Plain English Summary
Background and study aims
Atrial fibrillation (AF) is a heart condition that causes an irregular heartbeat. It affects up to 1 in 10 people over the age of 65. AF greatly increases the risk of stroke, but treatment with blood-thinning (anticoagulant) medication can stop this happening. About 10% of strokes happen in people unaware they have AF. Detecting AF can be difficult because it often comes and goes, and may not cause symptoms. Many clinicians think the NHS should promote AF screening. The UK National Screening Committee has highlighted a lack of evidence that detecting AF in people by screening would benefit them. Researchers are therefore undertaking a large 8-year programme of work to find out if screening for AF in people aged 70 and over does prevent stroke and other problems like heart attacks, does not cause significant harm, and represents good value-for-money for the NHS. This study will address these questions.
Who can participate?
People aged 70 or over
What does the study involve?
Participating practices will be randomly allocated to the intervention (AF screening) group or the control group. Consented participants in the intervention practices will be invited to attend an AF screening visit (see below for more details). The control group do not receive any interventions. Data collection will be electronic through GP medical notes and NHS Digital and health registry linkage, with an average of 5-year follow-up. Consented participants in intervention practices will be invited to take part in screening for atrial fibrillation. The screening will be carried out in two stages: an initial screen at the practice, followed by 3 weeks of intermittent screening at home. A member of staff will assign a single-lead ECG recorder (provided by Zenicor, a Swedish commercial partner in the trial) to the patient and instruct them how to use it. Each ECG records for 30 seconds. The participant will be sent home with the ECG recorder and instructions for use. The participant will be asked to take ECG recordings and to transmit each ECG recording to a remote database (via the mobile phone network) by pressing a button on the recorder. The researchers will request ECG traces are recorded four times per day (plus when symptomatic) for 3 weeks. Participants will be asked to return the ECG recorder to the practice at the end of the screening period. ECGs will be reviewed once the ECG recorder is returned. ECGs tagged as ‘possible AF’ or ‘other deviations’ will be reviewed by a trained individual for cases of AF or other important rhythm disturbances. The researchers will send diagnostic results to the practice. The practice will inform all participants of their screening result following their local process for informing patients of test results. Participants are informed that the results of the screening will be made available to their practice by 12 weeks from the date they return the ECG recorder to the practice. This timescale is in line with current NHS screening programmes. Participants with confirmed AF will be invited to attend an appointment with the GP (or other healthcare professional in accordance with local policies) to discuss the diagnosis and the appropriate management including initiation of anticoagulation therapy. Other rhythm abnormalities, in addition to AF, may also be detected. Such ECGs will be reviewed by a trained individual and reported to the GP via the same process as diagnoses of AF outlined above if they are clinically relevant. This completes the participant involvement in the screening process. Baseline and follow-up (approx. annually) data from the GP medical record will be collected in accordance with the consent provided. A random sample of patients will be assessed by postal questionnaire on up to three occasions. Participants from various stages of the study will be invited by the trial team to take part in interviews, either by telephone or face-to-face. Interviews will be audio recorded.
What are the possible benefits and risks of participating?
Taking part in the study may bring participants a sense of altruism in that it will be the world's largest randomised trial for AF screening. Knowing that the programme aims to prevent strokes and heart attacks in people just like them in the future may have a positive mental effect on the participant. Patients diagnosed with AF will be offered treatment which will greatly reduce their risk of having a stroke or heart attack, and possibly dementia. It is possible that a non-AF clinically significant cardiac rhythm abnormality may be diagnosed, in which case the patient will be referred to their GP for appropriate care - this may include further tests and/or treatment if necessary.
Where is the study run from?
