Objectives Diagnosing underlying arrhythmia in ED syncope patients remains problematic. This study investigates diagnostic yield, event prevalence, patient satisfaction and compliance, and influence on resource utilisation of an ambulatory patch monitor in unexplained ED syncope patients.
Methods Prospective pilot study conducted in a single tertiary ED in Scotland between 17 November 2015 and 16 June 2017 with a historical unmatched comparator group. Patients 16 years or over presenting within 6 hours of unexplained syncope were fitted in the ED with an ambulatory patch ECG recorder (Zio XT monitor), which continuously records a single-lead ECG for up to 14 days. Patients with an obvious underlying cause were excluded. An unmatched historical group of 603 syncope patients with no obvious diagnosis in ED, recruited to a prior cohort study (2007–2008), were used as a comparator. Primary endpoint was symptomatic significant arrhythmia at 90-day follow-up.
Results During the prospective study period, 86 patients were recruited. 90-day diagnostic yield for symptomatic significant arrhythmia was 10.5% (95% CI 4.0 to 16.9; 9 of 86) versus 2.0% (95% CI 0.9 to 3.1; 12 of 603) in the comparator group. 24 patients (27.9%) had a significant arrhythmia (five serious); 26 patients (30.2%) had serious outcomes (major adverse cardiac event and/or death). Blinded patch report review suggested the patch would significantly reduce requirement for standard outpatient ambulatory ECG monitoring. 56 of 76 returned patches had a diagnostic finding within±45 s of a triggered/diary event (73.7% diagnostic utility; 95% CI 63.7 to 83.6); 34 of 56 (61%) for sinus rhythm or ectopic beats only.
Conclusions Routine, early ambulatory ECG monitoring in ED patients with unexplained syncope is probably warranted. A large-scale trial comparing this approach to standard care with cost-effectiveness and safety analysis is now required.
Trial registration NCT02683174.
- cardiac care, arrythmia
Statistics from Altmetric.com
What is already known on this subject
Diagnosing underlying arrhythmia in ED syncope patients is difficult.
Many patients wait months for investigations that commonly fail to detect the underlying reason for the syncopal episode.
A recent study (monitoring of SYNcopes and/or sustained palpitations of suspected ARRhythmic origin-Flash (SYNARR-Flash)) showed that placement of an early (0–30 days after index event) external 4-week ECG monitor in patients with unexplained syncope and palpitations of likely arrhythmic origin had a 4-week diagnostic yield of 24.5% with an increasing likelihood of diagnostic events if used in the first 15 days.
What this study adds
In this prospective pilot study, using an ambulatory ECG monitor in ED patients with unexplained syncope results in identification of a symptomatic significant arrhythmia in 1 in 10 patients (compared with 2 in 100 historical unmatched comparators) and a diagnostic finding in 3 in 4.
An early ambulatory monitoring strategy has the potential to change current management of syncope patients by reducing hospital admissions, changing first-line monitoring from low diagnostic yield Holter to higher yield patches and allowing earlier diagnosis and treatment of clinically important arrhythmias in turn reducing morbidity and increasing quality of life.
Syncope is a common ED presentation, but the underlying diagnosis is not apparent in 60% of patients after assessment1 and serious adverse event rate is 11.1% at 1 month.2 Most serious outcomes are cardiovascular. Reed et al 3 showed there were 50% more cardiovascular serious outcomes than non-cardiovascular serious outcomes at 1 year, which are also more likely to be undetected during ED assessment. Many cardiovascular events are due to arrhythmia, which is difficult for clinicians to diagnose as examination and ECG findings may both be normal and symptoms have resolved by the time the patient gets to the ED. Currently, establishing a cardiac arrhythmia as the cause of syncope rests on correlating the arrhythmia with symptoms using monitoring devices, but these all have significant drawbacks. Holters are bulky and inconvenient to wear, and non-compliance and lack of extended monitoring reduce diagnostic yield.4 Event recorders can monitor over longer periods of time but must be activated and cannot detect asymptomatic arrhythmias. External continuous loop recorders are expensive, require electrodes and bulky recording devices and produce a large amount of data, which requires sifting. Implantable loop recorders are expensive and necessitate an invasive surgical procedure. In order to solve these problems, several novel ambulatory cardiac monitoring devices that can easily be applied to ED patients have recently been developed.
