UK GRACE Risk Score Intervention Study

The UK GRACE Risk Score Intervention Study (UKGRIS) aims to investigate the effectiveness of the GRACE risk score on the management and outcome of patients hospitalised with non-ST elevation acute coronary syndrome for public health purposes. University of Leeds wish to investigate whether implementing the GRACE risk score increases the use of Class 1 guideline-indicated therapies for the management of Non ST-elevation ACS (NSTEACS), within the guideline recommended time, compared with current standard care in this population.

The UKGRIS study will also investigate whether implementing the GRACE risk score reduces the composite endpoints of cardiovascular death, non-fatal myocardial infarction, new onset heart failure with hospitalisation and cardiovascular readmission at twelve months, compared with current standard care in this population.

The University of Leeds UKGRIS study require data from NHS Digital as there is no clinical follow up with the research teams at participating centres to provide this information which is needed for the co-primary and secondary endpoints.

The justification for processing under GDPR is:

• Article 6 (1)(e) Task in the public interest and Article 9(2)(j) archiving, research and statistics (with basis in law) are applicable to this application.

• Public Interest

Acute coronary syndrome (ACS) which includes ST-elevation myocardial infarction (STEMI), non ST-elevation myocardial infarction (NSTEMI) and unstable angina (UA) comprises the leading cause of emergency hospitalisation in Europe, a leading cause of death and disability and have major impacts on health economies. NSTEACS represent over half of the cases of Acute Coronary Syndrome (ACS) of which NSTEMI account for around 50,000 National Health Service hospitalisations per year

Evidence from randomised controlled trials (RCTs) as well as guideline recommendations from the European Society of Cardiology (ESC) and the National Institute for Health and Care Excellence (NICE) support the use of different strategies according to risk status because allocation of a treatment strategy significantly reduces subsequent cardiovascular events

In place of generic treatment of all patient’s, stratified care has the potential to achieve cost-effective patient-centred treatment as well as improved outcomes.

The state of clinical practice for the management of patients with ACS falls short of the state-of-the-art based on evidence and guidelines, which recommend risk-stratified patient management. Earlier work has shown that there is considerable diversity of clinical practice across centres and countries.

The University of Leeds work also reveals that, for patients eligible for care, those who had fewer missed treatment opportunities had improved survival and that this varies geographically within England

Additionally, evidence from the United Kingdom (UK) (Myocardial Ischemia National Audit Project) MINAP registry has shown that more comprehensive treatment was associated with improved outcome

Moreover, early ‘failure’ of receipt of an evidence-based care opportunity was significantly associated with subsequent missed guideline recommended treatments. In the UK the adoption of ACS therapies, including the diffusion of primary PCI, lags behind other countries. Between 2004 and 2010 this was associated with over 11,000 avoidable deaths. It is probable, therefore, that earlier and more widespread adherence to evidence-based ACS therapies at initial presentation will result in better outcomes. What is more, the failure to apply major guideline evidence-based treatments, in those without contraindications not only increases preventable deaths from cardiovascular disease, but diminishes the cost effectiveness of therapies.

• Background/study design

The UK GRACE Risk Score Intervention Study (UKGRIS) aims to find out whether there is a difference in a patient’s health following an unstable angina attack or a heart attack if treated according to a hospital’s usual care or if treated using the GRACE risk score tool.

Hospitals were randomly assigned to either continue with their current usual care or use of the GRACE risk score tool for the care of their patients admitted with a suspected NSTEACS (type of unstable angina attack or heart attack). Sites were only recruited into the study if their current usual care is not the use of the GRACE risk score tool.

When a patient has had an NSTEACS, the NHS has care processes (such as medication, further tests or health guidance) which are recommended for the patient based on how severe their NSTEACS is. These are called Class 1 guideline recommended care-processes. To decide whether there is a difference in patient’s health the study will compare each arm of the study (usual care vs GRACE risk score) by looking at how many of these Class 1 guideline care-processes a participant receives.

The study will also look at whether the participant goes on to experience any cardiovascular related events such as cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation following on from their NSTEACS to see whether there is a difference between the two arms of the study. To look at these cardiovascular related events the study requires hospital episodes and mortality data.

The GRACE Score is a prospectively studied scoring system to risk stratify patients with diagnosed Acute Coronary Syndrome (ACS) to estimate their in-hospital and 6-month to 3-year mortality.

The co-primary objective for the UKGRIS cluster-randomised trial is to assess whether the application of the GRACE risk score on patients with urgent acute coronary syndromes without ST-segment elevation NSTEACS admissions results in better uptake of Class I guideline recommended care-processes, and reduces the incidence of a composite cardiovascular endpoint (cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation). Secondary objectives are to assess the difference in terms of:

1. unscheduled revascularisations within 12 and 24 months

2. duration of hospitalisation within 12 and 24 months;

3. patient-reported quality of life at 12 months;

4. the four individual components of the co-primary composite endpoint of cardiovascular events at 12 and 24 months

The UKGRIS study was funded by the British Heart Foundation in June 2016 and opened in March 2017. The funding support has been recently extended until December 2022. Recruitment ended on 31st December 2019 and follow up ended 31st December 2020. The University of Leeds Clinical Trials Research Unit, (CTRU) a UKCRC-registered trials unit, manages the trial and comprises a multidisciplinary team of data managers, trial managers, information systems support and statisticians. All of this data is pseudonymised.

