) Further delays have occurred due to the Covid 19 pandemic. The primary researcher for the project was placed on full time furlough from April 2020 to September 2020. Full time work on the research recommenced as of October 2020, however the primary researcher has since left temporarily on maternity leave. Timelines have been further updated to reflect these additional delays.
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission for initial peer review: May 2020.
The expected outputs are as follows:
1) Peer-reviewed publications
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission for initial peer review: October 2021 – February 2022.
[1 paragraph unchanged]
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August
2021.
2022.
[3 paragraphs unchanged]
2).
2)
Wider academic dissemination of the research findings will also be made at
[13 words unchanged]
Cardiology conference (June 2021/2022) and the European Society of Cardiology Congress (August
2020/2021/2022).
2021/2022).
3).
3)
Lay summaries of the research findings will be generated and disseminated to the following key stakeholders:
[6 paragraphs unchanged]
Expected measurable benefits
This study will quantify the burden of hospitalisations and
long term
long-term
outcomes for patients who are admitted to NHS hospitals and surviving acute
[30 words unchanged]
as well as NHS commissioners via the formal networks discussed above (outputs).
[1 paragraph unchanged]
Quantifying the burden of hospitalisations and
long term
long-term
outcomes for patients surviving AMI in England will for the first time,
[35 words unchanged]
have resulted in improved survival rates for patients with AMI – the
long term
long-term
health burden of patients following their AMI is not yet known –
[24 words unchanged]
for future clinical trials, to ensure that not only the mortality or
short term
short-term
cardiovascular outcomes of AMI patients are considered, but also
longer term
longer-term
cardiovascular and non-cardiovascular outcomes in developing and testing new treatments in future.
[2 paragraphs unchanged]
Knowledge of geographical variation in hospitalisation and mortality from AMI will identify
[12 words unchanged]
NHS commissioners will enable them to act upon such inequalities to ultimately
drive up
drive-up
standards and provide direct and measurable benefit to patients and the NHS.
[64 words unchanged]
importantly, also on re-hospitalisation for a range of cardiovascular and non-cardiovascular conditions.
[6 paragraphs unchanged]
1) Research by Dondo and Hall et al (Dondo T, Hall M,
[26 words unchanged]
research also resulted in being invited to present at the Westminster health
forum,
forum
and forms the critical evidence for the non-ST elevation myocardial infarction NICE Implementation Collaborative.
[1 paragraph unchanged]
3) The geographic variation in AMI treatment by Dondo, Hall and Gale et al (BMJ Open 2016; 6 (7): e011600) has received widespread media
coverage,
coverage
and has led to a successful Department of Health / NHS England
[9 words unchanged]
feedback quality improvement programme for patients, hospitals and commissioners (work currently ongoing).
[3 paragraphs unchanged]
Benefits reported
The sheer size of these data (145 million hospitalization records) and complexity of analyses have meant the originally targeted
time-frames
timeframes
were not met, and as such, the benefits outlined above have not yet been realised.
A research ready
data-set
dataset
is now held, with an extensive set of preliminary analyses and is on target for the planned outputs as outlined within this agreement.
As a result of the coronavirus pandemic, the primary researcher was placed on furlough from April-September 2020, meaning that the research halted for several months. It is for this reason that there is no update to the yielded benefits section since the previous version of this application was submitted in early 2020.
[1 paragraph unchanged]
Objective for processing
The objective for processing HES data is to perform research into survival following heart attack in England.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease.
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving AMI in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
As a publicly funded organisation, the University of Leeds processes personal data to undertake scientific research which is in the public interest. Therefore, the legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
Over the last decade, there has been a substantial and sustained decline in mortality rates from cardiovascular disease in the UK. Despite this, cardiovascular disease remains the biggest killer in the UK and someone is admitted to an NHS hospital with a heart attack every three minutes. Moreover, improvements in acute myocardial infarction (AMI; heart attacks) survival are likely to be a major cause for the increasing incidence of heart failure (‘transferred morbidity’), which now affects around 900,000 individuals in the UK and accounts for 5% of all emergency hospitalisations. Presently, most patients with cardiovascular disease are elderly and because AMI survival has increased there are more patients living longer with co-morbidities. More frequently, such patients require specialist cardiovascular care in the form of invasive cardiac procedures including high and low voltage and resynchronization pacemakers and coronary revascularisation. Moreover, they frequently re-present to hospital – escalating the burden of admissions with heart failure.
Specifically, the research will aim to quantify the long-term outcomes and hospitalisation rates for survivors of acute myocardial infarction in England.
The objectives of the analysis are:
1. To describe hospitalisation patterns and endpoints (heart failure, cerebrovascular disease, coronary revascularisation, vascular dementia, severe bleeding, acute myocardial infarction, atrial fibrillation, all-cause mortality) for patients hospitalised with non-fatal AMI (i.e., survivors of the index hospital stay) compared to those who have no recorded AMI.
2. To identify factors associated with hospitalisation and endpoints for hospital survivors of index AMI compared to those who have no recorded AMI specifically focusing on geographical variation and the provision of timely percutaneous coronary intervention.
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease for example) following survival from AMI, the research group at the University of Leeds need to ensure they have a clean cohort for analysis to minimise confounding where possible as well as data of the hospitalisations occurring among patients who have no recorded AMI. Detailed justification for the request of this level of data is outlined below.
1. Reasons for requiring hospitalisations amongst patients with no recorded AMI
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease, heart failure and other outcomes) following survival from AMI, the research group needs to compare the number of each hospitalisation type occurring amongst AMI patients to the number of each hospitalisation type which occur in the background population (in this case, the background population is the population of patients admitted to hospital without an AMI in the same study period).
