Optimising Management of Patients with Heart Failure with Preserved Ejection Fraction in Primary Care (OPTIMISE HFpEF)

This Data Sharing Agreement permits the University of Cambridge to retain and reuse NHS Digital data, supplied under an earlier version of this Agreement, for the purpose of the following research project: Optimising Management of Patients with Heart Failure with Preserved Ejection Fraction in Primary Care (OPTIMISE HFpEF).

As the protocol publication about this study explains (Link: https://www.repository.cam.ac.uk/bitstream/handle/1810/297027/bjgpopen19X101675.full.pdf?sequence=3&isAllowed=y), Heart Failure with preserved Ejection Fraction (HFpEF) is less well understood than Heart Failure with Reduced Ejection Fraction (HFrEF)and is associated with greater diagnostic difficulty and management uncertainty. Half of all HF cases may be attributable to HFpEF, and prevalence is rising at a rate of 1% annually. Although mortality for all-cause HF in the UK has modestly improved, no treatment has yet been shown to improve mortality and morbidity in HFpEF. Lack of evidence for pathophysiological mechanisms underpinning the disease, effective pharmacotherapies, and disease management programmes specifically targeting HFpEF hamper progress. The OPTIMISE HFpEF project aims to explore the views of people with HFpEF and the multiple stakeholders involved in HFpEF care; phenotype a UK cohort; and undertake consensus methods to develop an optimised programme that would provide guidance to clinicians in diagnosing and managing HFpEF.

Part of phenotyping a cohort requires exploration of hospitalisation and healthcare utilisation, as understanding the reasons for hospitalisation may mean that healthcare professionals can intervene to prevent some of these. Therefore, part of the study involves exploring all-cause hospitalisation. The 152 patients in this cohort are older, have multiple comorbidities and data from other studies suggest that hospitalisation is as frequent for comorbid conditions as it is for heart failure exacerbation. Furthermore, hospitalisation is associated with high levels of readmissions and high mortality in some studies, and it would be useful to understand the factors associated with outcomes in this unique community cohort. Exploring hospitalisation is achieved in two ways 1) consultation of the participant and 2) review and extraction of hospitalisation data from their general practice record. However, both of these methodologies carry a high risk of inaccuracy (for example, length of hospital stay may not be recalled by participants and there will always be a lag time between discharge and GP record update, such that it may be missed at the record review points).

Returned pseudonymised data will be linked with the clinical database at the University of Cambridge for the period of the Agreement up until 30 November 2022. Returned pseudonymised data will be destroyed after that date.

The General Data Protection Regulation state that “Personal data shall be adequate, relevant and limited to what is necessary in relation to the purposes for which they are processed.” Following consent >400 data fields in 152 subjects has been collected in a face to face clinical interview and assessment (recruitment has now concluded). This will enable thorough characterisation of a community dwelling cohort of people with HFpEF. It is important to understand hospitalisations and health care use in the cohort, as this is expected to be high. Whilst every effort was made to collect this data during the face to face sessions and medical record review, this relies on memory and recording which can be unreliable, especially if the event was not in the recent past or the hospital visit did not generate a discharge summary for the general practice Patients are not always aware of their discharge diagnoses, and patients with HFpEF experience hospitalisations due both to heart failure and their other comorbidities. Therefore an accurate record of reasons for hospitalisation is necessary as this helps understand targets for intervention. Healthcare usage in HFpEF is variable as admissions are often related to co-morbid conditions rather than directly associated with HFpEF. For this reasons, both accident and emergency data and admitted patient care data is requested. It is also clear that many patients with HFpEF go undiagnosed experiencing multiple hospitalisations before they obtain the correct diagnosis, this is important to investigate within the UK. Admitted patient care data is important also as this will allow for exploration of length of stay as HFpEF may exacerbate co-morbid condition related hospital stays.

The University of Cambridge requested the data under GDPR provisions under Article 6(1)(e) which states the “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”. It is in the public interest to characterise this cohort of participants given the large proportion of people who suffer from HFpEF and the relevant scant information available on how best to identify and manage them. Linkage, processing and storing will also be made under GDPR provisions set out in Article 9(2)(j) which states “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.” University of Cambridge has a track record of excellent health research and have staff and the necessary expertise to undertake this role in the public interest.

