OPtimising Treatment for Mild Systolic hypertension in the Elderly: a randomised controlled trial (OPTiMISE)

The University of Oxford requires access to data for a long-term follow-up study of participants enrolled into the OPtimising Treatment for MIld Systolic hypertension in the Elderly (OPTiMISE) trial. Participants were enrolled into the OPTiMISE trial between April 2017 and September 2018 and underwent active follow-up until January 2019. The trial remains ongoing in passive follow-up (i.e., active participation from participants is no longer required).

This study aims to examine the effect of antihypertensive deprescribing on adverse events including cardiovascular disease, death and falls in patients enrolled into the OPTiMISE trial. The findings of this work is hoped to establish the longer-term impact of antihypertensive deprescribing and inform clinical guidelines on the treatment of high blood pressure in older age adults.

High blood pressure is the leading modifiable risk factor for cardiovascular disease and the most common co-morbid condition in older people with multi-morbidity. Antihypertensive treatment has been shown to be effective at preventing stroke and cardiovascular disease in older high-risk patients and approximately half of individuals aged 80 years or older are prescribed therapy.

Clinical guidelines recommend using clinical judgement when prescribing in frail older patients with multi-morbidity, emphasising a personalised approach to care which might include attempts to improve quality of life through deprescribing. They state that drugs which are used to prevent future events (e.g. antihypertensives) should be considered for withdrawal, taking into account an individual’s personal goals and the likelihood of benefit and harm from continued treatment. However, these guidelines are largely based on expert opinion and currently there are no trials examining the long-term effects of antihypertensive deprescribing.

The OPtimising Treatment for MIld Systolic hypertension in the Elderly (OPTiMISE) trial examined a structured approach to antihypertensive medication reduction in 569 older patients (aged 80+ years) with multi-morbidity and controlled systolic hypertension, prescribed two or more antihypertensives (Sheppard JP, Burt J, Lown M, et al. Effect of Antihypertensive Medication Reduction vs Usual Care on Short-term Blood Pressure Control in Patients With Hypertension Aged 80 Years and Older: The OPTIMISE Randomized Clinical Trial. Jama 2020;323(20):2039-51. doi: 10.1001/jama.2020.4871; https://jamanetwork.com/journals/jama/article-abstract/2766421). This study showed that deprescribing can be successfully achieved in two thirds of patients with no short-term impact on blood pressure control, quality of life or serious adverse events. However, to better understand whether deprescribing is safe and should be attempted, longer-term follow-up of patients enrolled into the trial is required.

This proposal aims to follow-up participants from this trial, 3 years after randomisation. Primary care data will be collected via the Oxford Royal College of General Practitioners (RCGP) Clinical Informatics Digital Hub (ORCHID), and secondary care data will be collected via Hospital Episode Statistics, Emergency Care Data and Civil Registration death data from NHS Digital.

The University of Oxford is the data controller for ORCHID and make the ORCHID database available as a resource for conducting research. ORCHID contains pseudonymised data including a data subject’s age, sex, ethnicity, symptoms, procedures, prescriptions, coded information about test results, hospital referrals, etc. The University of Oxford create a separate dataset for each approved project so that the researchers do not access data from other projects.

The University of Oxford and ORCHID database manages processes which pseudonymise data at two levels to prevent the approved researchers who will be granted access to data extracts from re-identifying individuals whose data resides in the ORCHID database. For example, the ORCHID database does not contain the addresses, dates of birth for subjects. It only contains the LSOA in case of addresses and only the rounded off age.

The pseudonymisation process used for linking ORCHID data to other datasets is designed so that ORCHID database administrators cannot identify individuals (e.g. by accessing both the key and the data to reverse the hashing process thereby getting to the original data). The University has established a mechanism to prevent this linkage by granting access to the encryption salt to two individuals who are not a part of the Clinical Informatics and Health Outcomes Research Group.