The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the research group will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
The research group will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations.
The research group considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population.
A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population-based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e., all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the research group.
The research group would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
Finally, the research group aim to determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
2. Ensuring derivation of a ‘clean’ cohort
The research group cannot be sure that patients admitted to hospital with AMI in a given period of HES data have not had a previous AMI, or, whether they have had any of the conditions they are considering as outcomes, such as cerebrovascular disease prior to their AMI. Therefore, the research group proposes to derive the cohort for analysis from admissions from 2008/09 to 2016/17, however, has requested for NHS digital to exclude all people from the data who have had a previous AMI or any of the hospitalisations the research group are considering as outcomes. The same filtering of prior conditions will be done by NHS digital for patients who have not had an AMI from 2008/09 onwards. This is a substantial minimisation effort by the research group, as without this step, data from 200-1/02 would have been required as part of the data application. Data from 2008/09 onwards will give sufficient data to look at time trends in hospitalisations and mortality as part of the analysis.
Details of all hospital attendances (not restricted to specific conditions with known associations with AMI) are required in order to understand the history of the patient and whether past (non-related) attendances have contributed in any way to that AMI attendance, or to any of the study outcomes including heart and non-heart related outcomes for patients in the AMI or non-AMI cohort. Post attendances also supports this (whether AMI contribute to non-heart related attendances).
Additionally, for each individual NHS Digital will provide a vital status indicator (alive/deceased) and, where individuals are deceased, the number of days between the data of admission and the date of death. The date of admission will not be supplied to the University of Leeds making it impossible for the research group to derive the date of death from the data supplied.
The proposed work to study the hospitalization patterns and outcomes for patients with acute myocardial infarction is part of a larger programme of work funded by the British Heart Foundation (Project Grant PG/13/81/30474) in order to fill an important knowledge gap of the long term hospital burden and non-fatal outcomes for patients with AMI using contemporary, large scale and national observational data. The British Heart Foundation Project Grant was entitled: “Cumulative Missed Opportunities for Care after Acute Myocardial Infarction: a linked national cardiovascular registries cohort study to identify preventable deaths“ and has now ended. NHS Digital data in relation to this DSA will not be used for any other programmes of work except those outlined in the specific objectives within this agreement.
DATA MINIMISATION:
Under GDPR, University of Leeds are adhering to the standard of data minimisation ensuring the data is adequate, relevant and limited to what is necessary for the stated purposes of processing. This has been achieved in the following ways:
1) Data is pseudonymised;
2) Data is filtered to exclude patients with AMI prior to 2008/09 and exclude any patients with any of the clinical conditions considered as outcomes (see section 2 ‘Ensuring derivation of a ‘clean’ cohort’ of the Objective for processing part of the application for full details of filtering)
3) Data fields are minimised by aggregating all date fields to Month and Year format
4) Date of death and date of birth are minimised to number of days survived with a mortality flag and age respectively.
University of Leeds are both the Data Controller and Data Processor for this study. There are no further data processors, or third parties, involved in the analysis of the data. Microsoft Limited supply provide Cloud Services for the University of Leeds and are therefore listed as a data processor. They supply support to the system, but do not access data.
The British Heart Foundation (Project Grant PG/13/81/30474) which was used to fund the initial data application has now ended. The cardiovascular epidemiology group are continuing to analyse NHS Digital data for the specific objectives and purpose outlined within this application to ensure the measurable benefits outlined within the application can be realised. Funding to support ongoing research has been obtained from the Wellcome Trust (ref 206470/Z/17/Z).
For the avoidance of doubt, the British Heart Foundation and the Wellcome Trust will not influence the results or dissemination of the research conducted, and the British Heart Foundation and the Wellcome Trust will have no role in the design, analysis or interpretation of the research.
Data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Expected output
The Cardiovascular Epidemiology research group dissemination strategy has remained the same as outlined within the original agreement with NHS Digital with the exception of the target dates.
There are several reasons which have factored into delays in producing the planned outputs according to the originally planned timeframes which include:
1) The application took a significant amount of time to complete which meant there were other projects competing for research staff time at the time of data receipt, and subsequently, there was a period of maternity leave for the primary researcher for the project.
2) The data extract is incredibly large (145 million hospitalisation episodes), which has meant loading, formatting, processing and cleaning of the data have taken a significantly longer period of time to complete than originally anticipated. The University of Leeds now have a research ready dataset, with an extensive set of preliminary analyses and are on target for the following extended target dates for the specific outputs, as outlined in the original application.
3) Further delays have occurred due to the Covid 19 pandemic. The primary researcher for the project was placed on full time furlough from April 2020 to September 2020. Full time work on the research recommenced as of October 2020, however the primary researcher has since left temporarily on maternity leave. Timelines have been further updated to reflect these additional delays.
The expected outputs are as follows:
1) Peer-reviewed publications
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission for initial peer review: October 2021 – February 2022.
This paper will quantify the hospitalisations and long term outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity.
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August 2022.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners.
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of percutaneous coronary intervention (PCI). Anticipated submission date: August 2022.
This paper will specifically look at the association between receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2) Wider academic dissemination of the research findings will also be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2021/2022) and the European Society of Cardiology Congress (August 2021/2022).
3) Lay summaries of the research findings will be generated and disseminated to the following key stakeholders:
- The British Cardiovascular Society (BCS),
- British Heart Foundation (BHF), TakeHeart,
- NHS commissioners and clinicians/health professionals involved in managing heart attack.
The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community.
In addition, the group website (Cardiovascular Epidemiology - https://lida.leeds.ac.uk/research-projects/566-2/) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project at appropriate stages.
All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.
