Use and impact of the pre-hospital 12-lead electrocardiogram in the primary PCI era. Mixed method study (PHECG-2)

Kingston University is seeking permission to obtain Civil Registry mortality data linked to the MINAP (Myocardial Ischaemia National Audit Project) dataset (data collected as part of National Clinical Audit of Acute Coronary Syndromes). NHS Digital will act as a trusted third party to receive the MINAP data and link it with the mortality data. The linked data will be used to answer research questions for the study “Use and impact of the pre-hospital 12-lead electrocardiogram in the primary PCI era: mixed methods study (PHECG-2)”.

Kingston University are requesting only mortality data. The data supplied by NHS Digital under this Agreement is not personal data as defined by the GDPR.

For the data flowing from NICOR to NHS Digital for the purpose of linkage of MINAP data to Civil registration mortality data and the subsequent processing of linked datasets by the named data processing organisation (Swansea University), the following legal basis for processing in research in living humans apply:

Article 6 (1) e: processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller

Article 9 (2) j: processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject.

The data released to the research team will be pseudonymised; the dataset is expected to generate approximately 510,000 records thus risk of identifying an individual is highly unlikely.

The mortality data requested will enable Kingston University to answer the following research questions covered in Work Package 1 (WP1):

• For patients presenting to Emergency Medical Services (EMS) with suspected Acute Coronary Syndrome (ACS), is there a difference in 30- day mortality and reperfusion strategy in patients managed outside hospital by EMS between those who do and do not receive a pre-hospital electrocaradioram (PHECG)? (WP1)

• Has the proportion of eligible patients who received PHECG changed since the national rollout of primary percutaneous coronary intervention (pPCI) networks? (WP1)

There are 3 other work packages which do not require the use of mortality data but complete the study protocol:

• Are patients that receive a PHECG different from those that do not in terms of social and demographic factors, and in their pre-hospital clinical presentation? (WP2)

• What factors do EMS clinicians report as influencing their decision to perform a PHECG? (WP3)

• Synthesis of the findings (WP4)

The study is an independent, one off study that builds on the findings from the Quinn et all (2014). PHECG is not part of any other project or study. It is funded by British Heart Foundation (ref: PG/18/13/33558), the grant commenced in 2018. BHF will not be in any way involved in handling or processing or controlling this data. BHF will receive an end of study report containing results from the study in aggregated format.

For the purpose of this application, Kingston University is interested in long term patient outcome data, in this case mortality/survival at 30 days and one year after admission. The requested Civil Registry data will be used in the work package 1 (WP1) for mortality/survival analysis. Mortality data will be linked to patients with acute coronary syndrome that appear in the MINAP (Myocardial Ischaemia National Audit Project) database.

Thus the cohort consists of all patient admissions available in MINAP database between 1 January 2010 and 31 December 2017 with confirmed diagnosis of ACS.

The inclusion criteria for analysis are:

• Aged 18 or older

• Admitted with ACS to a participating hospital in January 2010 - December 2017

• Brought to hospital by ambulance

The Pre-Hospital 12-lead electrocardiogram (PHECG) is a simple test that helps ambulance clinicians assess patients with suspected acute coronary syndrome (heart attack), and helps to inform ongoing care, such as direct transfer to a specialist heart attack centre. All NHS emergency ambulances carry this equipment. This project builds on previous work by this research team, which found that one in three eligible patients did not receive a PHECG, but those that did had a lowered risk of short-term death. Women, the elderly and people with more complex health status were less likely to receive PHECG. The dominant treatment for heart attack at the time of earlier analysis carried out by Kingston University was 'clot buster' drug therapy (fibrinolysis). In this study Kingston University will update that work, in the context of the shift in recent years to a more interventional strategy for treatment of heart attack (angioplasty and stents), and explore reasons for variations in practice -highlighting opportunities to improve care and outcomes.