When a data linkage is required and approved by the data provider, the designated encryption salt holder arranges to provide the linked data provider with the encryption salt; whereupon, the data provider will hash personal identifiers (for the data set requested) using a hashing algorithm and the encryption salt.

The hypothesis for this study is that medication reduction maybe associated with an increased risk of cardiovascular disease and adverse events (including hospitalisation, dementia and other conditions affecting quality of life). Using the above-mentioned data from NHS Digital, the study will answer the following questions:

1. Is medication reduction associated with an increased risk of serious adverse events (death or hospitalisation for any cause)?

2. What are the long-term benefits or harms of medication reduction in terms of adverse events, risk of death, cardiovascular disease, hospitalisations?

3. Are differences in outcomes modified by baseline characteristics of participants including frailty and multi-morbidity?

Study population: The study population consists of the remaining 558 participants enrolled into the OPTiMISE trial who have given consent for long-term follow-up. At baseline in the trial, participants had a mean age of 85±3 years, 49% were female, 60% were considered frail and 98% had multi-morbidity. Individuals had a mean blood pressure at baseline of 130/69 mmHg and took a median two antihypertensives to keep it under control. The population will be followed-up using primary care data from ORCHID and secondary care data from NHS digital. ORCHID currently includes 33/69 of the practices (47.8%) which participated in OPTiMISE, and 298/558 of the participant (53.4%). This number is likely to expand in the coming months and all OPTiMISE trial practices will be invited to join ORCHID as part of this study. Participants were randomised (1:1 ratio) to a strategy of antihypertensive medication reduction (removal of 1 drug [intervention], n = 282) or usual care (control, n = 287), in which no medication changes were mandated.

Measures of exposure: The exposure of interest in this study is medication reduction. The OPTiMISE trial mandated that all participants allocated to the medication reduction arm of the trial have one antihypertensive medication removed from their treatment regime. The choice of medication to remove was left at the discretion of the consulting general practitioner (GP), who were given advice on the most appropriate medications to withdraw.

Study outcomes: The primary outcome of this analysis will be the difference in rates of serious adverse events between intervention and control groups at 3-year follow-up. Serious adverse events will be defined as death or any hospitalisation (inpatient, outpatient admission or emergency department attendance). Secondary outcomes will explore differences in the rates of ischemic heart disease (including myocardial infarction), stroke, death (from any causes), falls, primary care consultation rates (i.e. workload) and non-serious adverse events (defined as those commonly reported during the trial such as ankle oedema, dizziness, headache, back pain, etc.).

The University of Oxford require latest available data covering the following information about consenting participants in the OPTiMISE trial from NHS Digital:

- A one off feed of the Emergency Care Data Set

- A one of feed of the Hospital Episode Statistics Admitted Patient Care

- A one off feed of the Hospital Episode Statistics Accident and Emergency

- A one off feed of civil registration data (death certificate information)

Only data relating to the 558 participants who gave consent for the trial (and long term follow-up) will be requested for the study. The data disseminated from NHS Digital to the University of Oxford will be pseudonymised record level information from between 2017 and 2022, i.e., the period between randomisation to the trial and the latest available data at time of data dissemination. The data will only relate to participants recruited in English healthcare settings during the trial. The primary outcome is serious adverse events which include hospitalisation and death. Therefore, only data relating to hospital visits and deaths, including diagnoses (for assessment of secondary outcomes) will be requested. The University of Oxford believe there are no alternative, less intrusive ways of achieving the purpose.

As far as possible practices are encouraged to join the network. If, however, the practices are not willing to join then they will be manually followed up by the trial team, who will go into the practices and undertake a manual notes review, without having the ORCHID database to be directly involved.

The ORCHID team will only ever link patients using pseudonymised NHS numbers. The trial team is expected to provide the list of pseudonymised NHS numbers to the ORCHID database for the purpose of linkage. Once the linkage is done within ORCHID and NHS Digital, it is then handed over to the trial team to further process the data.