This agreement relates to the second co-primary endpoint, and on secondary endpoints 1, 2 and 4. The data required to derive these endpoints are to be obtained through routine electronic health records.

Data from NHS Digital are critical to tracking the clinical follow-up of the patients who consented to take part. Other than a mailed postal questionnaire at 12 months to assess quality of life, The University of Leeds are not performing a clinical follow-up for hospital staff to review medical records to identify relevant clinical events, admissions and / or death information hence the University of Leeds are unable to perform the planned analyses of the UKGRIS trial without the timely delivery of the required NHS Digital data.

The UKGRIS study, as a stand-alone project is wholly confined within the United Kingdom. This research dataset will comprise self-reported questionnaire data, case report forms collected during the trial, and data from other national data collections, including HES data and mortality data. This data will be processed to create a research pseudonymised dataset without identifiers, such as date of birth, dates of events/episodes, and including derived variables such as age at registration, numbers of days since registration event/episode.

At the time of the original application, the UKGRIS study team were aware of their obligations under the Data Protection Act to not collect more data items than necessary. The items requested were therefore limited to those that UKGRIS believed would allow the study to derive and so analyse of the second co-primary outcome measure in a manner satisfactory for out trial. In the years since, UKGRIS have been made aware of efforts by other research teams to standardise analysis algorithms to derive cardiovascular adverse events from Office for National Statistics (ONS) mortality and HES-APC records. Discussions in confidence with these teams have informed the study that in order to be able to use such algorithms UKGRIS would require additional fields over and above those requested in our initial application. It is therefore necessary to request these additional fields so as to ensure that UKGRIS are able to analyse the UKGRIS trial dataset in a manner compliant with what UKGRIS anticipate to be the proposed standardised algorithms.

CTRU will be providing cohort details to NHS Digital including the following identifiers: participant ID number, NHS number, date of birth and postcode to allow for identification of participants and data linkage at NHS Digital. Data on this cohort will be returned to CTRU with all identifiers removed other than study ID number as a pseudonymised data set. This ensures that data is transferred between organisations with the minimum amount of identifiers as possible, whilst ensuring that the data is accurate and can be used for the analysis on a patient level at CTRU.

The UKGRIS study team are requesting data on all participants from 09/03/2017 (first participant recruited) to 31/12/2021 (2 years follow up for last participant) in order to report on secondary endpoints 1, 2 and 4.

Data is requested for participants from all hospitals taking part in the study, which is aligned with the geographical spread of data requested.

UKGRIS does not collect clinical data on participants post discharge and the planned analyses of the UKGRIS trial is not possible without NHS Digital data.

• Wider projects

The UKGRIS study team intend to also use MINAP and the National Audit of Cardiac Rehabilitation (NACR) datasets to complete missing data on adherence to guideline indicated therapies and hospital admissions. Data will be collated at patient level with analysis being performed on datasets from which public identifiers (including names, address information, NHS/Hospital identifiers) have been removed.

Planned secondary research in the UKGRIS trial includes deriving and analysing outcomes when using data provided from different routine data sources, including MINAP, Adult Cardiac Surgery, Adult Percutaneous Coronary Interventions and Heart Failure dataset. Results of deriving and analysing the same outcomes sourced from different datasets will be contrasted descriptively to those derived from the CRF data and from NHS Digital data sources.

The protocol has been harmonised with national studies undertaken with similar aims in Australia (AGRIS*) and Canada with the intention that CTRU will receive derived, de-identified datasets to perform an international collaborative individual participant data (IPD) meta-analysis. Doing so will allow the estimation of smaller differences in cardiovascular events observed during follow-up. No personal data will be received by CTRU from these countries. CTRU will be the recipient of data from the other countries and will not share any data with these countries. There will be no linkage of datasets received from other countries to the NHS Digital data set as the information received from other countries will be from a different set of participants (there is no overlap in data subjects between countries).

• Organisations involved

The sole Data Controller for UKGRIS is University of Leeds

The UKGRIS participating hospitals have sent personal participant data to CTRU

The UKGRIS international collaborators plan to send anonymised patient level data to CTRU

University of York will send link-anonymised patient level data to CTRU

CTRU will receive link-anonymised patient level data from HQIP

The British Heart Foundation has funded this study on the basis that The University of Leeds will be obtaining follow up data on the participants to answer the primary and secondary objectives.