Previous publications indicate that people with HFpEF experience multiple hospitalisation and often have many outpatient visits due to their multi-morbidity. Once hospitalised, readmission rates are high and mortality is increased. Linkage as set out above will enable characterisation of a community dwelling sample in order to improve care to provide essential information to primary care and specialist services. Having this information will improve diagnosis and management, and enable development and testing of interventions to improve outcomes. The request is for the minimum amount number of datasets and products necessary to capture this pattern of healthcare use and to determine accurate mortality rates.

This research is necessary as much of the information about HFpEF comes from other countries with different healthcare systems and from clinical trials of medications. It is widely acknowledged that clinical trials are very selective and do not often represent the clinical reality, therefore knowledge of HFpEF may be based on highly specific sub populations and not generalisable to the real world clinical care. In the UK there is evidence that patients with HFpEF are not supported by specialist services due to clinical commissioning restrictions which impacts on diagnosis, care and ultimately mortality. Previous research has also shown that clinicians are not very familiar with HFpEF, and often conflate this with a different type of heart failure (heart failure with reduced ejection fraction) which has very different clinical management strategies, some of which may be deleterious in HFpEF. This can lead to multiple hospitalisations and poor outcomes.

The programme of research to which this application pertains is exploring all of these factors and will take the learning from this to develop a new model and system of care that will be co-designed with HFpEF patients. There are five work packages (WP) that explore what currently happens in clinical practice (WP1 and WP2a), what patients with HFpEF are like in terms of physical function, health status and quality of life (WP2b), what things are like the relatives and carers of those with HFpEF (WP2c). All of these will be analysed and interrogated to establish problematic areas and potential new methods of working that will be assessed via consensus work in the final project (WP3). The programme has been designed to be comprehensive, inclusive and thorough. Hospitalisation data is requested as part of WP2b but will inform understanding across the work packages and any new ways of working proposed in WP3. A detailed description of the full programme of research has been published and can be accessed here: https://www.repository.cam.ac.uk/bitstream/handle/1810/297027/bjgpopen19X101675.full.pdf?sequence=3&isAllowed=y

The dataset has been minimised by recruitment, geography and time, however it is also minimised by careful selection of data products. For example, date of admission is requested but time is not; diagnosis is requested but assessments are not. Data return will be in pseudo-anonymised format. Every effort was made to reduce datasets requested via patient interview and medical record review, however an accurate picture of healthcare use via these methods is not possible as both a person’s memory and healthcare records are often incomplete. The data is already narrowed by geography based on the recruiting sites, Cambridge which recruited from the Cambridgeshire region, Oxford which recruited from the Oxfordshire region. A small time frame has been set: hospitalisation/s periods over one year was considered the smallest duration possible that would also capture the fullest picture and provide useful information for practice.

The data controller is University of Cambridge, who also process the data. University of Oxford are cohort contributors. Guys and St Thomas’ have recruited 18 patients to this study. There was a secure data transfer system in place and those organisations transferred their identifiable patient data in this way. Patient identifiable data from Guys is stored at Cambridge as per the ethics approval and study agreements and these participants will be included in the Cambridge cohort. Neither Oxford nor Guys are carrying out data controllership activities.

Optimising Management for Patients with Heart Failure with Preserved Ejection Fraction in Primary Care (Optimise HFpEF) is a collaborative programme of research involving four Universities (Cambridge, Oxford, Manchester, Keele). It is led by the University of Cambridge and the other sites are involved in various components (work packages) of the research. However only the University of Oxford is involved in the longitudinal cohort study, which is the clinical dataset for data linkage. Oxford recruited patients, conducted the baseline and follow-up assessments, and hold the patient identifiable data for their cohort at their site. Their funding was for their participation in the cohort study. The other two Universities, Manchester and Keele, were involved in qualitative research only as part of the programme of research but were not involved in the cohort study nor this data linkage. The University of Cambridge investigators are leading the analysis and are responsible for the overall programme of work. All papers and reports that come from the cohort study and linked aggregated data from NHS Digital will be reviewed and agreed by the relevant investigators at the sites involved, but analysis will be conducted by the University of Cambridge.