For patients from those GP practices who are not members of the ORCHID database, the flows of NHS Digital data in respect of those patients is appropriate. For patients from GP practices who are not members of the ORCHID network, the trial team will provide pseudonymised NHS numbers to the ORCHID database administrator who will then upload the entire cohort of the trial to NHS Digital to obtain related secondary care data. Therefore, for these particular patients, the ORCHID database will have only the secondary care data without any primary care data linkage from ORCHID.

The University of Oxford are the sole data controller who also process the data.

The University of Cambridge led on two qualitative studies for the main trial (scoping work to understand the perspectives of patients and GPs and to inform recruitment approaches, followed by an iterative examination of recruitment within the trial). Co-investigators from the Universities of Southampton and Bristol were involved in the trial and were responsible for recruiting participants from these regions of England. They are not involved in the long-term follow-up of participants and will not determine the purpose for processing as described in this Agreement or the manner in which the data will be processed. Nor will they have access to data provided by NHS Digital.

Legal basis: The University of Oxford will process data under 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"

Additionally, personal data will be processed under 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". These legal bases were applied because the University of Oxford are undertaking a project which is in the public interest and which is only possible through linkage of study data to hospital and death data. The study is of public interest because it is expected to be the first to determine the association between antihypertensive medication reduction and clinical outcomes such as hospitalisation and death due to cardiovascular disease. This is in keeping with the studies’ original aims to examine the safety and effectiveness of antihypertensive medication reduction. This further important information could inform the care of millions of older people taking antihypertensive therapy across the country. To determine this association accurately, the University of Oxford need to access hospital and civil registration death data specifically for participants recruited to the original medication reduction trial. Accessing these data through NHS Digital is the only way to achieve this linkage.

Patient and public involvement: In preparing this Agreement, the University of Oxford held two focus groups (at the Age UK day centre in Bakewell with 20 frail older people and at their academic department in Oxford with 7 older people aged 70+ years) and further individual discussions with older people and their carers. The project includes a patient co-applicant who has been involved in the study right from conception and is a member of the trial management group. The trial steering committee included two further patient representatives. They have emphasised the importance of this follow-up study and the obligation of the research team to undertake it having created an expectation that it will be completed in the information provided to patients at the beginning of the study.

The consent forms and patient information sheets were developed in conjunction with the University of Oxford's patient representatives and further input was sought from focus groups to ensure they were clear and easy to understand.

This study is funded by the British Heart Foundation who will provide funding for resources required to complete this study, but will not have access to any of the data collected and do not determine the purpose or the means of the processing.

1. Participants who have consented to the OPTiMISE trial will have their data initially linked with the primary care data which is already held as part of ORCHID.

2. The University of Oxford will provide NHS Digital with a list of NHS numbers and date of births along with a unique Study ID for the OPTiMISE cohort.

3. A specific member of the University of Oxford IT team within the IT department will share the identifiers with NHS Digital, not the research team.

4. NHS Digital will send back to University of Oxford the pseudonymised, record level cohort data with Hospital Episode Statistics (HES), Admitted Patient Care (APC) and Accident & Emergency (A&E) data, Emergency Care Dataset (ECDS) data and mortality data included. Files will be sent securely back to University of Oxford via the Secure Electronic File Transfer System (SEFT).

5. University of Oxford will download and store the data within their secure ORCHID trusted research environment.

6. University of Oxford will process the pseudonymised data to meet the aims and objectives of the study (as detailed in ‘Objectives for Processing’). This will include matching the pseudonymised data from NHS Digital with the pseudonymised study data using the unique study ID. There will be no subsequent flows of data from the University of Oxford which has not already been suppressed in accordance with the small number suppression rules in the HES Analysis Guide.

Once data have been provided by NHS Digital, the University of Oxford will have no need to re-identify participants for the purpose of data processing. All patient identifiers are stored (securely and) separately from the main study database. Patient identifiers will be permanently destroyed at the earliest opportunity, in line with ethical and GDPR requirements and University policy.