It is expected that a final report will be submitted to the Funder (BHF) on 31/06/2022 which will report results of the study as per their template available online via the following link: https://www.bhf.org.uk/for-professionals/information-for-researchers/managing-your-grant/reporting-progress

After final analysis The University of Leeds plan a single publication comprising all outcome data on guideline uptake, cardiovascular outcomes, quality of life, revascularisations and hospital stay duration. The University of Leeds believe that the clinical question is of sufficient strength to submit in the first instance to a high-impact peer-reviewed general medical journal (e.g. New England Journal of Medicine, The Lancet, Journal of the American Medical Association, Annals of Internal Medicine, BMJ). The University of Leeds believe that these outputs would be of interest to a general medical community, hence their initial submission strategy.

Should The University of Leeds UKGRIS be unsuccessful with such journals, they will prioritise publication in high-impact cardiology-specific journals, (e.g. Journal of the American College of Cardiology, Circulation). In order to allow access to these outputs, The University of Leeds would either ensure that the outputs were open access, or that author manuscripts were available in depositories for access. (The British Heart Foundation mandates open access as a condition of funding).

The trial has a separate methodological sub study relating to rates of return of postal questionnaires, which may warrant a separate publication (or inclusion in the main trial output). As this does not require the use of electronic health records, this is not relevant to the present application, and so The University of Leeds do not give further details here.

Oral Presentation:

UK GRACE Risk Score trial: a parallel-group cluster randomised registry-based controlled trial. European Society of Cardiology Digital Congress August 2021

Journal Publication:

UKGRIS protocol paper published 05/09/2021, BMJ Open http://dx.doi.org/10.1136/bmjopen-2019-032165

• Dissemination of results/outputs

Results on 24 month long term follow up, which are dependent upon the receipt of NHS Digital data, will be prepared upon receipt of data with an anticipated publication date of Summer/Autumn 2022 (depending upon timelines for receipt of data).

The following publications are also planned:

o UKGRIS Primary outcome study results Spring/Summer 2023

o International GRACE IPD MA results Spring/Summer 2023

o Comparison of GRACE score randomised sites to non-participating GRACE sites for guideline uptake Spring/Summer 2023

o Effect of time to GRACE risk scoring on outcomes Spring/Summer 2023

o Cohort profile paper, suspected NSTEACS Spring/Summer 2023

o Subset analysis of main study paper Spring/Summer 2023

o Comparison of UKGRIS participants and general population of patients who experience an NSTEACS Spring/Summer 2023

Results will also be available on the CTRU and Funder websites with associated social media posts for these institutions.

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

The data received to date has been used to provide the Data Monitoring and Ethics Committee with an unblinded analysis in 2019. This aided a decision by the committee whether the trial should continue (or be stopped early on safety/futility grounds). The recommendation was to continue the study in order the answer the research question, which would ultimately be of benefit to patients and the NHS.

The UK GRACE Risk Score Intervention Study (UKGRIS) aims to find out whether there is a difference in a patient’s health following an unstable angina attack or a heart attack if treated according to a hospital’s usual care or if treated using the GRACE risk score tool.

The UKGRIS study will randomly assign recruiting hospitals to either continue with their current usual care or use of the GRACE risk score tool for the care of their patients admitted with a suspected NSTEACS (type of unstable angina attack or heart attack). Sites are only recruited into the study if their current usual care is not the use of the GRACE risk score tool.

When a patient has had an NSTEACS, the NHS has care processes (such as medication, further tests or health guidance) which are recommended for the patient based on how severe their NSTEACS is. These are called Class 1 guideline recommended care-processes. To decide whether there is a difference in patient’s health the study will compare each arm of the study (usual care vs GRACE risk score) by looking at how many of these Class 1 guideline care-processes a participant receives.

The study will also look at whether the participant goes on to experience any cardiovascular related events such as cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation following on from their NSTEACS to see whether there is a difference between the two arms of the study. To look at these cardiovascular related events the study requires hospital episodes and mortality data.

The GRACE Score is a prospectively studied scoring system to risk stratify patients with diagnosed Acute Coronary Syndrome (ACS) to estimate their in-hospital and 6-month to 3-year mortality.

The co-primary objective for the UKGRIS cluster-randomised trial, which plans to recruit 3000 patients from a minimum of 30 hospitals, is to assess whether the application of the GRACE risk score on patients with urgent acute coronary syndromes without ST-segment elevation (NSTEACS) admissions results in better uptake of Class I guideline recommended care-processes, and reduces the incidence of a composite cardiovascular endpoint (cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation). Secondary objectives are to assess the difference in terms of:

1. unscheduled revascularisations within 12 months

2. duration of hospitalisation within 12 months;

3. patient-reported quality of life at 12 months;

4. the four individual components of the co-primary composite endpoint of cardiovascular events.

This agreement relates to the second co-primary endpoint, and on secondary endpoints 1, 2 and 4. The data required to derive these endpoints are to be obtained through routine electronic health records.

The UKGRIS study was funded by the British Heart Foundation in June 2016 and is currently recruiting patients, having opened in March 2017. The University of Leeds Clinical Trials Research Unit, a UKCRC-registered trials unit, manages the trial and comprises a multidisciplinary team of data managers, trial managers, information systems support and statisticians. All of this data is pseudonymised.