This study is jointly sponsored by University of Cambridge and Cambridge University Hospitals NHS Foundation Trust. Decisions regarding data processing are made solely by University of Cambridge and University of Cambridge are the data controllers. Cambridge University NHS Foundation Trust was included as a joint sponsor as the respective research and development departments work together to review and advise on clinical studies and face to face study visits were conducted within the Cambridge Clinical Research Facility which is located within Addenbrooke’s Hospital.

NHS Number and date of birth were provided to NHS Digital to enable linkage of data by each of the two sites holding identifiable data. Both sites were also sent postcode and gender. All participants have provided informed consent. Data also includes the study ID so that returned linked data could be pseudonymised using only the study ID.

The products requested include civil registration (deaths) data, HES A&E attendance (as well as Emergency care data set), HES admitted inpatient care , all of which are health data. University of Cambridge have requested pseudonymised record level data so that these data can be analysed within the extensive database that the University of Cambridge hold on patients by study ID. The research team want to be able to compare patients who are hospitalised with those who are not hospitalised to determine the most at-risk group and the characteristics that may differ between them. The University of Cambridge will also be able to determine factors associated with readmissions or multiple hospitalisations.

Linked data would not be transferred between sites. Data would only be shared in an aggregate anonymous format between Oxford and Cambridge, with small numbers suppressed in line with HES analysis. The analysis will be conducted at the University of Cambridge so there is not a reason for Oxford to access the raw NHS digital data. The University of Oxford will provide the patient identifiable data to allow linkage with their cohort, but the pseudonymised data will be held at the University of Cambridge. In this way, the University of Cambridge serves as the data processor and controller. This is consistent with the study set-up.

The flow of data would involve University of Oxford and University of Cambridge securely transferring NHS number and date of birth to NHS digital to link records. Linked pseudonymised data from NHS Digital would be transferred to the University of Cambridge. Data would be shared only in anonymous, aggregate format between the sites. As data are pseudonymised and only linked by study ID, all data should return to the University of Cambridge. The chief investigator of the study, who is a substantive employee of the data controller (University of Cambridge), will make decisions about the data and direct the analysis; however all investigators will be consulted. University of Oxford will participate in the analysis by reviewing drafts of papers and reports produced, they will not have direct access to the pseudo-anonymised return.

Both the University of Cambridge and the University of Oxford maintain separate securely hosted files of patient identifiable data (PID) for their cohorts. Each site will send patient study ID, NHS numbers, DOB, gender and postcode for their cohort to NHS Digital (as per the consent forms/material). Data for both cohorts (civil registration, hospitalisation, A& E, ECDS) will be sent back to the University of Cambridge in a pipe delineated pseudonymised file using only Study ID. These data will be combined with the clinical data so that the researchers can describe patient outcomes in aggregate form, compare differences in patients by specific outcomes (e.g. hospitalisation), and determine factors associated with outcomes.

Data will not be linked with other external data sets, only with the University’s own study database using study ID. There is no risk of re-identification as no further linkage will occur: only aggregate data with small numbers suppressed (as per HES analysis guidelines) are being used in descriptive publications. PID and patient clinical information are kept separate and the University of Cambridge are interested in the aggregate outcome. The University of Cambridge will make no attempt to reidentify any individuals.

Data will be presented in aggregate for the full cohort which is geographically dispersed, and small numbers will be suppressed, as per HES analysis guidelines. The research team are interested in aggregate data and not in individuals, therefore there will be no identification of individual patients.

Only the Chief Investigator (CI) will have access to the pseudo-anonymised data returned from NHS Digital to University of Cambridge. The CI will perform the data analysis relating to hospitalisations. The pseudo anonymised database will be held in password protected computers and accessible only by the investigators and their delegated representatives.

Patient identifiable data for each cohort are held in the Secure Data Hosting Service (SDHS) at each University . SDHS provides a dedicated network, separated from the production network by a firewall, for storing sensitive personal data and hosting computers involved in its management and analysis. All equipment connected to the SDHS will be located in the Clinical School Computing Service's physically secure server rooms.

Research group applications to store Sensitive Personal Data must be made on a per study basis, whereupon the data flows will be checked to make sure they are appropriate. Once approved, data are migrated to the SDHS network and access is provided by a secure Virtual Desktop.