The data will be controlled and processed by a group of substantive staff who are all based at the University of Oxford and under an employment contract. All staff are mandated to complete information governance training. The group is made up of analysts, academic fellows, Structure Query Language (SQL) developers, Royal College of General Practitioners Research and Surveillance Centre (RCGP RSC) practice liaison officers, a project manager and a head of department. All staff will access NHS Digital data from secure workstations or secure laptops with encrypted drives within the group’s secure network. No data will be accessed outside of the location where the data is stored.

Analysis will be done by the clinical trials unit within the Nuffield department of Primary Care Health Sciences at the University of Oxford.

All findings from the proposed research are intended to be published in peer-reviewed journals under an open-access license, with target journals including the British Medical Journal (BMJ), Journal of the American Medical Association (JAMA) Internal Medicine and Hypertension. Only aggregated data with small numbers suppressed in line with the HES Analysis Guide will be presented in published outputs. It is anticipated that the study findings will support better patient-centred management plans for the prevention of cardiovascular disease in older individuals and therefore will be available for the next iterations of the NICE hypertension and multi-morbidity guidelines.

The University of Oxford intend to present results at national scientific meetings (e.g. of the British and Irish Hypertension Society, Society for Academic Primary care, British Geriatrics Society) and international conferences (e.g., for the European Society of Cardiology, North American Primary Care Research Group). The University of Oxford will publish a summary of our findings on the study website (https://www.phctrials.ox.ac.uk/studies/optimise) and submit a final study report to the MHRA and ethics committee who provided approval for the study. Where appropriate, we will press release the results of the research in conjunction with their publication in scientific journals, and fully engage with any arising media inquiries that result. Social media (e.g. Twitter) will be used to draw attention to the work and stimulate debate, particularly when it is presented at conferences or published in the lay and scientific media.

Multiple channels of communication will be used to disseminate study findings including written feedback to study participants, plain English summaries, newsletters and community engagement events. We will work closely with our PPI representatives to ensure results are presented and disseminated in a patient friendly manner. An article discussing the issues raised by the research will be written for ‘The Conversation’, an online newspaper written by academics which is free and easy to read for the general public.

The University of Oxford will aim to present the findings of their analyses using these data at conferences in late 2022 and throughout 2023 and publish the final results in publications and online in early 2023.

This study aims to establish the longer-term impact of reducing blood pressure medications in older patients with mild hypertension, polypharmacy and multi-morbidity. This will be the largest study to examine the impact of deprescribing on adverse events, cardiovascular risk and falls and the first to provide important data on whether such a practice is safe and effective in the longer term. Such information could be incorporated into clinical guidelines soon after the publication of the study results and is likely to affect a large number of older people. In the UK, approximately 1.9 million adults are aged 80+ years and prescribed antihypertensive medications. If deprescribing is shown to be beneficial, it could lead to a reduction in inappropriate polypharmacy amongst older patients and improve their quality of life. Indeed, around 250,000 non-elective NHS hospital admissions (costing £530 million) are thought to be related to potentially avoidable adverse drug reactions, every year. If deprescribing is found to be beneficial, the number of hospital admissions due to potentially avoidable adverse drug reactions could be reduced. However, if the study shows deprescribing to be potentially harmful due to an increase adverse events, this work should prevent inappropriate deprescribing of antihypertensive medications in older patients with multi-morbidity (currently 30% of individuals over the age of 75 have their antihypertensive medications stopped for one reason or another). Therefore, the impact of this work can be measured in the number of non-elective NHS hospital admissions and the number of patients having antihypertensive medications deprescribed each year.

Furthermore, because there are limited data at present on optimal strategies for deprescribing in older patients, the results of this study will likely inform future trials of medication reduction in other clinical conditions.

It is hoped that the findings of this trial will support better patient-centred management plans for the prevention of cardiovascular disease in older individuals and will be made available for the next iterations of the NICE hypertension and multi-morbidity guidelines.

This is a new data request, so no benefits have yet been yielded.