Using routinely collected data from a large national audit, a review of ambulance records, and qualitative methods, Kingston University will assess the association of having PHECG with patient outcomes, and research patient, practitioner and contextual factors contributing to the decision to record (or not) a PHECG. The research study team will aim subsequently to develop an intervention to increase the proportion of eligible patients that receive a PHECG, and to produce a proposal for further funding to test this intervention in a subsequent randomised trial.

The proposed research has important implications for the NHS and people's health. Heart attack remains a common condition and a leading cause of death. The role of the ambulance service has evolved in recent years to provide early diagnosis of heart attack using the electrocardiogram ('heart trace' or ECG). When the test is done in the ambulance it is termed 'prehospital ECG (PHECG). A systematic review of studies combining results of several studies - of which the study by Quinn et al was by far the largest- confirmed the benefits of having PHECG in terms of improved survival in patients who had the test in the ambulance.

Since the previous study the main treatment of heart attack has changed from using 'clot buster' drugs to using angioplasty and stents. Kingston University does not know in this contemporary setting whether the ambulance ECG is being done in all patients who might benefit, or whether it is still associated with improved outcomes. Kingston University also know very little about how paramedics make decisions about whether or not to record an ECG on a patient.

Findings from the PHECG-2 study will provide important information about the use of the ECG test in contemporary heart attack care. It will also provide insights that have not been studied before about the wider factors (other than the purely clinical data from the national audit) that might influence whether a patient gets an ECG or not while under the care of the ambulance service.

Should the results show that there is still a proportion of heart attack patients who are not getting the ECG test, Kingston University hopes to be able to use the insights from their research to develop (and later test in another study) an intervention to narrow this gap, benefiting more patients.

Kingston University require Civil Registry mortality data including cause of death linked to MINAP dataset for the purpose of patient outcomes analysis, specifically 30 day and one year post admission all-cause mortality and cause of death.

Only pseudonymised data are required for this study. A defined cohort from the MINAP data will be linked to Civil Registry mortality data by NHS Digital and the relevant study team members at Swansea University will receive both datasets in pseudonymised form. Patient level data are required in order to apply risk adjustment when calculating mortality rates to make the results as comparable as possible.

The study requires data for patients from 2010 to 2017 inclusive. This will allow Kingston University to look at trends over time, increase the statistical power and cover the period when the primary angioplasty (or primary PCI – preferred treatment and system to treat patients with acute form of a heart attack as fast as possible) has been rolled out across England and Wales and to make analyses as contemporary and thus as relevant as possible.

MINAP collects data for patient in England, Wales and latterly in Northern Ireland. Kingston University are applying for data in England and Wales only as the mortality data are only available for these two countries. By including data from both countries, the results will be more generalisable compared to focusing on e.g. England only.

Kingston University confirm that they do not know of any other less intrusive way of achieving the purpose of the study in relation to WP1 which is the subject of this application.

Kingston University have requested pseudonymised mortality data linked to MINAP data by NHS Digital is sent to Swansea University who are the data processor thus eliminating the need for the research team to handle patient identifiable data related to this request.

Kingston University Faculty of Health, Social Care and Education, Kingston University & St George's, University of London are a joint faculty. However, in the context of this Agreement, the sole Data Controller is Kingston University, the sponsor organisation for this study. St George's, University of London will not determine the purpose of this study or the manner in which the data will be used. The data will be processed by the study statistical team based at Swansea University. No other organisations will be processing the data

The following three organisations are involved in WP2 and WP3:

- West Midlands Ambulance Service University NHS Foundation Trust

- South West Ambulance Service NHS Foundation trust

- Welsh Ambulance Services NHS Trust

The above three organisations are participating sites in the WP2 whereby a research paramedic identified at each participating site will be abstracting already collected data held by the ambulance Trust from their local patient record forms (PRF) on a sample of patients generated from the records in the MINAP dataset. None of the above organisations will have access to Civil Registry mortality data or MINAP data with exception of ambulance job number which enables them to link the selected records to their local PRFs.