Benefits reported
The sheer size of these data (145 million hospitalization records) and complexity of analyses have meant the originally targeted timeframes were not met, and as such, the benefits outlined above have not yet been realised.
A research ready dataset is now held, with an extensive set of preliminary analyses and is on target for the planned outputs as outlined within this agreement.
As a result of the coronavirus pandemic, the primary researcher was placed on furlough from April-September 2020, meaning that the research halted for several months. It is for this reason that there is no update to the yielded benefits section since the previous version of this application was submitted in early 2020.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the preliminary results now emerging to answer the proposed research questions will not be lost.
DARS-NIC-17649-G0X4B-v1.5 13 February 2020 to 12 February 2023
- Title
- Hospitalisation and Mortality after Acute Myocardial Infarction
- Commercial
- No
- Sublicensing
- No
- Datasets
- 1
- Files released
- 0
Datasets: Hospital Episode Statistics Admitted Patient Care (HES APC)
What changed from DARS-NIC-17649-G0X4B-v0.6
Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.
| Field | Was | Became |
|---|---|---|
| Start date | 2020-02-13 | |
| End date | 2023-02-12 | |
| Hospital Episode Statistics Admitted Patient Care (HES APC): common law duty of confidentiality | Does not include the flow of confidential data |
Objective for processing
The objective for processing
of these
HES
data is to perform research into survival following heart attack in England.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease.
[5 paragraphs unchanged]
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease for example) following survival from AMI, the
study team
research group
at the University of Leeds need to ensure they have a clean
[24 words unchanged]
justification for the request of this level of data is outlined below.
[1 paragraph unchanged]
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease, heart failure and other outcomes) following survival from AMI, the
study team
research group
needs to compare the number of each hospitalisation type occurring amongst AMI
[24 words unchanged]
patients admitted to hospital without an AMI in the same study period).
The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will then be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the study team will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
The study team will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations. The study team considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population. A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e. all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the study team. We would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the research group will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
Finally, the study team aim to additionally determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
The research group will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations.
The research group considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population.
A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e. all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the research group.
The research group would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
Finally, the research group aim to determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
[1 paragraph unchanged]
The
study team
research group
cannot be sure that patients admitted to hospital with AMI in a
[23 words unchanged]
as outcomes, such as cerebrovascular disease prior to their AMI. Therefore, the
study team
research group
proposes to derive the cohort for analysis from admissions from 2008/09 to
present day,
2016/17
however, has requested for NHS digital to exclude all people from the data who have had a previous AMI or any of the hospitalisations the
study team
research group
are considering as outcomes. The same filtering of prior conditions will be
[11 words unchanged]
AMI from 2008/09 onwards. This is a substantial minimisation effort by the
study team,
research group,
as without this step, data from 200-1/02 would have been required as
[15 words unchanged]
at time trends in hospitalisations and mortality as part of the analysis.
[1 paragraph unchanged]
Additionally, for each individual NHS Digital will provide a vital status indicator
[26 words unchanged]
be supplied to the University of Leeds making it impossible for the
study team
research group
to derive the date of death from the data supplied.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease.
The proposed work to study the
hospitalisation
hospitalization
patterns and outcomes for patients with acute myocardial infarction is part of
[31 words unchanged]
for patients with AMI using contemporary, large scale and national observational data.
The British Heart Foundation Project Grant was entitled: “Cumulative Missed Opportunities for Care after Acute Myocardial Infarction: a linked national cardiovascular registries cohort study to identify preventable deaths“ and has now ended. NHS Digital data in relation to this DSA will not be used for any other programmes of work except those outlined in the specific objectives within this agreement.
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving AMI in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
The legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
DATA MINIMISATION:
Under GDPR, University of Leeds are adhering to the standard of data minimisation ensuring the data is adequate, relevant and limited to what is necessary for the stated purposes of processing. This has been achieved in the following ways:
1) Data is pseudonymised;
2) Data is filtered to exclude patients with AMI prior to 2008/09 and exclude any patients with any of the clinical conditions considered as outcomes (see section 2 ‘Ensuring derivation of a ‘clean’ cohort’ of the Objective for processing part of the application for full details of filtering)
3) Data fields are minimised by aggregating all date fields to Month and Year format
4) Date of death and date of birth are minimised to number of days survived with a mortality flag and age respectively.
University of Leeds are both the Data Controller and Data Processor for this study. There are no further data processors or third parties involved in the analysis of the data.
Processing activities
Data will be stored on the University of Leeds' Secure Electronic Environment for Data (SEED) system. The data will only be accessible to authorized individuals in the Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. The data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Pseudonymised record level data was sent by NHS Digital to the University of Leeds under v0 of this agreement, for the purposes outlined within this application.
The data will be geocoded based on Lower Super Output Area (LSOA) to obtain information on higher aggregated geographical units (Clinical Commissioning Groups). No further linkage to the data will occur. Only summarised and aggregated data will be disseminated in the form of academic presentations and peer-reviewed journals.
Apart from this single transfer of data – there will be no other flows of data to or from NHS Digital as part of v1 of this agreement.
The data will not be used for commercial purposes, provided in record level form to any third party or used for any direct marketing.
Data will be stored on the University of Leeds' Secure Electronic Environment for Data (SEED) system. The data will only be accessible to authorized individuals in the Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds.
The study is funded by the British Heart Foundation. For the avoidance of doubt, the British Heart Foundation will not influence the results or dissemination of the research conducted, and the British Heart Foundation will have no role in the design, analysis or interpretation of the research.
The data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
The data will be geocoded based on Lower Super Output Area (LSOA) to obtain information on higher aggregated geographical units (Clinical Commissioning Groups).