Data from NHS Digital are critical to tracking the clinical follow-up of the patients who consented to take part. Other than a mailed postal questionnaire at 12 months to assess quality of life, The University of Leeds are not performing a clinical follow-up for hospital staff to review medical records to identify relevant clinical events, admissions and / or death information hence the University of Leeds are unable to perform the planned analyses of the UKGRIS trial without the timely delivery of the required NHS Digital data.

The UKGRIS study, as a stand-alone project is wholly confined within the United Kingdom. This research dataset will comprise self-reported questionnaire data, case report forms collected during the trial, and data from other national data collections, including HES data and mortality data. This data will be processed to create a research psuedonymised dataset without identifiers, such as date of birth, dates of events/episodes, and including derived variables such as age at registration, numbers of days since registration event/episode.

The protocol has been harmonised with a similar study in Australia (AGRIS* ) with the intention that the two trials (plus an additional planned Canadian study) will combine these derived, de-identified datasets to perform an international collaborative individual participant data (IPD) meta-analysis. Doing so will allow the estimation of smaller differences in cardiovascular events observed during follow-up. Derived data will only be shared with the international collaborators upon agreement with NHS digital that the data is sufficiently derived to satisfy the requirements of NHS digital. The Leeds Institute of Clinical Trials will work with NHS digital to reach agreement on the data to ensure that NHS digital agree that the data is derived sufficiently prior to being shared. A special condition which prevents any onward data sharing has been added to this agreement.

Whilst the main analysis for this study will commence once full patient recruitment has completed at the end of 2019 early 2020 data is being requested now from NHS Digital to allow the study to explore the methodology.

A download of data will be required upon approval so that the University of Leeds can present the data at the Trial Steering committee meeting

The British Heart Foundation has funded this study on the basis that The University of Leeds will be obtaining follow up data on the participants to answer the primary and secondary objectives. The University of Leeds provide them with 6 monthly updates on their progress.

The current timelines anticipate an end to recruitment by December 2019, with the final 12 months’ follow-up concluded by December 2020, after which analysis will begin to be complete by June 2021. Prior to that, the independent Data Monitoring Committee has requested an interim review of safety data available on the patients recruited so far, though no binding stopping rules will be in place.

A “Stopping rule” is a criterion (or set of criteria) that, if met, would indicate that the trial should be terminated early (that is, before the scheduled end of recruitment). Typical reasons for terminating early are futility (the final trial results are unlikely to show benefit to the intervention), harm (the intervention is leading to more adverse events, or fewer beneficial outcomes than the standard care arm) or efficacy (at this stage, the intervention benefit is such, that the study is already highly likely to have rejected its null hypothesis). In all of these cases, early termination avoids wasting the time of patients that have yet to be recruited (and avoids exposing them to unnecessary harm / lack of benefit shown in the data so far).

Stopping rules may be “Binding” or “Non-binding”/”Advisory”. Binding stopping rules are mandatory: if the criteria are met, the trial must be terminated early. Advisory stopping rules may be overruled by the DMEC if, on the balance of all the data available, there is still merit in continuing the study.

The study does not have any predetermined rules (Binding or otherwise) where the study would be stopped if the data showed a safety issue in either arm of the study. The data will be reviewed by the Data Monitoring and Ethics Committee and if they consider that it is not safe to continue with the study then it will be based on their medical and scientific expertise rather than any predefined stopping rule.

The Data Monitoring and Ethics Committee (DMEC) provide independent trial oversight and comprise independent members (that is members who are not co-applicants, not involved in trial conduct at the site level and not employees of the sponsoring or funding organisations and not from other participating organisations) with expertise relevant to the conduct of the study. The DMEC will review the safety and ethics of the trial by reviewing interim data during recruitment and will recommend on the continuation of the trial, or if any changes to the study design are indicated. Recommendations from the DMEC are made to the Trial Steering Committee, who make the final recommendation to the sponsor and funder that the study should be terminated early, if necessary. They will be provided with the following anonymised data, which is held in house at CTRU – having collected these on the case report forms - and are not requesting from NHS Digital;

• Aggregate recruitment by centre

• Aggregate screening data by centre by reason

• Aggregated numbers of deaths occurring before discharge and withdrawals

• Aggregated summary data on Protocol adherence

• Aggregate participant baseline characteristics (obtained from hospital completed data)

From the NHS Digital data download that is the subject of this application, It is expected to provide the following information to enable a single informal interim assessment (prior to completion of recruitment) of the safety profile of the GRACE and the Standard Care arms:

• Aggregate Numbers of post-discharge deaths (all-causes) in GRACE arm and the same in Standard care arm

• Aggregate Numbers of deaths (cardiac causes, pre and post discharge), in the GRACE arm and in the Standard care arm

• Aggregate Numbers of participants (and numbers of events) experiencing one or more of the component events of our composite clinical endpoint: Non-fatal myocardial infarction, New onset heart failure (admission), Cardiovascular readmission's (as well as cardiac cause of death), all for the GRACE arm, and separately for the standard care arm.