For the WP3, the above organisations will help with the recruitment of EMS clinical staff into focus groups, assist with identification of suitable venues and other tasks related to the organisation of the focus groups.

The data flow in respect of the data linkage of MINAP data with Civil Registration mortality data is illustrated below:

1. NICOR will generate the pseudonymised MINAP dataset/cohort consisting of all records with admission dates from 1 January 2010 up to 31 December 2017 inclusive with clinical data and send to Swansea Trials Unit to the statistical team. The pseudonymised MINAP dataset will include sex and ethnicity data items as Section 251 support is in place for these data items to be released as part of the analysis dataset provided to Swansea University.

2. NICOR will send a separate file of the same cohort with patient identifiable including study ID, NHS number, date of birth, gender and postcode to NHS Digital for the purpose of linkage to Civil Registry morality data. NHS Digital will perform the linkage, remove all identifying information and will send the mortality data with NICOR unique record IDs to the statistical team at Swansea Trials Unit. NICOR will not receive the mortality data linked to MINAP.

3. Statisticians at Swansea Trials Unit will ‘link’ both MINAP clinical dataset and the file with pseudonymised mortality data from NHS Digital using a common unique record identifier that identifies the records but not patients.

No special category data will be sent to NICOR or the research team other than the linked mortality data will be leaving NHS Digital for the purpose of this application. NB: mortality data will only be sent to Swansea University (data processor).

Data related to this application will be processed by the statistical team at Swansea University only.

No further flow of data will take place following the receipt of the Civil Registry mortality data with pseudo-identifiers by the statistical team at Swansea University (data processor) who will undertake the analyses.

Civil Registry mortality data will be linked with pseudo-identifiers by the study statistical team. The study statistical team will use the linked dataset to perform the following analyses:

*30-day mortality (primary outcome) - the proportion of patients who die within 30 days of the date of their ACS event

* Time-to-death - the length of time in days from date of ACS event to death. Time-to-death will be summarised using Kaplan-Meier survival curves.

* In-hospital mortality - the proportion of patients in each group who die during the (initial) hospital stay following the date of their ACS event.

*One-year mortality - the proportion of patients in each group who die within one year of date of their ACS event.

Specified MINAP data will be linked to Civil Registry mortality data by NHS Digital. NICOR will send a file with relevant patient identifiers to NHS Digital via a secure facility agreed by both parties for the linkage including relevant pseudo-identifiers. Subsequently non identifiable MINAP clinical data will be sent to the study team for the same cohort. NHS Digital will perform the linkage and include MINAP pseudo-identifiers in the file that will be send to the study team with the attached date of death/vital status and cause of death. This way NICOR does not need to handle Civil Registry mortality data and the study team does not need to handle patient identifiers. The two datasets will be linked using pseudo-IDs by the research team for the analysis.

Kingston University only require a one-off linkage for the purpose of this study. There might be a need for a repeat of the linkage in the unlikely event that the original MINAP dataset was incomplete. The re-identification is highly unlikely as the MINAP dataset is likely to generate approximately 510,000 records. In addition, with the use of pseudo-identifiers and no access to reverse the process, the risk of patient identification is mitigated. The data from ambulance patient record forms will be submitted to the study research team in an anonymised form i.e. any reference to the patient or any member of EMS staff will be anonymised by the research paramedic prior to submission to the study statistical team.

There neither is nor there will be any requirement or attempt to re-identify the individuals.

The data processing will be carried out by substantive employees of the data processor which has a signed collaborative agreement with the data controller. All employees have been appropriately trained in data protection and confidentiality.

The information security arrangements have been reviewed and approved by NHS Digital as part of a CAG review for the purpose of section 251 application.

The results will be published in variety of peer-reviewed journals and presented at different conferences (either oral and/or poster presentation).

