Immediate Ambulatory Electrocardiographic Monitoring in Syncope

Authors, Journal, Affiliations, Type, DOI

• Authors: Matthew J. Reed (first author); Steve Goodacre; Christopher J. Weir; Lynn Dinsmore; Amy Nugent; Titouan Kennel; Christopher C. Lang; Neil R. Grubb; Praveen Thokala; Karen Kilcoyne; Rachel O'Brien; Caroline Blackstock; Shona McDonald; Nicola Freeman; Ellise Clarke; Imad Adamestam; Louise Taylor — for the ASPIRED Investigators
• Affiliations: Emergency Medicine Research Group Edinburgh (EMERGE), Royal Infirmary of Edinburgh; University of Edinburgh; Edinburgh Clinical Trials Unit; and 45 UK National Health Service tertiary and district hospitals
• Journal: New England Journal of Medicine (published online 31 August 2026; early-release article)
• Type: Open-label, randomized, controlled superiority trial
• DOI: 10.1056/NEJMoa2605812
• Registration: ISRCTN10278811
• Funding: British Heart Foundation clinical study grant (CS/F/20/190016) and an NHS Research Scotland Career Researcher clinician award (to M.J. Reed). Boston Scientific provided all the BodyGuardian Mini 14-day ambulatory heart monitors used in the trial but had no other involvement. The funders had no role in design, data collection, analysis, interpretation, or writing

Overview

ASPIRED asked whether starting 14-day ambulatory ECG monitoring at the index emergency-department visit improves outcomes in patients whose syncope remains unexplained after ED assessment. In 2,234 patients across 45 UK hospitals it did not: patient-reported syncope episodes at 1 year were 1.37 versus 1.58 (IRR 0.89; 95% CI 0.68–1.18; P=0.42). The trial's value lies less in that null than in the dissociation it exposes — monitoring more than tripled the detection of clinically significant cardiac arrhythmia (22.0% vs 9.0%), substantially shortened the time to detection (22 vs 54.5 days), and increased downstream antiarrhythmic and pacemaker therapy, yet patients reported no reduction in syncope. All-cause mortality at 1 year was halved (1.5% vs 2.9%), but as a secondary outcome with no multiplicity adjustment it can only be hypothesis-generating.

Keywords

Syncope, unexplained syncope, emergency department, ambulatory electrocardiographic monitoring, cardiac monitoring, patch monitor, BodyGuardian Mini, arrhythmia detection, diagnostic yield, syncope recurrence, pacemaker implantation, implantable loop recorder, pragmatic trial

Key Takeaways

Question and Rationale

• Syncope accounts for ~650,000 ED visits annually in the UK and ~1.2 million in the US; no cause is identified after ED assessment in approximately 50% of patients
• Clinically important pathology is found in ~7% of patients with syncope within 1 month of an ED visit; diagnosing arrhythmic syncope requires symptom–ECG correlation, but conventional monitoring has limited availability and low yield
• Syncope recurs in ~50% of patients within 1 year, driving repeat visits and impaired quality of life; once an arrhythmia is identified and treated, 1-year recurrence falls to ~10%
• Prior evidence suggested arrhythmia detection is most likely when monitoring is initiated early, ideally at the index ED visit — the hypothesis ASPIRED set out to test

Design

• Open-label, randomized, controlled superiority trial in the emergency departments and acute medical units of 45 UK NHS tertiary and district hospitals
• Sites were eligible only if they were not already using cardiac monitoring at ED discharge as standard care — so standard care was genuinely free of the intervention
• Randomization: web-based, managed by the Edinburgh Clinical Trials Unit, 1:1, stratified, computer-generated sequences
• Patients were ≥16 years (the UK adult emergency-medicine threshold) presenting with syncope unexplained after initial assessment
• 2,234 randomized: 1,123 intervention, 1,111 standard care. Mean age 58.3 years, 52.1% male, ~90% white
• Primary analysis population after excluding those with no follow-up: 1,970 (1,004 intervention, 966 standard care)
• Patient and public involvement informed design, outcome selection, follow-up methods and patient-facing materials
• Follow-up: patient-reported data collected every 4 weeks for 2 years by automated text message or email, with telephone follow-up for non-responders; clinical outcomes and resource use from routine electronic health records
• Intention-to-treat throughout