The SYNARR-Flash study5 is the only previous study evaluating the role of an early external 4-week ECG monitor in patients with unexplained syncope and palpitations of likely arrhythmic origin. This study found a 4-week diagnostic yield of 24.5% with early recorder use (applied within 15 days from index event), history of supraventricular arrhythmia and frequent previous events increasing the likelihood of diagnostic events during the 4-week monitoring period. However, this study enrolled patients within 1 month after unexplained syncope. We therefore decided to look at the effect of starting monitoring very early (during ED attendance with the index event), a strategy that could potentially change current management of syncope in the ED. The reason for conducting a pilot study was to investigate whether the diagnostic yield of clinically important arrhythmias in this specific patient group was sufficient to warrant a large-scale trial.
The aim of this study is to investigate the diagnostic yield, prevalence of events, patient satisfaction, patch compliance and influence on subsequent treatment of an ED-placed ambulatory patch monitor in ED patients presenting with syncope, which remained unexplained after ED evaluation.
Study design and setting
This was a single-centre, prospective pilot study with a historical unmatched comparator group. The study was conducted in the ED of the Royal Infirmary of Edinburgh, a UK tertiary centre with 120 000 adult attendances per annum. Written consent was obtained from all patients.
Patients aged 16 years or over who presented between 17 November 2015 and 16 June 2017 within 6 hours of an episode of syncope and whose syncope remained unexplained after ED assessment were enrolled.
Exclusion criteria were obvious underlying cause after ED assessment (defined as: (A) clinical history of vasovagal syncope, ie, presyncope symptoms and low-risk patient according to current ESC (European Society of Cardiology) guidelines at the time of study commencement6 (incidentally revised in 2018 after the study had completed2); (B) arrhythmia on ED ECG thought to have caused syncope; (C) arrhythmia on prehospital ECG causing syncope; (D) pulmonary embolism diagnosed on computed tomography pulmonary angiography (CTPA) (or equivalent, eg, symptoms of PE plus positive leg ultrasound scan/ventilation–perfusion scan (USS/VQ)/echocardiography); (E) postural hypotension (postural drop >20 mm Hg in ED with presyncope symptoms during test and suggestive history); (F) myocardial infarction7; (G) CT brain or clinical signs/symptoms in ED showing cerebrovascular accident or subarachnoid haemorrhage; (H) evidence of haemorrhage in ED thought to have caused syncope; (I) other obvious cause of syncope apparent in ED: alcohol or illicit drugs as presumptive cause, epileptic seizure as presumptive cause (seizure activity with a >15 min witness reported postictal phase), stroke/transient ischaemic attack as presumptive cause, preceding head trauma, hypoglycaemia as presumptive cause; or (J) no consent available, that is, patient lacking capacity, previous recruitment into the study or patient in custody or prison).
The historical unmatched comparator group were 603 of 1067 patients who were recruited to the ROSE (risk stratification of syncope in the emergency department) study between 1 March 2007 and 21 July 2008 with no obvious diagnosis in ED according to the treating clinician and able to be followed-up at 90 days.1 The same exclusion criteria had been applied to the historical unmatched comparator group as were applied in the PATCH-ED (diagnostic yield of an ambulatory patch monitor in unexplained emergency department syncope) pilot study group. They were identified from the ‘intermediate risk’ pathway from the local ED syncope evidence-based guideline. These patients were initially assessed by ED medical staff and, depending on individual clinical factors, received one of the following: (1) discharge to general practitioner with outpatient investigation; (2) referral to a general physician for investigation and ongoing care; (3) referral to cardiologist for investigation and ongoing care; and (4) admission to medical or cardiology unit for immediate investigation. Echocardiography and standard (up to 72 hours) ambulatory ECG monitoring and patient-activated event monitoring were available as standard, but the clinical service did not include extended (eg, patch) ECG monitoring.
Usual care in PATCH-ED pilot study
Potentially eligible patients were identified by clinical staff in the ED and assessed for study inclusion by the attending clinician. If the patient fulfilled the study eligibility criteria, written consent was taken. A decision to enrol a patient was not later overturned. Patients had a case report form completed in the ED, comprising demographic, historical and examination characteristics, 12-lead ECG and radiology and standard laboratory investigations,8 and subsequently entered into a specially designed electronic database. ED tests not part of the study protocol were ordered at the discretion of the treating clinician, and patients were admitted, referred for outpatient investigation or discharged according to existing ED protocols and at the discretion of the treating clinician. If a patient was admitted, then the reason for admission was recorded.