No further linkage to the data will occur. Only summarised and aggregated data will be disseminated in the form of academic presentations, peer-reviewed journals and lay summaries (in line with the HES analysis guide)
The data will not be used for commercial purposes, provided in record level form to any third party or used for any direct marketing. There will be no requirement or attempt to re-identify any individuals within the data.
The study was funded by the British Heart Foundation. The British Heart Foundation did not influence the results or dissemination of the research conducted, and the British Heart Foundation had no role in the design, analysis or interpretation of the research.
Iron Mountain and University of York provide off-site secure backup facilities. Iron Mountain and University of York have no access to NHS Digital Data under this proposal.
Data will only be accessed by substantive employees of the University of Leeds and only used for the purpose of this project.
Expected output
Whilst the planned analyses will be disseminated to the academic and medical community in peer reviewed publications and presented at relevant conferences (see below), it is the clinical implications of the results for healthcare professionals, patients and regulators that are of greater virtue. It is clear that the results from the proposed study will help answer major gaps in the knowledge base of the health burden and ongoing hospitalisation for the increasing number of survivors following acute myocardial infarction which can therefore contribute to future healthcare policy. The Cardiovascular Epidemiology research team has established connections with numerous relevant groups through which findings will be disseminated to the NHS as well as patients.
The Cardiovascular Epidemiology research group dissemination strategy has remained the same as outlined within the original agreement with NHS Digital with the exception of the target dates.
These groups include: The NICE Indicator Advisory Group, the European Society of Cardiology Acute Cardiovascular Care Association, the European Society of Cardiology Acute Cardiovascular Care Association Quality of Care Group, the British Cardiovascular Society Guidelines and Practice Committee and the National Institute for Cardiovascular Outcomes Research (NICOR) Research Executive.
The University of Leeds are now applying for an extension of the data sharing agreement to allow target dissemination plans (with dates now shifted as per new estimated target dates).
The Cardiovascular Epidemiology research group is led by an Associate Professor of Cardiovascular Health Sciences at the University of Leeds who is also a member of the above groups and is additionally an honorary consultant Cardiologist at York Teaching Hospitals NHS trust and secretary of the European Society of Cardiology Acute Cardiovascular Care Association – offering further dissemination routes which will be utilised.
There are several reasons which have factored into delays in producing the planned outputs according to the originally planned timeframes which include:
NICE identifies awareness and knowledge as well as lack of motivation by healthcare professionals to be some of the key barriers to change in the NHS. Patients are at the heart of providing motivation for healthcare professionals to improve care in the NHS, therefore the Cardiovascular Epidemiology research group will focus on dissemination to patients as well through charities listed in point 3 below.
1) The application took a significant amount of time to complete which meant there were other projects competing for research staff time at the time of data receipt, and subsequently, there was a period of maternity leave for the primary researcher for the project.
The Cardiovascular Epidemiology research group dissemination strategy will be as follows:
2) The data extract is incredibly large (145 million hospitalisation episodes), which has meant loading, formatting, processing and cleaning of the data have taken a significantly longer period of time to complete than originally anticipated. The University of Leeds now have a research ready dataset, with an extensive set of preliminary analyses and are on target for the following extended target dates for our specific outputs, as outlined in the original application.
1). Peer-reviewed
publication
publications
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission
date: March 2018.
for initial peer review: May 2020.
[1 paragraph unchanged]
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August
2018.
2021.
[1 paragraph unchanged]
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of
timely
percutaneous coronary intervention (PCI). Anticipated submission date: August
2020.
2022.
This paper will specifically look at the association between
timely
receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2). Wider academic dissemination of the research findings will also be made
[5 words unchanged]
conferences as appropriate, such as the British Society of Cardiology conference (June
2017/18/19)
2021/2022)
and the European Society of Cardiology Congress (August
2018/19/20).
2020/2021/2022).
3). Lay summaries of the research findings will be generated and disseminated to the following key stakeholders: the British Cardiovascular Society (BCS), British Heart Foundation (BHF), TakeHeart, NHS commissioners and clinicians/health professionals involved in managing heart attack. The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community (see section 5d) In addition, the group website (Cardiovascular Epidemiology - https://medhealth.leeds.ac.uk/homepage/692/cardiovasucular_epidemiology-leeds_institute_of_cardiovascular_and_metabolic_medicine) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project.
3). Lay summaries of the research findings will be generated and disseminated to the following key stakeholders:
- The British Cardiovascular Society (BCS),
- British Heart Foundation (BHF), TakeHeart,
- NHS commissioners and clinicians/health professionals involved in managing heart attack.
The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community.
In addition, the group website (Cardiovascular Epidemiology - https://lida.leeds.ac.uk/research-projects/566-2/) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project at appropriate stages.
[1 paragraph unchanged]
Expected measurable benefits
This study will quantify the burden of hospitalisations and long term outcomes
[39 words unchanged]
the public as well as NHS commissioners via the formal networks discussed
in section 5c. Dissemination of the factors which could lead to increased hospitalisations and mortality to clinicians (via academic publications, presentations at clinical conferences and dissemination via the British Cardiac Society and British Heart Foundation) is envisaged to be a driver for improved patient care and has far reaching clinical and social benefits as outlined below.
above (outputs).
Quantifying the burden of hospitalisations and long term outcomes for patients surviving AMI in England will for the first time, on a national scale, provide NHS commissioners with the necessary evidence to plan effectively for service provision and resource allocation for the large proportion of patients who now survive their AMI. Although improvements in treatment have resulted in improved survival rates for patients with AMI – the long term health burden of patients following their AMI is not yet known – and this is what the study team propose to determine. In addition, the findings from this study can be used to inform new endpoints for future clinical trials, to ensure that not only the mortality or short term cardiovascular outcomes of AMI patients are considered, but also longer term cardiovascular and non-cardiovascular outcomes in developing and testing new treatments in future.