The Secure Data Hosting Service (SDHS) provides an ISO:27001 certified Safe Haven for members to store sensitive data, including Personally Identifiable Data. The service is managed by the Clinical School Computing Service (CSCS) in collaboration with the Information Governance Office on behalf of the School. The SDHS offers a logical network behind a firewall and secure file storage which is accessed via a browser-based Virtual Desktop. Data on the site can be viewed and edited from this remote desktop. It is not possible to copy the data, and it is not possible to use applications on the secure data. To access the SDHS users must:

- Have been approved in writing by the Study’s Data Manager

- Read the SDHS security policy

- Signed the SDHS acceptable use policy

- Configured their account with a 15 character password

- Received their 2-factor authentication token

Therefore access is restricted only to those with the above approvals in place. Within the SDHS University of Cambridge hold patient identifiable data (name, address, contact information, NHS and hospital record number) on a bespoke Access Database specifically designed for the study. The database is where all personal identifiable information is stored and it also functions as a record of contacts with study participants. The NHS Digital data returned to the research team would be stored within a separate folder within SDHS and not linked to the patient identifiable data. All clinical data collected as part of the study is stored separately in an anonymous format within a REDCAP database.

All risks are reviewed annually in February by information governance and technical staff prior to approval by the Council of the School and renewal of the NHS DSP Toolkit in March. Currently the School of Clinical Medicine is registered as approved, 'Standards Met'.

The University of Cambridge will use the data to produce at least one peer-reviewed journal paper with submission to journals such as Heart, European Journal of Heart Failure, European Journal of Cardiovascular Nursing, British Medical Journal and British Journal of General Practice. The University submitted a final report of the study to the National Institute for Health and Care Excellence (NIHR) School for Primary Care Research on 23 September, and has produced a research briefing with a link on the study website [Optimise HFpEF Study (cam.ac.uk)] or available here: Research Briefing: diagnosis and management problems for patients with hidden heart failure (adobe.com) However, both the final report and the research briefing refer only to the baseline clinical data from the cohort (plus the other studies in the Optimise HFpEF programme) and not to follow-up data including the NHS Digital linked data. The study team is currently analysing the data and plan to produce 1-2 papers using these data by March 2022.

Abstracts will be submitted abstracts to conferences such as European Society of Cardiology congress, EuroHeartCare, British Society for Heart Failure, and British Cardiovascular Society. The following may be produced:

• Reports to Grant Awarder (NIHR SPCR This will be a follow-up to the final report submitted 23 September.

• Submissions to peer reviewed journals

• Presentations

• Conferences

The University of Cambridge will facilitate the dissemination of the research and/or scientific work to stakeholders during the project and after its completion. The project has already published the protocol and has published papers.

The dissemination activities will target an audience of researchers, scientists and policy makers. Separate reports will be made for research participants with the help of a patient advisory group. Activities should also reach beyond the scientific community to engage with policy makers. Potential dissemination channels will include: journals, workshops, webinars, social media, public reports, co-hosted events.

The cohort is unique in the UK as a community-dwelling sample of patients with HFpEF recruited from primary care. These patients are often un-diagnosed and the condition is not well understood by many clinicians. However the prevalence of HFpEF is increasing, and it’s important to understand the clinical characteristics, problems and outcomes of this patient group, and what factors affect outcomes. The research’s findings with the information from NHS Digital will help the University of Cambridge to not only characterise this patient group but provide needed information on outcomes that will enable us to develop interventions and optimal management for this group. A final component of this programme of research is to build consensus related to best methods to diagnose and manage this group and develop interventions to test in practice. Having robust data to present to clinicians (GPs, cardiology specialist nurses and consultants, commissioners) is essential. The dissemination is through peer-reviewed publications and presentations as detailed above, reports, and through consensus-building activities as part of the research programme.