Intervention and Comparator

• Intervention: 14-day ambulatory heart monitor (BodyGuardian Mini, Boston Scientific) fitted as soon as possible after randomization, recording rhythm continuously, with a patient-triggered event button
• If the monitor identified a serious arrhythmia, the patient's local team was notified within 24 working hours; treatment decisions remained at the treating clinician's discretion
• Standard care: whatever each site normally provided for unexplained syncope, which could include inpatient telemetry and ambulatory monitoring chosen on the basis of the patient's presentation
• Both groups could be referred for syncope-specialist evaluation, ideally within 6 weeks of the index event — this is important for interpretation: the trial tested early monitoring, not specialist care versus no specialist care
• Adherence: 930 of 1,122 (82.9%) had an analysable ECG recording; 192 either did not wear or did not return the device. Mean wear time 11.7±3.1 days in 926 patients; 825 met the adherence threshold of ≥7 days

Primary Outcome — Null

• Primary outcome: mean number of patient-reported syncope episodes at 1 year
• 1.37±5.10 (intervention) vs 1.58±8.56 (standard care); incidence rate ratio 0.89 (95% CI 0.68–1.18); P=0.42 — not significant
• Results were consistent across all prespecified sensitivity analyses
• Analysed by negative binomial regression with site as a random effect and follow-up duration as an offset term

Secondary Outcomes — Where the Trial Becomes Interesting

The secondary outcomes point consistently in a positive direction while the primary does not:

• Implantable loop recorder placement was essentially identical in both arms (7.5% vs 7.6%). This is the natural downstream consequence of a 14-day monitor being non-diagnostic: both groups proceeded to ILR at the same rate, so early patch monitoring did not reduce — or increase — the eventual need for an ILR. It is directly relevant to the wiki's existing claim that ICM has 55% diagnostic yield versus 19% for conventional testing, since ASPIRED shows that yield advantage did not change how often an ILR was ultimately implanted
• Numbers of short-term monitors were lower per patient in the intervention arm (0.29 vs 0.40; IRR 0.72), consistent with the 14-day patch substituting for other short-term monitoring rather than adding to it

Arrhythmia subtypes detected more often with monitoring, with wide intervals:

• Non-sustained VT: 109 (10.0%) vs 13 (1.2%); OR 9.50 (5.29–17.06)

• Sustained VT: 16 (1.5%) vs 6 (0.6%); OR 2.67 (1.04–6.86)

• Pause ≥6 seconds: 30 (2.7%) vs 13 (1.2%); OR 2.32 (1.20–4.48)

• Bradycardia: 19 (1.7%) vs 8 (0.7%); OR 2.37 (1.02–5.49)

• Second-degree AV block Mobitz II: 13 (1.2%) vs 7 (0.6%); OR 1.85 (0.73–4.65)

• Complete or third-degree heart block: 11 (1.0%) vs 8 (0.7%); OR 1.36 (0.55–3.41)

• The authors' own multiplicity caveat: "For secondary outcomes, the widths of the confidence intervals have not been adjusted for multiplicity and should not be used to infer treatment effects." This applies to the mortality result and to every row above

• Approximately seven patients needed to be monitored to detect one clinically significant arrhythmia

The Central Finding — Detection Did Not Translate Into Benefit

• Monitoring achieved exactly what it was designed to achieve on the diagnostic axis: detection 22.0% vs 9.0%, and much earlier (median 22 vs 54.5 days)
• It also changed treatment: more antiarrhythmic initiation, more pacemakers, more ILR placement
• Yet the patient-reported primary outcome did not improve, and syncope recurrence was identical (27.9% vs 28.6%)
• This is the trial's principal contribution: a large, well-powered randomized demonstration that improving diagnostic yield and time-to-diagnosis in unexplained syncope does not by itself reduce syncope, at least not within 1 year and not from monitoring initiated at the ED visit alone
• It also qualifies the reasoning that motivated the trial — the observation that treated arrhythmic syncope has ~10% recurrence — since the causal chain from detection → treatment → fewer episodes did not hold at trial scale