Dissemination of the factors which could lead to increased hospitalisations and mortality to clinicians (via academic publications, presentations at clinical conferences and dissemination via the British Cardiac Society and British Heart Foundation) is envisaged to be a driver for improved patient care and has far reaching clinical and social benefits as outlined below.
Quantifying the burden of hospitalisations and long term outcomes for patients surviving AMI in England will for the first time, on a national scale, provide NHS commissioners with the necessary evidence to plan effectively for service provision and resource allocation for the large proportion of patients who now survive their AMI. Although improvements in treatment have resulted in improved survival rates for patients with AMI – the long term health burden of patients following their AMI is not yet known – and this is what the research group propose to determine. In addition, the findings from this study can be used to inform new endpoints for future clinical trials, to ensure that not only the mortality or short term cardiovascular outcomes of AMI patients are considered, but also longer term cardiovascular and non-cardiovascular outcomes in developing and testing new treatments in future.
[12 paragraphs unchanged]
Several members of the Cardiovascular Epidemiology research group have experience with Public
&
and
Patient Involvement, meeting patients to discuss a range of different research proposal and allowing
them to
influence and
be part of
participation in
the research agenda as well as
access to
disseminate research
findings back to them,
findings,
whilst some members of the research group hold very close involvement with
[11 words unchanged]
who therefore act as a powerful conduit for change and knowledge dissemination.
The research the Cardiovascular Epidemiology research group propose here is of direct
[89 words unchanged]
need for cardiovascular care, as well as where, when and in whom.
The research group has an excellent track record to ensure this research outputs are far reaching with high impact.
Whilst exact estimates of the numbers of people in the UK who have survived AMI are difficult to estimate – a report in 2015 estimated the prevalence of AMI in the UK to be around 915,000 people (Bhatnagar P et al. Heart 2015;101:1182–9. https://doi.org/10.1136/heartjnl-2015-307516). Knowledge of the hospitalisation patterns and endpoints following AMI will lead to better service provision planning which can improve the quality of health care for more than 900,000 people in the UK who have suffered AMI, but also for people who may experience AMI in future as well as their families and carers.
Benefits reported
Yielded Benefits is not a requirement for new applications.
The sheer size of these data (145 million hospitalization records) and complexity of analyses have meant the originally targeted time-frames were not met, and as such, the benefits outlined above have not yet been realised.
A research ready data-set is now held, with an extensive set of preliminary analyses and is on target for the planned outputs as outlined within this agreement.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the preliminary results now emerging to answer the proposed research questions will not be lost.
Objective for processing
The objective for processing HES data is to perform research into survival following heart attack in England.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease.
Over the last decade, there has been a substantial and sustained decline in mortality rates from cardiovascular disease in the UK. Despite this, cardiovascular disease remains the biggest killer in the UK and someone is admitted to an NHS hospital with a heart attack every three minutes. Moreover, improvements in acute myocardial infarction (AMI; heart attacks) survival are likely to be a major cause for the increasing incidence of heart failure (‘transferred morbidity’), which now affects around 900,000 individuals in the UK and accounts for 5% of all emergency hospitalisations. Presently, most patients with cardiovascular disease are elderly and because AMI survival has increased there are more patients living longer with co-morbidities. More frequently, such patients require specialist cardiovascular care in the form of invasive cardiac procedures including high and low voltage and resynchronization pacemakers and coronary revascularisation. Moreover, they frequently re-present to hospital – escalating the burden of admissions with heart failure.
Specifically, the research will aim to quantify the long term outcomes and hospitalisation rates for survivors of acute myocardial infarction in England.
The objectives of the analysis are:
1. To describe hospitalisation patterns and endpoints (heart failure, cerebrovascular disease, coronary revascularisation, vascular dementia, severe bleeding, acute myocardial infarction, atrial fibrillation, all-cause mortality) for patients hospitalised with non-fatal AMI (i.e. survivors of the index hospital stay) compared to those who have no recorded AMI .
2. To identify factors associated with hospitalisation and endpoints for hospital survivors of index AMI compared to those who have no recorded AMI specifically focusing on geographical variation and the provision of timely percutaneous coronary intervention.
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease for example) following survival from AMI, the research group at the University of Leeds need to ensure they have a clean cohort for analysis to minimise confounding where possible as well as data of the hospitalisations occurring among patients who have no recorded AMI. Detailed justification for the request of this level of data is outlined below.
1. Reasons for requiring hospitalisations amongst patients with no recorded AMI
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease, heart failure and other outcomes) following survival from AMI, the research group needs to compare the number of each hospitalisation type occurring amongst AMI patients to the number of each hospitalisation type which occur in the background population (in this case, the background population is the population of patients admitted to hospital without an AMI in the same study period).
The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the research group will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
The research group will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations.
The research group considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population.
A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e. all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the research group.
The research group would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
Finally, the research group aim to determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
2. Ensuring derivation of a ‘clean’ cohort
The research group cannot be sure that patients admitted to hospital with AMI in a given period of HES data have not had a previous AMI, or, whether they have had any of the conditions they are considering as outcomes, such as cerebrovascular disease prior to their AMI. Therefore, the research group proposes to derive the cohort for analysis from admissions from 2008/09 to 2016/17 however, has requested for NHS digital to exclude all people from the data who have had a previous AMI or any of the hospitalisations the research group are considering as outcomes. The same filtering of prior conditions will be done by NHS digital for patients who have not had an AMI from 2008/09 onwards. This is a substantial minimisation effort by the research group, as without this step, data from 200-1/02 would have been required as part of the data application. Data from 2008/09 onwards will give sufficient data to look at time trends in hospitalisations and mortality as part of the analysis.