The study has a website and the research group a twitter account to aid dissemination. The study featured on the BBC Look East and the qualitative paper was highlighted through an interview with lead author and a patient and carer on the Naked Scientist. The University of Cambridge have produced one newsletter so far and plan to revise the research briefing to make it more patient friendly and disseminate to participants The data and papers produced from the data linkage with NHS digital are part of the information and evidence generated from the overall programme of research. Cambridge have also produced numerous abstracts for national (Society for Academic Primary Care) and international (European Society of Cardiology Congress and Preventive Cardiology Congress, and EuroHeartCare) from baseline data and other components of Optimise HFpEF. Thus, the University of Cambridge have a good track record of dissemination and publication and the plans for the NHS Digital linkage dataset are as follows:

Analysis with statistical support began as soon as data are returned from NHS Digital, and will be combined with our data from baseline, 6 and 12 month follow-ups. Although much of the analysis has been done, some work remains along with the drafting of the paper. The outcomes paper is estimated to be finalised in early 2022 and data will be reported as aggregated data only. A report for the funder was completed for the funder in September 2021 but focused on the baseline data. University of Cambridge will tweet, blog and disseminate the outcomes paper through presentations and a brief patient report posted on the study website.

GPs find that HFpEF patients are challenging to diagnose, and are often unrecognised in practice. Previous research from University of Cambridge has shown that practices often don’t have the information needed to diagnose HFpEF in patients: https://bjgpopen.org/content/2/3/bjgpopen18x101606/tab-article-info. Another component of this research has been qualitative research with clinicians, patients and carers, and the major themes are related to diagnostic difficulty, unclear illness perceptions and management disparity. Two papers have been published from the qualitative data (see below) and two are in draft

Having robust outcome data will help provide the impetus to improve awareness and care. It is anticipated that the cohort study will help illuminate the holistic picture of patients. Baseline analysis has shown that they are have high levels of obesity and comorbid conditions, are very functionally impaired (slow gait speed and limited distance walked in six minutes), experience significant frailty and have many symptoms such as breathlessness, fatigue and dizziness to name a few. Few were referred to cardiac rehabilitation, and most had not been supported in heart failure self-care, such as monitoring for weight gain due to fluid retention.

No commercial exploitation is anticipated given the topic of investigation. A Patient Advisory Group gives us input into results and how best to communicate with patients. University of Cambridge will also work with the Pumping Marvellous patient charity to provide information to patients. In the final work package the study team conducted a consensus survey about changes to practice, and held a workshop with healthcare providers and members of the public about next steps in improving care for patients with HFpEF in September. A paper is being written regarding the findings of the survey. Both the survey and the workshop are contributing to development of a systems approach to improving the diagnosis and management of patients with HFpEF in primary care.

• Protocol Paper: published 2019

• Baseline Papers from cohort: Two papers published in September 2021

• Physical Activity Paper from cohort: published May 2021.

• 12 Month Follow-up Paper: in process, target date no later than March 2022

HFpEF is an under recognised and poorly managed condition. The studies and publications thus far have raised awareness amongst general practitioners (sometimes identifying their lack of knowledge) who are at the forefront of identifying symptoms to begin the cascade of investigations that would lead to diagnosis. For example:

• Some GPs participating in the qualitative study said that they had reviewed information about HFpEF prior to the interview and realised that they did not know much about it.

• GP colleagues at Cambridge have been using the publications and study information in teaching primary care trainees

• Two of the main publications have been published in the main general practice journals: BJGP and BJGP Open so likely to be read by practicing GPs.

• Anonymised quotes from GPs involved in the consensus work:

"I feel I have a significant learning need about HFpEF as I'm not familiar with it."

"As a GP I have often found it hard to talk to patients about this diagnosis."

"I think this is an area GPs really struggle…"

"Very glad you're taking it [HFpEF management] forward."

Identifying the condition early in the disease trajectory has many beneficial outcomes including optimal medication management, lifestyle changes and other beneficial interventions such as supported self-care and rehabilitation. Having robust hospitalisation and mortality data allows us to determine the factors associated with these outcomes, and provides the impetus to implement changes to practice leading to earlier diagnosis and better management. These data, along with information collected from patients (quantitative and qualitative) tell an important story about a vulnerable group. The objective is to develop optimal methods for diagnosis and management that can be tested and implemented into practice.