Details of the full set of data items to be downloaded for the final analysis is given in the original application.

After final analysis (expected completion June 2021) The University of Leeds plan a single publication comprising all outcome data on guideline uptake, cardiovascular outcomes, quality of life, revascularisations and hospital stay duration. The University of Leeds believe that the clinical question is of sufficient strength to submit in the first instance to a high-impact peer-reviewed general medical journal (eg New England Journal of Medicine, The Lancet, Journal of the American Medical Association, Annals of Internal Medicine, BMJ). The University of Leeds believe that these outputs would be of interest to a general medical community, hence their initial submission strategy. Should The University of Leeds be unsuccessful with such journals, they will prioritise publication in high-impact cardiology-specific journals, (eg Journal of the American College of Cardiology, Circulation). Ahead of publication, The University of Leeds would also seek to present their findings at a leading peer-reviewed international cardiology congress (eg European Society of Cardiology). In order to allow access to these outputs, The University of Leeds would either ensure that the outputs were open access, or that author manuscripts were available in depositories for access. (The British Heart Foundation mandates open access as a condition of funding).

The trial has a separate methodological sub study relating to rates of return of postal questionnaires, which may warrant a separate publication. (or inclusion in the main trial output) As this does not require the use of electronic health records, this is not relevant to the present application, and so The University of Leeds do not give further details here.

After final analysis the British Heart Foundation have agreed to an article on their website regarding the outcomes of the study.

The CTRU website will also be updated with information regarding the outcome of the study so this can be disseminated to the general public.

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

The yielded benefits section must be updated within any subsequent agreement

The UK GRACE Risk Score Intervention Study (UKGRIS) aims to find out whether there is a difference in a patient’s health following an unstable angina attack or a heart attack if treated according to a hospital’s usual care or if treated using the GRACE risk score tool.

The UKGRIS study will randomly assign recruiting hospitals to either continue with their current usual care or use of the GRACE risk score tool for the care of their patients admitted with a suspected NSTEACS (type of unstable angina attack or heart attack). Sites are only recruited into the study if their current usual care is not the use of the GRACE risk score tool.

When a patient has had an NSTEACS, the NHS has care processes (such as medication, further tests or health guidance) which are recommended for the patient based on how severe their NSTEACS is. These are called Class 1 guideline recommended care-processes. To decide whether there is a difference in patient’s health the study will compare each arm of the study (usual care vs GRACE risk score) by looking at how many of these Class 1 guideline care-processes a participant receives.

The study will also look at whether the participant goes on to experience any cardiovascular related events such as cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation following on from their NSTEACS to see whether there is a difference between the two arms of the study. To look at these cardiovascular related events the study requires hospital episodes and mortality data.

The GRACE Score is a prospectively studied scoring system to risk stratify patients with diagnosed Acute Coronary Syndrome (ACS) to estimate their in-hospital and 6-month to 3-year mortality.

The co-primary objective for the UKGRIS cluster-randomised trial, which plans to recruit 3000 patients from a minimum of 30 hospitals, is to assess whether the application of the GRACE risk score on patients with urgent acute coronary syndromes without ST-segment elevation (NSTEACS) admissions results in better uptake of Class I guideline recommended care-processes, and reduces the incidence of a composite cardiovascular endpoint (cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation). Secondary objectives are to assess the difference in terms of:

1. unscheduled revascularisations within 12 months

2. duration of hospitalisation within 12 months;

3. patient-reported quality of life at 12 months;

4. the four individual components of the co-primary composite endpoint of cardiovascular events.

This agreement relates to the second co-primary endpoint, and on secondary endpoints 1, 2 and 4. The data required to derive these endpoints are to be obtained through routine electronic health records.

The UKGRIS study was funded by the British Heart Foundation in June 2016 and is currently recruiting patients, having opened in March 2017. The University of Leeds Clinical Trials Research Unit, a UKCRC-registered trials unit, manages the trial and comprises a multidisciplinary team of data managers, trial managers, information systems support and statisticians. All of this data is pseudonymised.

Data from NHS Digital are critical to tracking the clinical follow-up of the patients who consented to take part. Other than a mailed postal questionnaire at 12 months to assess quality of life, The University of Leeds are not performing a clinical follow-up for hospital staff to review medical records to identify relevant clinical events, admissions and / or death information hence the University of Leeds are unable to perform the planned analyses of the UKGRIS trial without the timely delivery of the required NHS Digital data.

The UKGRIS study, as a stand-alone project is wholly confined within the United Kingdom. This research dataset will comprise self-reported questionnaire data, case report forms collected during the trial, and data from other national data cokllections, including HES data and mortality data. This data will be processed to create a research psuedonymised dataset without identifiers, such as date of birth, dates of events/episodes, and including derived variables such as age at registration, numbers of days since registration event/episode.