Details of all hospital attendances (not restricted to specific conditions with known associations with AMI) are required in order to understand the history of the patient and whether past (non-related) attendances have contributed in any way to that AMI attendance, or to any of the study outcomes including heart and non-heart related outcomes for patients in the AMI or non-AMI cohort. Post attendances also supports this (whether AMI contribute to non-heart related attendances).
Additionally, for each individual NHS Digital will provide a vital status indicator (alive/deceased) and, where individuals are deceased, the number of days between the data of admission and the date of death. The date of admission will not be supplied to the University of Leeds making it impossible for the research group to derive the date of death from the data supplied.
The proposed work to study the hospitalization patterns and outcomes for patients with acute myocardial infarction is part of a larger programme of work funded by the British Heart Foundation (Project Grant PG/13/81/30474) in order to fill an important knowledge gap of the long term hospital burden and non-fatal outcomes for patients with AMI using contemporary, large scale and national observational data. The British Heart Foundation Project Grant was entitled: “Cumulative Missed Opportunities for Care after Acute Myocardial Infarction: a linked national cardiovascular registries cohort study to identify preventable deaths“ and has now ended. NHS Digital data in relation to this DSA will not be used for any other programmes of work except those outlined in the specific objectives within this agreement.
LEGAL BASIS
The University of Leeds is an independent corporation established by Royal Charter. Its objects, powers and framework of governance are set out in the Charter and its supporting statutes, amendments. The University of Leeds is responsible for conducting scientific research for academic and public benefit. Under this project, the University aim to quantify the burden of hospitalisations and long term outcomes for patients surviving AMI in England, thereby providing NHS commissioners, clinicians and patients with the necessary evidence to allow commissioners to effectively plan for service provision and resource allocation, and to ensure high standards of quality of health care, for the large number of patients who have AMI.
The legal basis for processing the pseudonymised patient data requested under this agreement is GDPR Article 6 (1) (e), public task and, for special categories of data (including health information and information concerning ethnicity and sexual orientation), Article 9.2(j), archiving, research and statistics.
DATA MINIMISATION:
Under GDPR, University of Leeds are adhering to the standard of data minimisation ensuring the data is adequate, relevant and limited to what is necessary for the stated purposes of processing. This has been achieved in the following ways:
1) Data is pseudonymised;
2) Data is filtered to exclude patients with AMI prior to 2008/09 and exclude any patients with any of the clinical conditions considered as outcomes (see section 2 ‘Ensuring derivation of a ‘clean’ cohort’ of the Objective for processing part of the application for full details of filtering)
3) Data fields are minimised by aggregating all date fields to Month and Year format
4) Date of death and date of birth are minimised to number of days survived with a mortality flag and age respectively.
University of Leeds are both the Data Controller and Data Processor for this study. There are no further data processors or third parties involved in the analysis of the data.
Expected output
The Cardiovascular Epidemiology research group dissemination strategy has remained the same as outlined within the original agreement with NHS Digital with the exception of the target dates.
The University of Leeds are now applying for an extension of the data sharing agreement to allow target dissemination plans (with dates now shifted as per new estimated target dates).
There are several reasons which have factored into delays in producing the planned outputs according to the originally planned timeframes which include:
1) The application took a significant amount of time to complete which meant there were other projects competing for research staff time at the time of data receipt, and subsequently, there was a period of maternity leave for the primary researcher for the project.
2) The data extract is incredibly large (145 million hospitalisation episodes), which has meant loading, formatting, processing and cleaning of the data have taken a significantly longer period of time to complete than originally anticipated. The University of Leeds now have a research ready dataset, with an extensive set of preliminary analyses and are on target for the following extended target dates for our specific outputs, as outlined in the original application.
1). Peer-reviewed publications
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission for initial peer review: May 2020.
This paper will quantify the hospitalisations and long term outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity.
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August 2021.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners.
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of percutaneous coronary intervention (PCI). Anticipated submission date: August 2022.
This paper will specifically look at the association between receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2). Wider academic dissemination of the research findings will also be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2021/2022) and the European Society of Cardiology Congress (August 2020/2021/2022).
3). Lay summaries of the research findings will be generated and disseminated to the following key stakeholders:
- The British Cardiovascular Society (BCS),
- British Heart Foundation (BHF), TakeHeart,
- NHS commissioners and clinicians/health professionals involved in managing heart attack.
The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community.
In addition, the group website (Cardiovascular Epidemiology - https://lida.leeds.ac.uk/research-projects/566-2/) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project at appropriate stages.
All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.
Benefits reported
The sheer size of these data (145 million hospitalization records) and complexity of analyses have meant the originally targeted time-frames were not met, and as such, the benefits outlined above have not yet been realised.
A research ready data-set is now held, with an extensive set of preliminary analyses and is on target for the planned outputs as outlined within this agreement.
The extension requested within this application is crucial to ensure the important expected measurable benefits as outlined above can be realised within the new proposed time frames and so that efforts to date in processing and readying the data for statistical analyses as well as the preliminary results now emerging to answer the proposed research questions will not be lost.
DARS-NIC-17649-G0X4B-v0.6 13 February 2017 to 12 February 2020
- Title
- Hospitalisation and Mortality after Acute Myocardial Infarction
- Commercial
- No
- Sublicensing
- No
- Datasets
- 1
- Files released
- 9
Datasets: Hospital Episode Statistics Admitted Patient Care (HES APC)
Objective for processing
The objective for processing of these data is to perform research into survival following heart attack in England.