Heart failure affects 900,000 people in the UK, half of which will be HFpEF. Despite this, the number of publications and interest in HFpEF is significantly less than for Heart failure with reduced ejection fraction (HFrEF). The study's publications focusing on HFpEF will not only raise awareness amongst practitioners but also amongst other stakeholders (charities, commissioners etc.) opening the way to changes in practice which would benefit patients with HFpEF. A research briefing [Research Briefing: diagnosis and management problems for patients with hidden heart failure (adobe.com)] has been shared with healthcare providers, the public and local commissioners through email and social media, and will be presented at an update to regional primary care colleagues in October. An increased spotlight on HFpEF is likely due to the publication of the Emperor-Preserved Trial, which has been the first pharmacological trial (emagliflozin, a SGLT2 inhibitor) to show an improvement in the composite outcome (cardiovascular death and heart failure hospitalisation) in HFpEF. Other trials of SGLT2 inhibitors have shown benefit in improving quality of life in patients with HFpEF. Thus it is a good time to move forward with the next phase of the work, which will be testing interventions in practice.

To date, UK publications on HFpEF have focused on establishing prevalence and hospitalisation rates of HFpEF, and not holistically described a UK population. There is a paper that describes the characteristics of the cohort with HFpEF that was recruited mainly from primary care (BJGPOpen in press). With NHS Digital data linkage we will be able to determine patient outcomes (death and hospitalisation) and factors that increased risk of mortality and morbidity (analysis in process). Analysis shows the cohort recruited are representative of patients seen and managed in primary care, and are invariably different from patients that are enrolled in clinical trials. Being able to link with NHS Digital data means that University of Cambridge can combine the extensive cohort data on patient clinical characteristics and their own reported outcomes of quality of life, psychological status, and symptoms with accurate hospitalisation and mortality data. This will reveal the illness trajectory of this patient group and the factors associated with outcomes. The aim of the overall programme of work is to develop an optimised management programme for this group of patients, and this would be impossible without understanding the patients, their needs, problems and outcomes. The data from NHS Digital are integral to this goal.

The overall aim is to improve patient care for patients with HFpEF. This programme of research will achieve this aim through the multi-faceted work package approach that pursued multiple perspectives to explore and identify problems in current practice and understand system factors that make it difficult for optimal management to be achieved. The final work has been to take what has been learned thus far and conduct a survey and stakeholder workshop to gain ideas about what is needed in practice. Knowledge gained from the baseline work is further enhanced through ongoing analysis of the NHS Digital data on hospitalisations and mortality. As detailed above and in publications from this research, Heart Failure with preserved Ejection Fraction (HFpEF) is a common clinical syndrome which has been referred to as ‘the greatest unmet need in cardiology.’ The condition was first recognised more than 40 years ago, but widespread understanding and pro-active management is lacking. Central to optimising care for any chronic disease is being able to identify people with the condition, but confusion around the diagnosis and management of HFpEF has hindered progress. Making a clinical diagnosis of HFpEF has been described as cumbersome, difficult, based on exclusion and even ‘not clinically relevant’. In the United Kingdom (UK) this challenge is compounded by referral pathways and commissioned services designed to diagnose and treat Heart Failure with reduced Ejection Fraction (HFrEF) only.

The outcome of a system that has evolved around one heart failure phenotype is uncertainty around roles and responsibilities, variable service provision and management disparity for the growing number of people with HFpEF. This is the first UK study to comprehensively describe a substantial UK cohort and has important implications for learning about diagnosis, management and prognosis of patients with HFpEF. The illness trajectory and prognosis of patients with HFpEF is important to communicate, as this provides the impetus for better management. Hospitalisation is the major financial cost associated with care of patients with heart failure, and understanding its frequency and the factors associated with it can lead to specific interventions to support patients and prevent hospitalisation.

As stated previously, the benefits are the provision of robust and accurate information about patients with HFpEF in primary care (demographic and clinical characteristics, problems, patient reported measures from the database), and their illness trajectory including outcomes: changes in mental and physical function, symptoms, quality of life from our database combined with data on hospitalisation and mortality. The University of Cambridge will be able to analyse the factors affecting outcomes and identify areas for intervention. These data will be combined with the other components of Optimise HFpEF (qualitative data) to provide a rich picture of patients with HFpEF.

The framework for dissemination includes:

• peer-reviewed journals (building on earlier outputs) with 8 papers already published

• presentations at national and international primary care, nursing and cardiovascular conferences

• reports and patient materials (input from the Patient Advisory Group to support this). One of the investigators is on the steering committee of the patient HF charity Pumping Marvellous, so their support will be enlisted in dissemination of patient materials, links and downloadable information on the study website. The investigators and collaborators include clinicians from primary care and cardiology services so they will assist in reaching the widest possible audiences.