1. NIHR CRN: Eastern
2. NIHR CRN: North East and North Cumbria
3. NIHR CRN: North West Coast
4. NIHR CRN: Yorkshire and Humber
5. NIHR CRN: Greater Manchester
6. NIHR CRN: East Midlands
7. NIHR CRN: West Midlands
8. NIHR CRN: West of England
9. NIHR CRN: Thames Valley and South Midlands
10. NIHR CRN: Kent, Surrey and Sussex
11. NIHR CRN: Wessex
12. NIHR CRN: South West Peninsula
13. NIHR CRN: North Thames
14. NIHR CRN: South London
15. NIHR CRN: North West London
When is the study starting and how long is it expected to run for?
August 2017 to March 2026
Who is funding the study?
NIHR Programme Grant for Applied Research (UK)
Who is the main contact?
Mr Andrew Dymond
SAFER@medschl.cam.ac.uk
Trial website
Contact information
Type
Scientific
Primary contact
Dr Andrew Dymond
ORCID ID
Contact details
SAFER Study Coordinator
Strangeways Research Laboratory
Worts Causeway
Cambridge
CB1 8RN
United Kingdom
+44 (0)1223 761457
safer@medschl.cam.ac.uk
Additional identifiers
EudraCT number
Nil known
ClinicalTrials.gov number
Nil known
Protocol/serial number
CPMS 43549, IRAS 272184
Study information
Scientific title
The SAFER Trial: Screening for Atrial Fibrillation with ECG to Reduce stroke - a randomised controlled trial
Acronym
SAFER
Study hypothesis
This study aims to find out if screening for atrial fibrillation (AF) in people aged 70 and over does prevent stroke and other problems like heart attacks and represents good value-for-money for the NHS.
Ethics approval
Approved 23/10/2019, London Central NRES (3rd Floor, Barlow House, 4 Minshull Street, Manchester, M1 3DZ, UK; +44 (0)207 104 8012; NRESCommittee.London-Central@nhs.net), REC ref: 19/LO/1597
Study design
Randomised; Both; Design type: Screening, Other, Qualitative
Primary study design
Interventional
Secondary study design
Randomised controlled trial
Trial setting
Hospitals
Trial type
Screening
Patient information sheet
Not available in web format, please use the contact details to request a patient information sheet
Condition
Atrial fibrillation
Intervention
This trial has been carefully designed in consultation with the UK National Screening Committee to mimic as closely as possible a 'real life' national screening programme. There is no point in testing an artificial 'research' version of a screening programme, as if this was found to be feasible and rolled out by the NHS it would most likely fail. The researchers will recruit 360 practices, who will each invite around 1,000 eligible patients to take part in the trial. Participants will provide written consent by post. Following this, practices will be randomised to intervention (AF screening) or control in the ratio 1:2. Consented participants in the intervention practices will be invited to attend an AF screening visit (see below for more details). The control group do not receive any interventions. Data collection for the trial will be electronic through GP medical notes and NHS Digital and health registry linkage, with an average of 5-year follow-up.
Some participants will also be invited to complete questionnaires and to take part in qualitative work including interviews and observations. There is an extensive health economic modelling work package in the trial, as a critical factor in determining the success of the screening programme will be value for money for the NHS. There is an internal pilot phase of 36 practices (12 screening, 24 control). There are 3 types of participants in this trial; patients (the largest group), practice staff, and additional healthcare professionals/external stakeholders. The latter two groups will be invited to participate in qualitative components only. Each of these participant types will be considered in turn:
1. Patients
The researchers will recruit 360 GP practices in the UK. Practices will identify potentially eligible patients >= 65 years with an electronic record search and invite about 1,000 to take part in the trial. The researchers may also use this opportunity to over-sample specific substrata e.g. the very elderly, if they observe at early stages that they are under-represented. The practice will send identified patients an invitation pack consisting of a participant information sheet (PIS), covering letter, consent form, reply slip and Freepost envelope. The PIS and consent form will focus on participation in the wider trial and the request to access and follow up the patient’s electronic health records. It will explain that this is for a trial of stroke prevention, and that participants may be contacted to participate in screening, and can decide at that point whether to accept or decline the screening invitation. The reply slip will enable patients who do not wish to consent to respond anonymously indicating their reasons for declining the trial invitation. It will also provide an option for patients who do not wish to participate to indicate that they would be happy for a researcher to contact them to discuss further their reasons for deciding to not participate. This is a key part in understanding whether there is anything about the concept of screening that is putting patients off consenting. If the patient wishes to take part in the trial they will be invited to complete the consent form and return it to the trial team in the Freepost envelope provided or to complete an online consent form (this process will be managed by the trial team).