Over the last decade, there has been a substantial and sustained decline in mortality rates from cardiovascular disease in the UK. Despite this, cardiovascular disease remains the biggest killer in the UK and someone is admitted to an NHS hospital with a heart attack every three minutes. Moreover, improvements in acute myocardial infarction (AMI; heart attacks) survival are likely to be a major cause for the increasing incidence of heart failure (‘transferred morbidity’), which now affects around 900,000 individuals in the UK and accounts for 5% of all emergency hospitalisations. Presently, most patients with cardiovascular disease are elderly and because AMI survival has increased there are more patients living longer with co-morbidities. More frequently, such patients require specialist cardiovascular care in the form of invasive cardiac procedures including high and low voltage and resynchronization pacemakers and coronary revascularisation. Moreover, they frequently re-present to hospital – escalating the burden of admissions with heart failure.
Specifically, the research will aim to quantify the long term outcomes and hospitalisation rates for survivors of acute myocardial infarction in England.
The objectives of the analysis are:
1. To describe hospitalisation patterns and endpoints (heart failure, cerebrovascular disease, coronary revascularisation, vascular dementia, severe bleeding, acute myocardial infarction, atrial fibrillation, all-cause mortality) for patients hospitalised with non-fatal AMI (i.e. survivors of the index hospital stay) compared to those who have no recorded AMI .
2. To identify factors associated with hospitalisation and endpoints for hospital survivors of index AMI compared to those who have no recorded AMI specifically focusing on geographical variation and the provision of timely percutaneous coronary intervention.
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease for example) following survival from AMI, the study team at the University of Leeds need to ensure they have a clean cohort for analysis to minimise confounding where possible as well as data of the hospitalisations occurring among patients who have no recorded AMI. Detailed justification for the request of this level of data is outlined below.
1. Reasons for requiring hospitalisations amongst patients with no recorded AMI
In order to quantify the incidence of a range of hospitalisations (cerebrovascular disease, heart failure and other outcomes) following survival from AMI, the study team needs to compare the number of each hospitalisation type occurring amongst AMI patients to the number of each hospitalisation type which occur in the background population (in this case, the background population is the population of patients admitted to hospital without an AMI in the same study period). The hospitalisations occurring in the non-AMI population are used to determine the expected number of each hospitalisation type for someone hospitalised in the same year, and of the same age and sex as someone who has had AMI. The observed numbers of hospitalisations for those with AMI will then be compared to the expected number of hospitalisations amongst those without AMI to determine whether patients with AMI have more hospitalisations than expected (the excess hospitalisation incidence rate). Without this quantification of the excess hospitalisation incidence rate, the results will have no context as the study team will be unable to ascertain whether those with an AMI are more or less likely to experience certain conditions following their AMI than the background population. Determining this is the primary aim of the study.
The study team will require all hospitalisations amongst patients with no recorded AMI (subject to filtering described under “2. Ensuring derivation of a ‘clean’ cohort”) rather than a sample of hospitalisations. The study team considered the feasibility of selecting a reduced number of geographical areas to represent the nation rather than requesting national data but ruled out this approach because it would affect the validity of the findings. This is because an incidence rate is calculated from the observed hospitalisations amongst the AMI patients divided by the observed hospitalisations in the non-AMI population. A sample of hospitalisations (obtained from a reduced set of geographical areas) would change the study design from a population based cohort study to a case-control study, through which it is not possible to calculate incidence rates. Whilst it is possible to obtain relative risks from a case-control study, selecting a sample of ‘controls’ which are representative across a range of hospitalisations (i.e. all the study outcomes) in England could not be guaranteed. Without a representative cohort, the relative risks obtained would be prone to bias. Given the potential impact of the findings on healthcare users and NHS policy, it is essential to minimise uncertainty. In addition, a case-control study design is not suitable for studying multiple outcome measures as proposed by the study team. We would therefore be unable to achieve the study objective of defining the incidence of multiple hospitalisation outcomes following AMI if restricting the non-AMI patient data to a sample of the population, nor guarantee the validity of results which are obtainable under a case-control study design.
Finally, the study team aim to additionally determine the extent of geographical variation in hospitalisations and mortality following AMI, which require the calculation of incidence rates (and therefore require a full population denominator) for all areas in England.
2. Ensuring derivation of a ‘clean’ cohort
The study team cannot be sure that patients admitted to hospital with AMI in a given period of HES data have not had a previous AMI, or, whether they have had any of the conditions they are considering as outcomes, such as cerebrovascular disease prior to their AMI. Therefore, the study team proposes to derive the cohort for analysis from admissions from 2008/09 to present day, however, has requested for NHS digital to exclude all people from the data who have had a previous AMI or any of the hospitalisations the study team are considering as outcomes. The same filtering of prior conditions will be done by NHS digital for patients who have not had an AMI from 2008/09 onwards. This is a substantial minimisation effort by the study team, as without this step, data from 200-1/02 would have been required as part of the data application. Data from 2008/09 onwards will give sufficient data to look at time trends in hospitalisations and mortality as part of the analysis.
Details of all hospital attendances (not restricted to specific conditions with known associations with AMI) are required in order to understand the history of the patient and whether past (non-related) attendances have contributed in any way to that AMI attendance, or to any of the study outcomes including heart and non-heart related outcomes for patients in the AMI or non-AMI cohort. Post attendances also supports this (whether AMI contribute to non-heart related attendances).
Additionally, for each individual NHS Digital will provide a vital status indicator (alive/deceased) and, where individuals are deceased, the number of days between the data of admission and the date of death. The date of admission will not be supplied to the University of Leeds making it impossible for the study team to derive the date of death from the data supplied.