The information will also be valuable in future work around building consensus regarding diagnosis and management of this patient group, which has already begun.

The outputs are both realistic and comprehensive. The study team anticipates growing attention and interest in HFpEF amongst charities who provide patient information, general practitioners who initiate diagnosis and commissioners who establish funding of services.

900,000 people in the UK have heart failure, we do not know how many of these people have HFpEF however epidemiological studies estimate this to be 50%. HF consumes around 2% of the TOTAL NHS budget (principally due to hospitalisations) and this is set to rise with ageing population who are living longer with a higher burden of co-morbidities. Initial research from Ireland has estimated the cost of the two types of heart failure (HFrEF and HFpEF) and found that costs for HFpEF are greater than HFrEF. Epidemiological analyses from the US have found that hospitalisation for HFrEF is decreasing, while that for HFpEF is increasing.

This research aims to benefit patients with HFpEF, and health services in the long-term if improvement to diagnosis and management lead to a decrease in avoidable hospitalisations. Greater awareness of HFpEF and understanding of patient characteristics and their outcomes will lead to earlier diagnosis and improved management. At the least it will provide the impetus to change services to decrease the disparity in access currently experienced by patients with HFpEF. These results will also lead to further research to develop and test methods for optimal management that can improve outcomes. The controller will benefit from recognition associated with peer review publications.

Benefit will be measured by the impact factor and altimetric of publications, citations, media exposure, speaker invitations. The findings of this research will also lead to further research to develop and test interventions to improve patient outcomes and well-being. The programme of research begun here will also provide opportunities to develop research capacity by stimulating doctoral (one doctoral project has already received funding to conduct a study) and post-doctoral research.

The publication plan has been outlined in the sections above. To date eight papers have been published from the Optimise HFpEF study, with three of these from the longitudinal cohort (baseline). Initial outputs from the follow-up data (including NHS Digital data linkage) are expected in early 2022.

Benefits thus far (despite disruption and delays due to the COVID-19 pandemic):

• A thorough understanding of diagnosis and management of HFpEF across healthcare sectors, including challenges and problems, experiences of patients and perspectives of primary care and specialist healthcare providers. The understanding has been gained through:

More than 140 interviews with patients, carers and healthcare providers

Survey of 66 providers regarding needed changes to practice

Stakeholder workshop

• Robust phenotyping of patients with HFpEF and their outcomes in the community

Recruitment of 152 patients from 30 primary care practices and 2 specialist centres

Sixty-one percent confirmed as HFpEF (n = 93) and followed-up over one year

In process analysis of linked data from NHS Digital to determine morbidity and mortality

• Work begun to develop interventions to improve diagnosis and management of HFpEF using a systems approach

• Obtained funding from the Evelyn Trust to conduct a study leading to a PhD: A blended lifestyle intervention to preserve lean mass, muscle strength, exercise tolerance and quality of life in multi-morbid older people. The intervention was developed from findings of the study and discussion with Patient Advisory Group. £23,069

• Developed collaborations with other HFpEF researchers: REACH-HFpEF trial and submission of grant application with US colleagues.

Outputs:

Research Briefing: diagnosis and management problems for patients with hidden heart failure (adobe.com)

Lin H, Hartley P, Forsyth F, Pilling M, Hobbs FDR, Taylor CJ, Schiff R, Deaton C on behalf of the Optimise HFpEF investigators. Clinical and demographic correlates of accelerometer-measured physical activity in participants enrolled in the OPTIMISE HFpEF study. European Journal of Cardiovascular Nursing. 2021 10.1093/eurjcn/zvab028

Forsyth F, Brimicombe J, Cheriyan J, Edward D, Hobbs FDR, Jalaludeen N, Mant J, Pilling M, Schiff R, Taylor CJ, Zaman MJ, Deaton C. Diagnosis of Patients with Heart Failure with Preserved Ejection Fraction in Primary Care: Cohort Study. ESC Heart Failure. 2021. http://doi.org/10.1002/ehf2.13612