Following participant consent, the practice will be randomised to intervention (AF screening) or control at a ratio of 1:2. Consented participants in intervention practices will be invited to take part in screening for atrial fibrillation. They will be sent an additional PIS about the screening process and its potential benefits and harms – consistent with a national screening programme leaflet. This invitation needs to be separate from the main research trial invitation as the researchers need to be able to specifically test a screening invitation leaflet of the type (in terms of content, language and tone) that would be used in an NHS national screening programme. Conflating the two information leaflets would make this impossible.
The researchers may also use SMS messages from the practice as part of the recruitment process. This method of communication is increasingly commonly used by practices for communication about routine care in patients in this age group with a great deal of success. It is likely that it would be utilised in national roll-out of an AF screening programme. Some practices may wish to also invite patients to a trial-specific meeting, as part of the invitation to the screening process. Here AF screening is presented and patients are able to ask questions. This approach is used in research practices in the local region to good effect.
During the trial in addition to screening invitation letters, the researchers will also encourage practices to opportunistically approach consented patients, for example at routine appointments, health checks, flu vaccinations, etc. Again, reminder letters and/or emails/SMS will be used if necessary. Procedures at individual practices will be adapted to fit existing practice systems as much as possible, to minimise burden and replicate the delivery of a national screening programme. As a critical factor to the success of the trial, response rates will be closely monitored throughout and approaches and processes changed as required to optimise uptake.
Some participants will also be invited at various points in the screening process to take part in additional, optional qualitative interviews. The initial trial invitation PIS will introduce these elements stating that additional information will be provided at the point of invitation to these elements, and the participant is free to decide at that point whether to take part or not. A decision not to take part will not affect their participation in the main trial. The researchers will carefully sample participants for an invitation to take part in these elements to ensure that they are not overburdened by invitations to participate in multiple qualitative aspects. Please see the qualitative section towards the end of this question for more details of these components.
The screening will be carried out in two stages: an index ECG recorded at the practice, followed by 3 weeks of intermittent screening at home.
The AF screening visit will be with an appropriately trained member of staff, either from the practice or a research nurse from the Clinical Research Network. Participants will be asked to confirm verbal consent to participate in the screening process. The member of staff will assign a single-lead ECG recorder (provided by Zenicor, a Swedish commercial partner in the trial) to the patient and instruct them how to use it. Each ECG records for 30 seconds. Initial screen: an index ECG recording will be performed by the member of staff who will check it using the online system for quality and heart rate. The individual viewing the index ECG trace will not have access to the analysis and resulting classification provided by the system algorithm, and will not be instructed to review the trace in a diagnostic capacity. Further participant instruction and repeat index ECGs will be made if the participant is not able to record an ECG trace of sufficient quality, or if other issues are identified. In a small percentage (about 1%) of individuals it will not be possible to obtain a single lead ECG trace of sufficient quality, purely for physiological reasons. In such a situation (and if the participant is unable to record a good quality trace at the visit) the member of staff will be instructed to take the pulse manually. If the pulse is regular the participant will not undergo further ECG screening as part of the trial, but will remain in the trial for the purposes of data collection. If the pulse is irregular, the participant will leave the trial AF screening process and be managed as per local clinical practice, most likely with a 12 lead ECG. They will not undergo further ECG screening as part of the trial but will remain in the trial for the purposes of data collection. If the heart rate of the index ECG is outside a pre-defined range, the member of staff will be instructed to follow the routine clinical pathway for such patients. The heart rate check will facilitate the early detection of conditions such as fast AF, atrial flutter, supraventricular tachycardia or complete heart block of a degree which may possibly require clinical review in a more timely manner than the routine screening pathway. The patient will remain on the trial AF screening pathway at the discretion of the GP.