All enrolled patients were fitted in the ED with a novel ambulatory patch monitor (Zio XT monitor, iRhythm Technologies, San Francisco, California, USA), which continuously records a single-lead ECG for up to 14 days. This was placed on the patient in the ED by a trained research team member and was left on for 2 weeks. The study participant was also given a diary in which to record any symptoms occurring during the recording period, for example, light-headedness/dizziness, syncope/presyncope and a patch satisfaction questionnaire. This was a bespoke unvalidated questionnaire designed specifically for this study. The study participant was given details of the PATCH-ED research team in case of any patch monitor problems during the wear period. At the end of the patch period, the patient sent the patch back to the iRhythm Clinical Centre for data processing by a certified electrocardiographic technician specialised in advanced arrhythmia detection adhering to Medicare Independent Diagnostic Performance Standards.9 The patch report was then conveyed to the Edinburgh study team electronically by iRhythm. The time for any arrhythmias to appear on the ambulatory patch will therefore be less than the time for clinical detection as the latter includes postal transit, analysis of data, interpretation and reporting process steps prior to the clinical team becoming aware of the arrhythmia.
Any study participant with a serious significant arrhythmia (see figure 1 for definition) on the patch report was contacted immediately and appropriately referred to the hospital cardiac electrophysiology service. Patients were followed up at 90 days after presentation through hospital and primary care electronic patient record (EPR) systems. Patients who had left the NHS Lothian area after 1 month were contacted by telephone. Serious significant arrhythmia and significant arrhythmia (figure 1) were defined based on Standardized Reporting Guidelines for ED Syncope Risk Stratification Research8 and were recorded as an endpoint whether diagnosed by ambulatory patch monitor or usual care.
Arrhythmias were also defined as symptomatic (ie, concurrent light-headedness/dizziness and syncope/presyncope with arrhythmia) noted via triggered and diary entries or asymptomatic. Any significant symptomatic arrhythmia was discussed with the hospital cardiac electrophysiology service. Non-recruited but potentially eligible patients were identified by a daily search of all ED EPRs to assess for potential recruitment bias.
Outcome measures and assessment
The primary endpoint was symptomatic significant arrhythmia at 90-day follow-up. Diagnostic utility was also measured and was defined as the percentage of returned patches with a diagnostic finding (both significant arrhythmia and sinus rhythm/ectopic beats) associated with symptoms.
Secondary endpoints were: (A) median time to detection of symptomatic significant arrhythmia; (B) prevalence during the 90-day follow-up of arrhythmia (serious significant arrhythmia, significant arrhythmia and symptomatic arrhythmia); (C) patient patch monitor satisfaction (brief postal questionnaire including Likert scales, yes/no questions and option for free-text comments; see online supplementary appendix 1); (D) patch monitor compliance (median device wear time); (E) number of patients requiring referral and/or cardiac investigations (standard outpatient ambulatory ECG monitoring, inpatient and outpatient echocardiogram and treadmill testing) with and without the findings of the ambulatory patch monitor; and (F) all-cause serious outcome at 90 days. All-cause serious outcome was defined as death and/or major adverse cardiac event (MACE) comprising myocardial infarction,7 significant arrhythmia,8 significant structural heart disease,8 positive electrophysiology study findings,8 permanent pacemaker or defibrillator placement, coronary artery bypass graft or coronary artery stent, cardiac valve surgery, elective cardioversion in the absence of objective evidence that tachyarrhythmia is responsible for the syncope, balloon pump insertion, heart transplant, initiation of antiarrhythmia medical therapy and ventricular assist device.
Secondary endpoint (E) involved blinded review of the clinical notes and ambulatory patch monitor reports by two cardiologists with electrophysiology expertise. These experts decided independently using their experience and standard practice whether prior to the ambulatory patch monitor report being available, each patient would have required further specialist follow-up and/or cardiac investigations (and which ones). This assessment was then repeated after viewing the ambulatory patch monitor report and reported for each expert individually. This process was not the same as endpoint adjudication.
Eighty-five patients were required to be enrolled to show a sixfold increase in the proportion of patients reaching the primary endpoint in the intervention group (12%) compared with the historic standard care strategy group (2%) with an 80% power at a two-sided 5% significance level assuming the historical comparator group was six times the size of the intervention group. At the time of designing the study, no comparable data on ambulatory patch monitors were available. A sixfold improvement in primary endpoint was chosen as four previous studies10–13 comparing implantable loop recorders with standard diagnostic assessment showed between a twofold and ninefold increase in syncope aetiological diagnosis. We aimed to enrol 100 patients to ensure any unreturned or lost ambulatory patch monitors would not affect our primary and secondary aims.