Forsyth F, Brimicombe J, Cheriyan J, Edward D, Hobbs FDR, Jalaludeen N, Mant J, Pilling M, Schiff R, Taylor CJ, Zaman MJ, Deaton C. Characteristics of Patients with Heart Failure with Preserved Ejection Fraction in Primary Care: Cross-sectional Analysis. BJGP Open. 2021 (in press)

Forsyth F, Sowden E, Hossain M, Tuffnell R, Blakeman T, Chew-Graham C, Deaton C. Clinicians’ and patients’ experiences of managing heart failure during the COVID-19 pandemic. BJGP Open. 2021: DOI: https://doi.org/10.3399/BJGPO.2021.0115

Hossain M, Chew-Graham C, Sowden E, Blakeman T, Deaton C. Challenges in the management of people with heart failure with preserved ejection fraction (HFpEF) in primary care: a qualitative study of general practitioner perspectives. Chronic Illness. 2021 DOI: 10.1177/1742395320983871

Sowden E, Hossain M, Chew-Graham C, Blakeman T, Tierney S, Wellwood I, Rosa F, Deaton C. Understanding the management of Heart Failure with Preserved Ejection Fraction: a qualitative multi-perspective study. Br J Gen Pract. 2020 Nov 2:bjgp20X713477. doi: 10.3399/bjgp20X713477. PMID: 33139334

Kalogirou F, Forsyth F, Kyriakou M, Mantle R, Deaton C. Heart Failure Disease Management: A Systematic Review of Effectiveness in Heart Failure with Preserved Ejection Fraction. ESC Heart Failure. 2020; doi: 10.1002/ehf2.12559

Forsyth F, Mant J, Taylor C, Hobbs R, Chew-Graham C, Blakeman T, Sowden E, Long A, Hossain M, Edwards D, Deaton C. Optimising Management of Patients with Heart Failure with Preserved Ejection Fraction in Primary Care (OPTIMISE-HFpEF): Rationale and Protocol for a Multi-Method Study. British Journal of General Practice Open. November 2019 https://doi.org/10.3399/bjgpopen19X101675

Presentations and Abstracts:

2021 Florence Nightingale Lecture on Nurse-led Research (virtual, ESC Congress, August 2021). A Multidisciplinary Approach to Diagnosing and Managing Heart Failure with Preserved Ejection Fraction.

Keeping the plates spinning: a qualitative study of the multifaceted role of caregiving in HFpEF (virtual presentation, EuroHeartCare 2021)

Management of patients with cardiovascular disease during and after the COVID-19 pandemic: what has and should change? (virtual presentation, Society for Academic Primary Care, 2021)

Characteristics and health status of patients with and without confirmed HFpEF (virtual poster, ESC Congress 2020)

Daily physical activity levels in patients with heart failure with preserved ejection fraction: clinical correlates and subjective perception of activity (virtual poster, ESC Preventive Cardiology Congress. 2020)

Multi-morbidity and self-care in older patients with heart failure. (Presentation, EuroHeartCare. 2019)

Expected Benefits

The intention is to use the knowledge gained from Optimise HFpEF and the NHS Data Linkage to develop a programme of research testing interventions and changes to the patient pathway that will improve diagnosis and management. The importance of the data linkage is that it broadens our understanding of the trajectory of patients with HFpEF in the community. How often are patients hospitalised and readmitted? What is the average length of stay? What is the mortality rate? What factors make patients more likely to be hospitalised? All of this information is useful when developing the rationale for further research and communicating to healthcare providers, commissioners and integrated care systems. The University of Cambridge have begun collaboration with a healthcare engineering expert and held a recent stakeholder meeting to consider a systems approach to addressing issues in HFpEF diagnosis and management. There were 20 participants from primary care, HF specialist services and the public, who had not previously been involved in the research. They had all received a research briefing and discussed both problems and possible solutions.

Along with ongoing data analysis, we will begin to develop a Programme Grant for Applied Research to be submitted in Spring 2022. The intent will be to test interventions and changes to practice to determine if they will make a difference in time to diagnosis of HFpEF and improving management.

A PhD student is already working on a blended diet and activity intervention that will be tested to address identified problems of sarcopenia, nutritional deficiencies, poor activity tolerance and sedentary behaviour.