Some practices may wish to train participants to use the ECG recorders at group sessions at the practice, both as well as or instead of individual training by the member of staff at the AF screening visit. This approach is reported to work well in other studies. The researchers will monitor patients who attend such groups to determine if it confers additional compliance, and if so, may encourage it in subsequent practices.
Intermittent screening: The participant will be sent home with the ECG recorder and instructions for use. The participant will be asked to take ECG recordings and to transmit each ECG recording to a remote database (via the mobile phone network) by pressing a button on the recorder. In the STROKESTOP trial, sufficient adherence (at least 15 ECG traces) was achieved by 99% of participants (aged 75). Observations from the SAFER feasibility study suggest excellent adherence to 4 ECGs per day across age groups.
The researchers will request ECG traces are recorded 4 times per day (plus when symptomatic) for 3 weeks. They will continue to monitor compliance, problems, ECG quality, incidence of AF etc. As a result they may revise the frequency and/or duration to achieve the optimal combination. In order to maximise adherence to ECG recording they may intervene by contacting the participant if fewer than the expected number of ECG traces are received (e.g. less than 50%), if the Zenicor system alerts that no ECGs have been received from a device for a certain number of days, or if a significant proportion of poor quality traces are recorded.
When participants start their screening period the researchers will send an email (submitted for review) reiterating good ECG recording technique including the link to the patient ECG recording instruction video on the trial website (previously seen by the committee as part of the feasibility study). This is to try and ensure good quality traces are recorded in order to facilitate review and diagnosis by the cardiologists.
Participants will be asked to return the ECG recorder to the practice at the end of the screening period. Practices will use various means of return, for example in person or by post, or offering specific return appointments, adapting the approach as the trial progresses to ensure optimal ease of return for the participant whilst minimising recorder loss (as they are expensive to replace).
ECG reading system:
The ECG traces are not displayed on the recorder, but are transmitted via mobile signal to a central secure Zenicor database from where they are viewed via a web-based platform. At the end of the screening period when the ECG recorder is returned to the practice, the patient will be flagged on the system as complete and ready for review. The system algorithm classifies and tags each ECG trace as ‘no tag’, ‘possible AF’, ‘other deviations’ or ‘poor quality’. ECGs tagged as ‘possible AF’ or ‘other deviations’ will be reviewed by a first filter (an experienced cardiac technician or similar) who will forward suspected cases of AF or other abnormalities to a trial cardiologist to confirm if AF or any other important rhythm disturbance is present. The researchers will not require a 12-lead ECG to be undertaken, as single lead ECGs read by cardiologists have high accuracy. If at the end of screening the cardiologist is unable to determine a diagnosis a further assessment may be offered using a different modality (e.g. ECG patch). The details of these patches are still to be confirmed and will be submitted in a future amendment. Review and subsequent classification of ECGs, including diagnosis and additional information for the GP occurs within the web-based platform.
Screening results:
The trial team will log into the Zenicor system to retrieve the diagnostic information of all patients and send it to the practice. The practice will record the information on the patient medical notes. In addition, a full report including cardiologist diagnosis, comments and a copy of relevant ECG traces will be sent to the practice for each patient with a positive (AF or other) diagnosis. The GP will also be provided with access to the Zenicor online system to view and download the ECG traces and reports for all their patients taking part, should they wish to do so. An expedited results pathway will be available for any screening results the cardiologist identifies as clinically urgent. All participants with cardiologist confirmed AF (or other non-AF clinically significant finding) will be notified of their screening result. The researchers will encourage the practice to inform all participants of their screening result following their local process for informing patients of test results. Participants are informed that the results of the screening will be made available to their practice within 8-12 weeks from the date they return the ECG recorder to the practice. This timescale is in line with current NHS screening programmes.