The protocol has been harmonised with a similar study in Australia (AGRIS* ) with the intention that the two trials (plus an additional planned Canadian study) will combine these derived, de-identified datasets to perform an international collaborative individual participant data (IPD) meta-analysis. Doing so will allow the estimation of smaller differences in cardiovascular events observed during follow-up. Derived data will only be shared with the international collaborators upon agreement with NHS digital that the data is sufficiently derived to satisfy the requirements of NHS digital. The Leeds Institute of Clinical Trials will work with NHS digital to reach agreement on the data to ensure that NHS digital agree that the data is derived sufficiently prior to being shared. A special condition which prevents any onward data sharing has been added to this agreement.

Whilst the main analysis for this study will commence once full patient recruitment has completed at the end of 2019 early 2020 data is being requested now from NHS Digital to allow the study to explore the methodology.

A download of data will be required upon approval so that the University of Leeds can present the data at the Trial Steering committee meeting

The British Heart Foundation has funded this study on the basis that The University of Leeds will be obtaining follow up data on the participants to answer the primary and secondary objectives. The University of Leeds provide them with 6 monthly updates on their progress.

The current timelines anticipate an end to recruitment by December 2019, with the final 12 months’ follow-up concluded by December 2020, after which analysis will begin to be complete by June 2021. Prior to that, the independent Data Monitoring Committee has requested an interim review of safety data available on the patients recruited so far, though no binding stopping rules will be in place.

A “Stopping rule” is a criterion (or set of criteria) that, if met, would indicate that the trial should be terminated early (that is, before the scheduled end of recruitment). Typical reasons for terminating early are futility (the final trial results are unlikely to show benefit to the intervention), harm (the intervention is leading to more adverse events, or fewer beneficial outcomes than the standard care arm) or efficacy (at this stage, the intervention benefit is such, that the study is already highly likely to have rejected its null hypothesis). In all of these cases, early termination avoids wasting the time of patients that have yet to be recruited (and avoids exposing them to unnecessary harm / lack of benefit shown in the data so far).

Stopping rules may be “Binding” or “Non-binding”/”Advisory”. Binding stopping rules are mandatory: if the criteria are met, the trial must be terminated early. Advisory stopping rules may be overruled by the DMEC if, on the balance of all the data available, there is still merit in continuing the study.

The study does not have any predetermined rules (Binding or otherwise) where the study would be stopped if the data showed a safety issue in either arm of the study. The data will be reviewed by the Data Monitoring and Ethics Committee and if they consider that it is not safe to continue with the study then it will be based on their medical and scientific expertise rather than any predefined stopping rule.

The Data Monitoring and Ethics Committee (DMEC) provide independent trial oversight and comprise independent members (that is members who are not co-applicants, not involved in trial conduct at the site level and not employees of the sponsoring or funding organisations and not from other participating organisations) with expertise relevant to the conduct of the study. The DMEC will review the safety and ethics of the trial by reviewing interim data during recruitment and will recommend on the continuation of the trial, or if any changes to the study design are indicated. Recommendations from the DMEC are made to the Trial Steering Committee, who make the final recommendation to the sponsor and funder that the study should be terminated early, if necessary. They will be provided with the following anonymised data, which is held in house at CTRU – having collected these on the case report forms - and are not requesting from NHS Digital;

• Aggregate recruitment by centre

• Aggregate screening data by centre by reason

• Aggregated numbers of deaths occurring before discharge and withdrawals

• Aggregated summary data on Protocol adherence

• Aggregate participant baseline characteristics (obtained from hospital completed data)

From the NHS Digital data download that is the subject of this application, It is expected to provide the following information to enable a single informal interim assessment (prior to completion of recruitment) of the safety profile of the GRACE and the Standard Care arms:

• Aggregate Numbers of post-discharge deaths (all-causes) in GRACE arm and the same in Standard care arm

• Aggregate Numbers of deaths (cardiac causes, pre and post discharge), in the GRACE arm and in the Standard care arm

• Aggregate Numbers of participants (and numbers of events) experiencing one or more of the component events of our composite clinical endpoint: Non-fatal myocardial infarction, New onset heart failure (admission), Cardiovascular readmission's (as well as cardiac cause of death), all for the GRACE arm, and separately for the standard care arm.

Details of the full set of data items to be downloaded for the final analysis is given in the original application.

After final analysis (expected completion June 2021) The University of Leeds plan a single publication comprising all outcome data on guideline uptake, cardiovascular outcomes, quality of life, revascularisations and hospital stay duration. The University of Leeds believe that the clinical question is of sufficient strength to submit in the first instance to a high-impact peer-reviewed general medical journal (eg New England Journal of Medicine, The Lancet, Journal of the American Medical Association, Annals of Internal Medicine, BMJ). The University of Leeds believe that these outputs would be of interest to a general medical community, hence their initial submission strategy. Should The University of Leeds be unsuccessful with such journals, they will prioritise publication in high-impact cardiology-specific journals, (eg Journal of the American College of Cardiology, Circulation). Ahead of publication, The University of Leeds would also seek to present their findings at a leading peer-reviewed international cardiology congress (eg European Society of Cardiology). In order to allow access to these outputs, The University of Leeds would either ensure that the outputs were open access, or that author manuscripts were available in depositories for access. (The British Heart Foundation mandates open access as a condition of funding).