Data analysis and statistics
Two researchers (MJR and AM) independently reviewed all clinical data including patch monitor findings and assigned endpoints with any disagreements resolved by consensus. Categorical outcomes were compared between the intervention and historical comparator groups using the χ2 test. Normally distributed variables are presented as mean±SD; otherwise, the median and IQR were used as summary measures.
Eighty-six patients were recruited to the study. Baseline characteristics are shown in table 1. Figure 2 details the study recruitment diagram. Ten ambulatory patch monitors were not returned or lost in transit. Although only 76 patients had ambulatory patch monitors informing their outcomes, arrhythmia could still potentially be detected via usual care in these patients; therefore, all 86 patients have informed the primary endpoint, but 76 patients are included in ambulatory patch specific results.
Nine of 86 patients in the PATCH-ED study had a symptomatic significant (including serious) arrhythmia endpoint (table 2). Diagnostic yield of the patch monitor for symptomatic significant (including serious) arrhythmia was 10.5% (95% CI 4.0 to 16.9; 9 of 86) compared with 2.0% (95% CI 0.9 to 3.1; 12 of 603) in the comparator group. Fifty-six of 76 (73.7% diagnostic utility; 95% CI 63.7 to 83.6) returned ambulatory patch monitors had a diagnostic finding within ±45 s of a triggered or diary event with 34 of 56 (61%) being for symptomatic sinus rhythm or ectopic beats only.
Time to detection
In the PATCH-ED study, the median time to clinical detection (ie, clinical team becoming aware of arrhythmia) of symptomatic significant arrhythmia was 19 (IQR 4–30) days (n=9). This compares with a median time to detect a significant symptomatic arrhythmia of 8 (IQR 5–17) days (n=12) in the comparator group.1 Eight of the 9 PATCH-ED symptomatic significant (including serious) arrhythmias were first detected by the ambulatory patch monitor (table 3). One patient presented to the ED in symptomatic complete heart block, 3 days after the index ED attendance. The ambulatory patch monitor was removed and sent away and ultimately reported this serious significant arrhythmia but obviously after the diagnosis had already been made through usual care.
Figure 3 shows the time in days for the first detection on the ambulatory patch monitor of symptomatic significant (including serious) arrhythmias, and of all significant (serious and non-serious) arrhythmias whether they were symptomatic or not (n = 24). Online supplementary efigure 1 shows the cumulative incidence of detection of symptomatic significant arrhythmias in ambulatory patch and historical unmatched comparator groups. .
Prevalence of arrhythmias (serious significant, significant and symptomatic arrhythmia) at 90-day follow-up
In the PATCH-ED study, 24 patients had a significant arrhythmia detected by 90 days with 5 of these having a serious significant arrhythmia. All but one patient was diagnosed by the ambulatory patch monitor. This patient was diagnosed through usual care strategies, but the arrhythmia was also present subsequently on the patient’s ambulatory patch monitor report (table 3).
Patient patch satisfaction
Reponses to the patient patch monitor satisfaction questionnaire were received from 47 patients (55%). Forty-three patients (91% of respondents) agreed or strongly agreed that the patch monitor was easy to use, and 34 (72%) agreed or strongly agreed that the patch monitor was comfortable to wear. Thirty-nine patients (82%) agreed or strongly agreed that they were able to carry out normal activities, and 38 patients (80%) would use the patch monitor if required in the future. Nineteen participants (40%) submitted free-text comments that indicated the patch irritated the skin (six participants, 13%) and the patch lost adherence to the skin (seven participants, 15%) (see online supplementary etable 1).
Patch monitor compliance
Median ambulatory patch monitor wear time was 13.6 days (IQR 11.8–14.0; online supplementary etable 2).
Patients with ambulatory patch ECG monitor findings requiring referral
After ED assessment (prior to the ambulatory patch monitor reports being available), no patients had been referred to the hospital cardiac electrophysiology service (table 1). By 90 days, 12 patients had been referred (table 3).
Blinded review of the ambulatory patch monitor reports by the two cardiologists suggested the patch monitor would significantly reduce requirement for standard outpatient ambulatory ECG monitoring from 84% to 1% (expert 1) and from 81% to 8% (expert 2). There would be a non-significant increase in outpatient echocardiograms from 28% to 49% (expert 1) and from 15% to 18% (expert 2) and a non-significant increase in those patients requiring no further cardiac investigation from 11% to 29% (expert 1) and from 14% to 77% (expert 2) (online supplementary etable 3 and 4).