Management of participants with AF:
Participants with confirmed AF will be invited to attend an appointment with the GP (or other healthcare professional in accordance with local policies) to discuss the diagnosis and the appropriate management including initiation of anticoagulation therapy.
Incidental cardiac findings:
The algorithm that interrogates each ECG trace can also detect and report other rhythm abnormalities in addition to AF. Such ECGs will be reviewed by a cardiologist and reported to the GP via the same process as diagnoses of AF outlined above if they are clinically relevant. The report will recommend to the GP if any further action is necessary, including possible referral. The PIS makes it clear to participants that other (non-AF) rhythm abnormalities may be detected as a result of screening.
This completes the participant involvement in the screening process. Baseline and follow-up data from the GP medical record will be collected in accordance with the consent provided.
Questionnaires – a random sample of patients (1,800; 1,500 invited to screening and 300 not invited to screening, matched for age, sex and postcode deprivation score where possible) will be assessed by postal questionnaire at baseline (pre-invitation) and post-invitation (8 weeks and 6 months). Participants will be sent the questionnaire (consisting of 3 validated measures to assess psychological and functional status) with a covering letter and freepost envelope. Consent will be assumed by completion and return.
Qualitative components:
In addition to the screening process outlined above a small number of patients will also be invited to participate in one or more qualitative components of the trial. The qualitative components of this programme of work will explore how screening is implemented at a practice level, and examine in some detail how practice staff and patients experience the AF screening intervention. Understanding the necessary and sufficient conditions for increased detection of AF and increased use of anticoagulation for screen-detected cases is essential in developing a feasible and sustainable screening programme, both within the main trial and nationally. These activities will be conducted in a small sub-sample of screening practices.
Interviews – participants from various stages of the trial will be invited by the trial team to take part in interviews, either by telephone or face-to-face. These may be one-off interviews (patients who decline the offer of taking part in the trial but agree to speak to a researcher about their reasons, and participants who decline the offer of screening), longitudinal interviews (following individual participants through the screening process with up to 4 interviews, the initial interview face-to-face with subsequent interviews conducted by telephone) or graphic elicitation interviews (interviewees will be supported by the interviewer to complete relational and/or sequential diagrams outlining their medical history, the meaning and implications of an AF diagnosis within this history, and the potential impact of anticoagulation on their daily medication regimens). Interviews will be audio recorded. Please see A30-1 for more details on the consent approach for interviews.
Observations - ethnographic observations of relevant practice procedures, alongside repeated visits to observe the day-to-day workings of the practice (including time spent in reception, waiting rooms, administrative offices, and patient consultations (if appropriate)), to inform understanding of the context within which the screening programme is being delivered. These will result in the observation of patients.
2. Practice staff
Interviews – the trial team will identify staff involved in the delivery of screening at the practice and approach them directly with an invitation to take part in an interview. Interviews will be conducted with GPs, nurses, practice managers, receptionists, other administrative staff, involved in the delivery of the screening programme to explore reactions to and attitudes towards the various components of the trial AF screening process. Interviews may be repeated as the delivery of the AF screening programme progresses.
Ethnographic observations - ethnographic observations of relevant practice meetings and other practice procedures, alongside repeated visits to observe the day-to-day workings of the practice (including time spent in reception, waiting rooms, administrative offices, and patient consultations (if appropriate)), to inform understanding of the context within which the screening programme is being delivered. No photos or videos will be taken. Field notes will not include any identifiable information.
3. External stakeholders
Part of the practice case studies interview work may involve identifying relevant external stakeholders to the practice and inviting them to take part in an interview. These individuals may include local care commissioners, clinical pharmacists, training providers, for example. Interviews may be repeated as the delivery of the AF screening programme progresses.