The trial has a separate methodological sub study relating to rates of return of postal questionnaires, which may warrant a separate publication. (or inclusion in the main trial output) As this does not require the use of electronic health records, this is not relevant to the present application, and so The University of Leeds do not give further details here.

After final analysis the British Heart Foundation have agreed to an article on their website regarding the outcomes of the study.

The CTRU website will also be updated with information regarding the outcome of the study so this can be disseminated to the general public.

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

The UK GRACE Risk Score Intervention Study (UKGRIS) aims to find out whether there is a difference in a patient’s health following an unstable angina attack or a heart attack if treated according to a hospital’s usual care or if treated using the GRACE risk score tool.

The UKGRIS study will randomly assign recruiting hospitals to either continue with their current usual care or use of the GRACE risk score tool for the care of their patients admitted with a suspected NSTEACS (type of unstable angina attack or heart attack). Sites are only recruited into the study if their current usual care is not the use of the GRACE risk score tool.

When a patient has had an NSTEACS, the NHS has care processes (such as medication, further tests or health guidance) which are recommended for the patient based on how severe their NSTEACS is. These are called Class 1 guideline recommended care-processes. To decide whether there is a difference in patient’s health the study will compare each arm of the study (usual care vs GRACE risk score) by looking at how many of these Class 1 guideline care-processes a participant receives.

The study will also look at whether the participant goes on to experience any cardiovascular related events such as cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation following on from their NSTEACS to see whether there is a difference between the two arms of the study. To look at these cardiovascular related events the study requires hospital episodes and mortality data.

The GRACE Score is a prospectively studied scoring system to risk stratify patients with diagnosed Acute Coronary Syndrome (ACS) to estimate their in-hospital and 6-month to 3-year mortality.

The co-primary objective for the UKGRIS cluster-randomised trial, which plans to recruit 3000 patients from a minimum of 30 hospitals, is to assess whether the application of the GRACE risk score on patients with urgent acute coronary syndromes without ST-segment elevation (NSTEACS) admissions results in better uptake of Class I guideline recommended care-processes, and reduces the incidence of a composite cardiovascular endpoint (cardiac death, heart failure, new onset myocardial infarction, cardiovascular hospitalisation). Secondary objectives are to assess the difference in terms of:

1. unscheduled revascularisations within 12 months

2. duration of hospitalisation within 12 months;

3. patient-reported quality of life at 12 months;

4. the four individual components of the co-primary composite endpoint of cardiovascular events.

This agreement relates to the second co-primary endpoint, and on secondary endpoints 1, 2 and 4. The data required to derive these endpoints are to be obtained through routine electronic health records.

The UKGRIS study was funded by the British Heart Foundation in June 2016 and is currently recruiting patients, having opened in March 2017. The University of Leeds Clinical Trials Research Unit, a UKCRC-registered trials unit, manages the trial and comprises a multidisciplinary team of data managers, trial managers, information systems support and statisticians. All of this data is pseudonymised.

Data from NHS Digital are critical to tracking the clinical follow-up of the patients who consented to take part. Other than a mailed postal questionnaire at 12 months to assess quality of life, The University of Leeds are not performing a clinical follow-up for hospital staff to review medical records to identify relevant clinical events, admissions and / or death information hence the University of Leeds are unable to perform the planned analyses of the UKGRIS trial without the timely delivery of the required NHS Digital data.

The UKGRIS study, as a stand-alone project is wholly confined within the United Kingdom. This research dataset will comprise self-reported questionnaire data, case report forms collected during the trial, and data from other national data cokllections, including HES data and mortality data. This data will be processed to create a research psuedonymised dataset without identifiers, such as date of birth, dates of events/episodes, and including derived variables such as age at registration, numbers of days since registration event/episode.

The protocol has been harmonised with a similar study in Australia (AGRIS* ) with the intention that the two trials (plus an additional planned Canadian study) will combine these derived, de-identified datasets to perform an international collaborative individual participant data (IPD) meta-analysis. Doing so will allow the estimation of smaller differences in cardiovascular events observed during follow-up. Derived data will only be shared with the international collaborators upon agreement with NHS digital that the data is sufficiently derived to satisfy the requirements of NHS digital. The Leeds Institute of Clinical Trials will work with NHS digital to reach agreement on the data to ensure that NHS digital agree that the data is derived sufficiently prior to being shared. A special condition which prevents any onward data sharing has been added to this agreement.

Whilst the main analysis for this study will commence once full patient recruitment has completed at the end of 2019 early 2020 data is being requested now from NHS Digital to allow the study to explore the methodology.