Safety and Acceptability

• A total of 57 adverse events: 49 in the intervention group and 8 in the standard-care group (50 mild, 5 moderate, 2 severe)
• Most intervention-group events were minor skin irritation from the device
• Two serious adverse events, one in each group, both judged unrelated to the trial intervention: infection of an implantable loop recorder requiring device removal (intervention) and a haematoma requiring outpatient antibiotics (standard care)
• Device acceptability was high: 91.3% found it easy to use, 73.3% comfortable, 82.5% willing to use it again, 91.9% reported no restriction in daily activities
• The 49-versus-8 imbalance is a real cost of the intervention even though the events were minor: a monitoring strategy generates adverse events in the group that receives it

Limitations of the document

• Open-label design with the potential for performance bias; the authors note that follow-up intensity — including duration of index hospital stay, clinic attendance and cardiac investigations — was similar between groups
• The primary outcome is subjective and patient-reported, and therefore susceptible to detection and reporting bias. The authors state that objective secondary outcomes are less susceptible to bias and "probably more robust" — which cuts both ways, since those are the outcomes showing benefit
• Secondary outcomes were not adjusted for multiplicity, and the trial explicitly says their confidence intervals should not be used to infer treatment effects. The mortality finding (OR 0.50) is the most quotable and the most caveated result in the paper
• Lower-than-anticipated event rates reduced statistical power. The power calculation assumed 42.5% 1-year recurrence with standard care; observed recurrence was ~28% in both groups. The trial may therefore have been underpowered for the effect size originally hypothesised, although consistency across sensitivity analyses supports the null
• Non-response for the primary outcome (questionnaire-based) was mitigated by consistency with medical-record review, but it remains a source of imprecision
• The trial tested monitoring initiated at the ED visit, not a complete syncope pathway. Both groups could access syncope-specialist referral ideally within 6 weeks, so the trial isolates one component of care rather than comparing a full early-monitoring strategy against usual care
• Standard care was site-defined and heterogeneous, though the eligibility criterion (sites not already monitoring at ED discharge) means it was free of the intervention
• Single-country health system (45 UK NHS hospitals), which may limit generalisability of both the standard-care comparator and the resource-use findings to other systems
• 1-year follow-up — a monitoring strategy might plausibly affect outcomes over a longer horizon, particularly for the mortality signal
• The follow-up narrative is only partly recoverable: the PDF conversion rendered the Discussion section as fragmented table content, so the authors' own mechanistic explanation for the detection–outcome dissociation could not be extracted in full. The results table, design, and limitations paragraph were readable and are reproduced above; the interpretive discussion between them is incomplete in this source file
• External validation is pending from the ongoing REMOSYNCED trial (NCT05066347) in different care settings

Key Concepts Mentioned

• concepts/Syncope — the condition; ASPIRED is the first randomized test of whether the diagnostic yield of ambulatory monitoring converts into patient benefit, and it does not
• concepts/Vasovagal-Syncope — the commonest cause of syncope and the population in which monitoring is least likely to be diagnostic
• concepts/Carotid-Sinus-Syndrome — a reflex cause that ambulatory ECG monitoring cannot detect, relevant to the trial's null result
• concepts/Orthostatic-Hypotension — another non-arrhythmic mechanism unaffected by rhythm monitoring
• concepts/Sudden-Cardiac-Death — relevant to the arrhythmia subtypes detected more often (non-sustained VT, pauses, bradycardia)
• entities/Atrial-Fibrillation — detected arrhythmia driving anticoagulation decisions in some trial patients

Key Entities Mentioned

• Boston Scientific — supplied the BodyGuardian Mini monitors for the trial and had no other involvement. No wiki page exists and none is warranted from this disclosure alone
• British Heart Foundation and NHS Research Scotland — funders with no role in design, analysis or writing

Wiki Pages Updated

• wiki/sources/Syncope-ASPIRED-NEJM-2026.md — created (this page)
• wiki/concepts/Syncope.md — monitoring sections qualified with the first randomized outcome evidence; new contradiction on diagnostic yield versus patient benefit
• wiki/concepts/Vasovagal-Syncope.md — relevance of the null result to the commonest syncope subtype
• wiki/sourceindex.md — index row
• log.md — ingest entry