Intervention type
Other
Phase
Drug names
Primary outcome measure
Stroke incidence, defined as stroke of any severity but excluding transient ischaemic attack. Ischaemic and haemorrhagic stroke will be combined. Collection of stroke incidence will be via electronic health registries (including HES, ONS and SSNAP) and GP medical records, approximately annually for an average of 5 years.
Secondary outcome measures
Collected via electronic health registries (including HES, ONS, MiNAP and SSNAP) and GP medical records, approximately annually for an average of 5 years:
1. Bleeding episode requiring hospital admission
2. Myocardial infarction
3. Cardiovascular events (myocardial infarction + stroke + other admissions for cardiovascular disease – including heart failure)
4. All-cause mortality
5. Cardiovascular mortality
6. Dementia
7. AF detection rates
8. Anticoagulation rates
9. New diagnoses of depression and anxiety
10. Psychological effects of screening and impact on functional status assessed using the Spielberger state trait anxiety inventory (STAI), EQ5-D and Short-form 8 (SF-8) questionnaire posted to a random sample of participants in intervention and control practices at baseline and post-invitation to the trial
Overall trial start date
08/08/2017
Overall trial end date
31/03/2026
Reason abandoned (if study stopped)
Eligibility
Participant inclusion criteria
1. Have given informed consent to participate
2. Be aged 70 years or over
Participant type
All
Age group
Senior
Gender
Both
Target number of participants
Planned Sample Size: 126000; UK Sample Size: 126000
Participant exclusion criteria
1. Long-term anticoagulation therapy for stroke prevention
2. On the palliative care register
3. Resident in a nursing/care/residential home
Recruitment start date
01/04/2020
Recruitment end date
31/12/2022
Locations
Countries of recruitment
United Kingdom
Trial participating centre
NIHR CRN: Eastern
NR1 1QQ
United Kingdom
Trial participating centre
NIHR CRN: North East and North Cumbria
NE3 3HD
United Kingdom
Trial participating centre
NIHR CRN: North West Coast
L7 8XP
United Kingdom
Trial participating centre
NIHR CRN: Yorkshire and Humber
S10 2SB
United Kingdom
Trial participating centre
NIHR CRN: Greater Manchester
M13 9WL
United Kingdom
Trial participating centre
NIHR CRN: East Midlands
LE1 5WW
United Kingdom
Trial participating centre
NIHR CRN: West Midlands
CV3 2TX
United Kingdom
Trial participating centre
NIHR CRN: West of England
BS1 2NT
United Kingdom
Trial participating centre
NIHR CRN: Thames Valley and South Midlands
OX3 9DU
United Kingdom
Trial participating centre
NIHR CRN: Kent, Surrey and Sussex
ME8 0NZ
United Kingdom
Trial participating centre
NIHR CRN: Wessex
SO30 2UN
United Kingdom
Trial participating centre
NIHR CRN: South West Peninsula
PL6 8BX
United Kingdom
Trial participating centre
NIHR CRN: North Thames
W1T 7HA
United Kingdom
Trial participating centre
NIHR CRN: South London
SE1 9RT
United Kingdom
Trial participating centre
NIHR CRN: North West London
W12 0HT
United Kingdom
Sponsor information
Organisation
University of Cambridge
Sponsor details
c/o Carolyn Read
University of Cambridge School of Clinical Medicine
Box 111 Cambridge Biomedical Campus
Cambridge
CB2 0SP
United Kingdom
+44 (0)1223 769291
cad50@medschl.cam.ac.uk
Sponsor type
University/education
Website
Funders
Funder type
Government
Funder name
NIHR Central Commissioning Facility (CCF); Grant Codes: RP-PG-0217-20007
Alternative name(s)
Funding Body Type
Funding Body Subtype
Location
Results and Publications
Publication and dissemination plan
The researchers will publish the trial protocol in due course. Planned publication in a high-impact peer reviewed journal.
IPD sharing statement
The data sharing plans for the current study are unknown and will be made available at a later date/
Intention to publish date
31/03/2027
Participant level data
To be made available at a later date
Basic results (scientific)
Publication list