The research will be undertaken by the established Cardiovascular Epidemiology Research Group within the Leeds Institute of Cardiovascular and Metabolic Medicine at the University of Leeds. This research group has a remit of using large scale routine data and clinical registries alongside advanced analytical epidemiological techniques to better understand and improve the quality of care of patients with cardiovascular disease. The proposed work to study the hospitalisation patterns and outcomes for patients with acute myocardial infarction is part of a larger programme of work funded by the British Heart Foundation (Project Grant PG/13/81/30474) in order to fill an important knowledge gap of the long term hospital burden and non-fatal outcomes for patients with AMI using contemporary, large scale and national observational data.
Expected output
Whilst the planned analyses will be disseminated to the academic and medical community in peer reviewed publications and presented at relevant conferences (see below), it is the clinical implications of the results for healthcare professionals, patients and regulators that are of greater virtue. It is clear that the results from the proposed study will help answer major gaps in the knowledge base of the health burden and ongoing hospitalisation for the increasing number of survivors following acute myocardial infarction which can therefore contribute to future healthcare policy. The Cardiovascular Epidemiology research team has established connections with numerous relevant groups through which findings will be disseminated to the NHS as well as patients.
These groups include: The NICE Indicator Advisory Group, the European Society of Cardiology Acute Cardiovascular Care Association, the European Society of Cardiology Acute Cardiovascular Care Association Quality of Care Group, the British Cardiovascular Society Guidelines and Practice Committee and the National Institute for Cardiovascular Outcomes Research (NICOR) Research Executive.
The Cardiovascular Epidemiology research group is led by an Associate Professor of Cardiovascular Health Sciences at the University of Leeds who is also a member of the above groups and is additionally an honorary consultant Cardiologist at York Teaching Hospitals NHS trust and secretary of the European Society of Cardiology Acute Cardiovascular Care Association – offering further dissemination routes which will be utilised.
NICE identifies awareness and knowledge as well as lack of motivation by healthcare professionals to be some of the key barriers to change in the NHS. Patients are at the heart of providing motivation for healthcare professionals to improve care in the NHS, therefore the Cardiovascular Epidemiology research group will focus on dissemination to patients as well through charities listed in point 3 below.
The Cardiovascular Epidemiology research group dissemination strategy will be as follows:
1). Peer-reviewed publication
Paper 1: Hospitalisation and mortality after acute myocardial infarction. Anticipated submission date: March 2018.
This paper will quantify the hospitalisations and long term outcomes for patients surviving acute myocardial infarction as well as determine the factors which lead to increased hospitalisation and morbidity.
Paper 2: Geographical variation in hospitalisation and mortality for patients surviving acute myocardial infarction. Anticipated submission date: August 2018.
This paper will quantify the potential geographical variation in hospitalisation for patients surviving AMI to identify potential healthcare inequalities for NHS Commissioners.
Paper 3: Hospitalisation and mortality for patients surviving acute myocardial infarction according to receipt of timely percutaneous coronary intervention (PCI). Anticipated submission date: August 2020.
This paper will specifically look at the association between timely receipt of PCI and the long term hospitalisations and outcomes for patients surviving AMI.
2). Wider academic dissemination of the research findings will also be made at major national and international conferences as appropriate, such as the British Society of Cardiology conference (June 2017/18/19) and the European Society of Cardiology Congress (August 2018/19/20).
3). Lay summaries of the research findings will be generated and disseminated to the following key stakeholders: the British Cardiovascular Society (BCS), British Heart Foundation (BHF), TakeHeart, NHS commissioners and clinicians/health professionals involved in managing heart attack. The Cardiovascular Epidemiology Research group have previously liaised with these organisations to ensure wide reaching research impact, beyond the academic community (see section 5d) In addition, the group website (Cardiovascular Epidemiology - https://medhealth.leeds.ac.uk/homepage/692/cardiovasucular_epidemiology-leeds_institute_of_cardiovascular_and_metabolic_medicine) as well as the research team's twitter account (@UoLCardioEpi) will be used to update the public, the network of stakeholders, charities, and health professionals throughout the project.
All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.
Benefits reported
Yielded Benefits is not a requirement for new applications.
Register history
When this agreement appeared in, or was edited in, each monthly edition of the register. Built by comparing every edition this site holds, the earliest of which is July 2021.
-
July 2021 —
already listed in the earliest edition this site holds, so it may be older. 2 versions: DARS-NIC-17649-G0X4B-v0.6, DARS-NIC-17649-G0X4B-v1.5
-
August 2021
1 version added: DARS-NIC-17649-G0X4B-v2.4
-
December 2022
Register-wide edit DARS-NIC-17649-G0X4B-v0.6, DARS-NIC-17649-G0X4B-v1.5 — Datasets: legal basis: “
s261(1) and” taken out. Made to 639 agreements in this edition, so it is reported once, on the changes page, and not counted as an amendment of this agreement. -
March 2023
1 version added: DARS-NIC-17649-G0X4B-v3.2
Cite this page
NHS England (2026) Data Uses Register, September 2026 edition, agreement DARS-NIC-17649-G0X4B, “Hospitalisation and Mortality after Acute Myocardial Infarction”. Read via NHS Data Access Explorer (unofficial), https://healthdatauses.uk/agreements/dars-nic-17649-g0x4b/ (accessed [date]).
This address stays the same, but the page is rebuilt with each monthly edition, so the citation names the edition it shows. Every edition's data is kept in the facts store.
Source: datausesregister_september2026.xlsx, September 2026 edition of the NHS England Data Uses Register. Search that workbook for DARS-NIC-17649-G0X4B to see the original rows.