All-cause serious outcome at 90 days
In the PATCH-ED study at 90 days, 26 patients (30.2%) had a serious outcome (defined as MACE or death).
In this prospective pilot study using an ambulatory ECG monitor in ED patients with unexplained syncope, we detected a symptomatic significant arrhythmia in 1 in 10 patients (compared with 2 in 100 in the historical unmatched comparator group) and a diagnostic finding associated with symptoms for every three of four monitored patients.
Diagnosing an arrhythmic cause for patients presenting to the ED with syncope remains problematic.14 15 The gold standard (ESC syncope guidelines2 6) for the diagnosis of syncope is when a correlation between symptoms and a documented arrhythmia is recorded.2 6 In the absence of this correlation, the presence of some asymptomatic significant arrhythmias is also diagnostic.2 6 The ESC syncope guidelines2 6 also suggest that ECG monitoring is indicated only when there is a high pretest probability of identifying an arrhythmia associated with syncope. There is clearly a balance between the early and effective detection of culprit arrhythmias versus detecting incidental non-culprit rhythms, and it is this conflict that has led to ambulatory ECG monitoring remaining within the confines of elective outpatient cardiology services in most healthcare settings.
Although ambulatory ECG monitoring has been studied in an ED population before,16 and in other emergency presentations,17 18 this is the first study using ambulatory ECG solely in patients presenting to the ED with syncope and also the first fitting the monitor at the time of the index presentation. This strategy has the potential to change current management of syncope patients by reducing hospital admissions, changing first-line monitoring from low diagnostic yield Holter to higher yield patches and allowing earlier diagnosis and treatment of clinically important arrhythmias in turn reducing morbidity and increasing quality of life.
In both the ambulatory patch monitor and historical unmatched comparator group, we excluded those patients who had an obvious diagnosis for their ED syncope presentation, including an arrhythmic diagnosis, as clearly these patients did not require further ECG monitoring to make a diagnosis. We also excluded all low-risk patients who had a likely vagal or postural cause of syncope in both groups. It is this remaining ‘intermediate’ group that are so difficult for physicians to manage. They are at risk of a cardiac cause for their syncope, but it is not clear what is the best management strategy for them.
We also only selected symptomatic arrhythmias (the ESC gold standard for the diagnosis of syncope) as our primary endpoint ensuring that incidental non-culprit rhythms (even asymptomatic serious arrhythmias), although recorded, were not included in our primary endpoint. In spite of this high threshold for arrhythmia detection, 10.5% of this ‘intermediate’ group had a primary endpoint and 30% of patients had a serious outcome. These data suggest that an aggressive early ECG monitoring approach in patients presenting to the ED with syncope unexplained after ED evaluation (the ‘intermediate’ risk group) may be warranted.
The proportion of patients in the historical control group with cardiac risk factors (eg, known ischaemic heart disease, clinical signs of heart failure present, previous myocardial infarct and heart murmur heard on examination) is higher than in the current patient population, which would suggest that the lower diagnostic rate for arrhythmia in the historical group is not likely to be due to a healthier population. The proportion of patients admitted in the historical control group was higher than in the current patient population. One could speculate that the placement of the ambulatory ECG monitor made ED clinicians more accepting of discharge.
The diagnostic yield for important arrhythmias in our study using the Zio XT monitor compares favourably to studies looking at the diagnostic yield of other available ambulatory ECG monitoring devices.19 Four studies of ambulatory monitoring for periods between 24 hours and 48 hours20–23 showed incidences of syncope with ECG abnormality during the monitor time of between 0% and 1%. Four studies of external loop monitoring for periods between 4 weeks and 8 weeks showed incidences of syncope with ECG abnormality during the monitor time of between 1% and 24.5%.5 21 24 25 Our diagnostic utility of 74% also compares favourably to other studies (14%–56%) with 61% being for symptomatic sinus rhythm or ectopic beats only. This is another important finding: ECG documentation of the absence of an arrhythmia during a similar episode of syncope suggests a low likelihood of arrhythmia being the cause and may allow a clinician to reassure and discharge a patient safely or to focus further assessment on non-arrhythmogenic causes.