A download of data will be required upon approval so that the University of Leeds can present the data at the Trial Steering committee meeting

The British Heart Foundation has funded this study on the basis that The University of Leeds will be obtaining follow up data on the participants to answer the primary and secondary objectives. The University of Leeds provide them with 6 monthly updates on their progress.

The current timelines anticipate an end to recruitment by December 2019, with the final 12 months’ follow-up concluded by December 2020, after which analysis will begin to be complete by June 2021. Prior to that, the independent Data Monitoring Committee has requested an interim review of safety data available on the patients recruited so far, though no binding stopping rules will be in place.

A “Stopping rule” is a criterion (or set of criteria) that, if met, would indicate that the trial should be terminated early (that is, before the scheduled end of recruitment). Typical reasons for terminating early are futility (the final trial results are unlikely to show benefit to the intervention), harm (the intervention is leading to more adverse events, or fewer beneficial outcomes than the standard care arm) or efficacy (at this stage, the intervention benefit is such, that the study is already highly likely to have rejected its null hypothesis). In all of these cases, early termination avoids wasting the time of patients that have yet to be recruited (and avoids exposing them to unnecessary harm / lack of benefit shown in the data so far).

Stopping rules may be “Binding” or “Non-binding”/”Advisory”. Binding stopping rules are mandatory: if the criteria are met, the trial must be terminated early. Advisory stopping rules may be overruled by the DMEC if, on the balance of all the data available, there is still merit in continuing the study.

The study does not have any predetermined rules (Binding or otherwise) where the study would be stopped if the data showed a safety issue in either arm of the study. The data will be reviewed by the Data Monitoring and Ethics Committee and if they consider that it is not safe to continue with the study then it will be based on their medical and scientific expertise rather than any predefined stopping rule.

The Data Monitoring and Ethics Committee (DMEC) provide independent trial oversight and comprise independent members (that is members who are not co-applicants, not involved in trial conduct at the site level and not employees of the sponsoring or funding organisations and not from other participating organisations) with expertise relevant to the conduct of the study. The DMEC will review the safety and ethics of the trial by reviewing interim data during recruitment and will recommend on the continuation of the trial, or if any changes to the study design are indicated. Recommendations from the DMEC are made to the Trial Steering Committee, who make the final recommendation to the sponsor and funder that the study should be terminated early, if necessary. They will be provided with the following anonymised data, which is held in house at CTRU – having collected these on the case report forms - and are not requesting from NHS Digital;

• Aggregate recruitment by centre

• Aggregate screening data by centre by reason

• Aggregated numbers of deaths occurring before discharge and withdrawals

• Aggregated summary data on Protocol adherence

• Aggregate participant baseline characteristics (obtained from hospital completed data)

From the NHS Digital data download that is the subject of this application, It is expected to provide the following information to enable a single informal interim assessment (prior to completion of recruitment) of the safety profile of the GRACE and the Standard Care arms:

• Aggregate Numbers of post-discharge deaths (all-causes) in GRACE arm and the same in Standard care arm

• Aggregate Numbers of deaths (cardiac causes, pre and post discharge), in the GRACE arm and in the Standard care arm

• Aggregate Numbers of participants (and numbers of events) experiencing one or more of the component events of our composite clinical endpoint: Non-fatal myocardial infarction, New onset heart failure (admission), Cardiovascular readmission's (as well as cardiac cause of death), all for the GRACE arm, and separately for the standard care arm.

Details of the full set of data items to be downloaded for the final analysis is given in the original application.

After final analysis (expected completion June 2021) The University of Leeds plan a single publication comprising all outcome data on guideline uptake, cardiovascular outcomes, quality of life, revascularisations and hospital stay duration. The University of Leeds believe that the clinical question is of sufficient strength to submit in the first instance to a high-impact peer-reviewed general medical journal (eg New England Journal of Medicine, The Lancet, Journal of the American Medical Association, Annals of Internal Medicine, BMJ). The University of Leeds believe that these outputs would be of interest to a general medical community, hence their initial submission strategy. Should The University of Leeds be unsuccessful with such journals, they will prioritise publication in high-impact cardiology-specific journals, (eg Journal of the American College of Cardiology, Circulation). Ahead of publication, The University of Leeds would also seek to present their findings at a leading peer-reviewed international cardiology congress (eg European Society of Cardiology). In order to allow access to these outputs, The University of Leeds would either ensure that the outputs were open access, or that author manuscripts were available in depositories for access. (The British Heart Foundation mandates open access as a condition of funding).

The trial has a separate methodological sub study relating to rates of return of postal questionnaires, which may warrant a separate publication. (or inclusion in the main trial output) As this does not require the use of electronic health records, this is not relevant to the present application, and so The University of Leeds do not give further details here.

After final analysis the British Heart Foundation have agreed to an article on their website regarding the outcomes of the study.

The CTRU website will also be updated with information regarding the outcome of the study so this can be disseminated to the general public.

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

Yielded Benefits is not a requirement for new applications.