Kingston University will produce a report for the funder, REC and NICOR/HQIP and will develop an infographic(s) and/or other lay summaries of the research which they also plan to share with participants in the research, participating and non-participating ambulance Trusts and the National Ambulance Research Steering Group. Kingston University will also use a variety of PPI platforms and networks through PPI colleagues, INVOLVE and British Heart Foundation (the funder of this study) to reach out as many members of the public and patients as possible. The results will also be available on Kingston University website.

Some of the publications in peer reviewed journals might be publicised through press releases, most likely in conjunction with the funder (BHF) and collaborating institutions to further raise the study's profile.

The data will be reported in aggregate format; small numbers are unlikely given the size of the dataset. However, in an unlikely event that issue of small numbers should occur, these will be suppressed in line with the HES analysis guide.

The study Steering Committee has approved a Publication Charter of which role is to:

• ensure the timely creation and publication of high quality abstracts, presentations, and manuscripts based on the PHECG-2 study.

• solicit, review, refine and prioritize projects and analysis plans following submission of suggestions for outputs arising from the research.

• define timeline, track progress and determine peer-reviewed destination for each output (and within which of the following types of academic works: abstract, presentation, and manuscript)

• review all abstracts/presentations/manuscripts prior to submission/publication to ensure quality. (‘Safeguarding the PHECG-2 brand.’)

Kingston University aim to publish the results from the study within one year from the end of the study.

Kingston University are also required to produce a progress report to the REC 13 months from the date the ethical approval was given, and a summary of the final report on the research will be provided to the REC within 12 months of the conclusion of the study.

The study is funded until 30 November 2019.

The sponsor’s institution has an open access policy for the dissemination of findings. The funder also has separate funding arrangements that form part of standard terms and conditions of the grant covering the cost of open access publishing.

Data will not be shared with third parties.

Understanding the clinical and non-clinical factors influencing EMS clinicians’ decisions to record a PHECG will enable Kingston University to develop (and later test through in a randomised trial) an intervention with the potential to improve PHECG uptake and patient outcomes following an ACS event. With over 85,000 admissions with acute myocardial infarction recorded in MINAP each year, and millions worldwide, the potential for this research to improve outcomes nationally and internationally is substantial.

With the improved uptake of PHECG, more patients are likely to benefit from having PHECG recorded and as such have improved clinical outcomes following their episode of acute coronary syndrome.

The results will be shared via peer-reviewed journals, conferences, made publicly available to anyone with interest in this area of care to inform national and international guidelines and practice. In addition, Kingston University intend to develop an intervention for testing to improve the uptake of PHECG for all that might benefit from it.

The results will be shared via peer-reviewed journals, conferences, made publicly available to anyone with interest in this area of care to inform national and international guidelines and practice.

National and international guidelines for heart attack care, and education of paramedics and other relevant health professionals are informed by the published scientific literature; the evidence base. Thus publication in peer-reviewed journals is a key lever of research impact on patients and the public. Work in this area previously published by the study research team was influential in changing national and international guidelines.

Improvements in heart attack care in people presenting to ambulance services. The precise nature will not of course be known until the research has been undertaken.

The previous study by Quinn et al (2014) reported that approximately a third of patients with heart attack who came by ambulance did not receive PHECG and had worse outcomes than those who did. Optimising use of this simple test has the potential to save lives following heart attack.

Guideline implementation will be at local level, informed by national and international guidelines. Individual clinician behaviour may be influenced by the overall study findings (e.g. a paramedic’s threshold for performing PHECG in the elderly, women or co-morbid patients). This will of course depend on study findings.

The benefits can be measured through local audit (e.g. for Ambulance Quality Indicators), national audit (MINAP) and international (which may be through national registries)

When it will be achieved depends on study findings and any subsequent guideline recommendations to inform practice.

References:

Quinn T, Johnsen S, Gale CP, et al. Myocardial Ischaemia National Audit Project (MINAP) Steering Group. Effects of pre-hospital 12-lead ECG on processes of care and mortality in acute coronary syndrome: a linked cohort study from the Myocardial Ischaemia National Audit Project. Heart 2014;100:944-50

Yielded Benefits is not a requirement for new applications.