In order to assess the impact on cardiology services from an early ECG monitoring approach in patients presenting to the ED with syncope, we looked at the blinded expert review of the patch monitor ECG reports that showed a significant reduction in the need for standard outpatient ambulatory ECG monitoring. The median time to the clinical team becoming aware of a significant symptomatic arrhythmia was greater in the ambulatory patch monitor cohort compared with the historical unmatched comparator group (19 days vs 8 days). The wear duration of monitoring contributes to the majority of this difference as usual care monitoring is around 1–3 days and ambulatory patch monitoring is 14 days. Delay for return of the device, analysis, report creation and notice to the clinician is therefore around 5 days for both strategies. Combining the patch ECG monitor technology with inhouse analysis or an immediate upload facility for patients may speed up the time to diagnosis in both groups. The incidence of detection seems greater at all time points for the ambulatory patch monitor cohort (notwithstanding the possibility of censoring) due to the greater incidence of significant symptomatic arrhythmia detection.
It is unclear presently how long patients selected for ambulatory monitoring should be monitored and various times have been suggested from 24 hours to 28 days. In our study, the majority of symptomatic significant arrhythmias are captured on the ambulatory patch ECG monitor within the first 7 days. All serious symptomatic significant arrhythmias were captured within 4 days, but some significant arrhythmias (mainly non-serious and asymptomatic) were picked up between days 8 and 14. It is likely that an optimum ambulatory ECG monitoring period is therefore somewhere between 7 and 14 days and that shorter term (24–48 hour) Holter type monitoring is not sufficient in this patient group.26
While there is a lag time in the detection of arrhythmia using the ambulatory ECG monitor, the earlier diagnosis of clinically important arrhythmias compared with historical controls could well be safer and associated with less morbidity than standard care. Whether ED clinicians would be happy in practice to discharge patients from the ED with ambulatory ECG monitoring will require an interventional trial. Of note: (1) admission for syncope has never been shown to improve outcome and (2) while the most recent estimates of death (0.8%) and non-fatal severe outcome (10.3%) within 30 days seem high, two-thirds of this non-fatal severe outcome (6.9%) is identified in the ED.2
Our study has some limitations. The comparator group was a historical cohort. The patients from this cohort had the same inclusion and exclusion characteristics, were collected from the same centre and also recruited during index ED visit but were enrolled between 2007 and 2009. Care of patients with syncope may have changed in intervening years. The study was at a single centre, and it was not in the scope of the study to perform cost-effectiveness analysis. It must also be noted that further assessment of the role and benefit of patch monitoring in other healthcare systems is needed before widespread adoption as access to investigations such as ambulatory patch ECG monitoring from the ED may vary considerably.
Patients with unexplained syncope had a sizeable 90-day serious outcome. The diagnostic yield of an ambulatory patch ECG monitor for important symptomatic arrhythmias in patients presenting to the ED with syncope unexplained after ED evaluation was fivefold better than in a historical unmatched comparator group and revealed a symptomatic significant arrhythmia in 1 of 10 patients. An aggressive early ECG monitoring approach in this unexplained group is probably warranted, but a large-scale multicentre, open-label randomised controlled trial of an early ambulatory patch ECG monitor approach versus standard care with cost-effectiveness and safety analysis is now required to confirm these findings and assess the impact on clinical pathways and patient and health service outcomes.
Thanks to the medical and nursing staff at the Royal Infirmary of Edinburgh for their support with PATCH-ED and to the Emergency Medicine Research Group Edinburgh research nurses.
Contributors MJR, NRG, AJG, CCL and KS were responsible for the conception and design of the study; KS and AM were responsible for acquisition of data; MJR, AM and CJW were involved in data analysis; and all authors were involved in drafting the article and revising it critically for important intellectual content. All authors approve the manuscript.
Funding This study was funded by a minor research award from Chest, Heart and Stroke Scotland (£4950). MJR was supported by an NHS Research Scotland Career Researcher Clinician award. iRhythm Technologies, Inc. provided the Zio XT monitors and ECG analysis service free of charge. CJW was supported in this work by NHS Lothian via the Edinburgh Clinical Trials Unit.
Disclaimer iRhythm Technologies, Inc. and the funder had no involvement in the design, conduct, analysis or reporting of the study.
Competing interests None declared.
Patient consent Not required.
Ethics approval The study was granted ethical approval by the South East Scotland Research Ethical Committee (15/SS/0072).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All data are available on request.
Presented at This work was presented at the European Heart Rhythm Association Congress on 18–20 March 2018.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.