Syncope is one of those presentations that can be both straightforward and remarkably difficult. In many patients the history gives us a pretty convincing diagnosis of vasovagal or orthostatic syncope, but there remains a group where the diagnosis is uncertain and where we worry about an underlying cardiac arrhythmia. The problem is that arrhythmias are typically intermittent. A normal ECG in the emergency department tells us what the rhythm is at that moment, but not necessarily what happened immediately before the patient collapsed, or what might happen next week, and some of those rhythms are potentially life threatening.
That leaves us with the familiar problem of deciding who can go home, who needs admission and who needs further investigation. In that regard, there is great uncertainty, multiple guidelines and many studies. In Virchester we have risk assessed the patients and sometime admit patients for cardiac monitoring, but there are obvious limitations to that approach. An intermittent arrhythmia may not conveniently recur during the few hours or days that a patient is in hospital. In summary it’s an area of practice where I see variable practice and uncertain outcomes. I’m pretty sure we can do better.
This month the NEJM has published the ASPIRED trial looking at early ambulatory ECG monitoring following unexplained syncope. It’s accompanied by a useful recent Clinical Practice review on syncope from Rose Anne Kenny, which gives us an opportunity to think a little more widely about how we investigate these patients. The abstract is below, but as always we recommend you read the full paper yourself and come to your own conclusions.
What did they actually do?
ASPIRED was a pragmatic, open-label randomised controlled trial conducted across 45 UK hospitals. We did not recruit at St Emlyn’s so I have no prior knowledge of this trial. The investigators recruited adults aged 16 or over whose syncope remained unexplained following their initial emergency department or acute medical assessment.
Patients were randomised within 72 hours to either standard care or a continuous 14-day ambulatory ECG patch, fitted as soon as possible after randomisation. The monitor recorded continuously and allowed patients to mark symptomatic events, with clinically important findings reported back to the treating team.
The primary outcome was the number of patient-reported syncopal episodes over the following year. Secondary outcomes included clinically significant and symptomatic arrhythmias, mortality, hospital admission, further investigations and therapeutic interventions. This is therefore a very pragmatic trial. It’s not really asking whether monitoring for longer finds more arrhythmias. We already have good reason to think that it will. The more interesting question is whether putting that technology into the clinical pathway soon after an ED presentation makes a meaningful difference to patients. That said, we will return to the choice of primary outcome later.
Tell me about the patients
They randomised 2234 patients, roughly equally between the two groups. The mean age was 58 years and just over half were male. The groups were well matched at baseline. Importantly, these weren’t simply low-risk patients being discharged from the ED. Around 39% were admitted following their initial presentation, although their syncope remained unexplained in that admission. The intervention was in addition to other investigations and management the treating clinicians considered appropriate.
Adherence wasn’t perfect. Of the 1122 patients allocated to monitoring, 930 produced an analysable recording, with an average wear time of just under 12 days. Even relatively simple diagnostic interventions don’t always happen exactly as planned in clinical practice.
What about the outcomes?
It’s an interesting primary outcome for a diagnostic rather than a therapeutic intervention. Rather than simply asking whether they found more arrhythmias, the investigators chose the number of patient-reported episodes of syncope at one year. This was influenced by patients and the funder, who prioritised recurrent symptoms rather than arrhythmia detection. I sort of get that, but at first look I was more interested in what is diagnosed, I suppose that in this approach that information would then be available to clinicians who may, or may not be able to influence the underlying cause and number of events. So there is a link, but it presume/hopes that the data found would be acted upon and that would make a difference in terms of number of events. Is that what we as clinicians are most interested in? Well, it’s yes and no.
On the one hand it’s arguably a good approach for patients, as finding an abnormality on a monitor is one thing but making patients better is rather more important. On the other, in a trial like this where diagnostics may affect care it might hide other important clinical outcomes. This has resulted in a little tension between the primary and secondary outcomes in this trial which will no doubt create a lot of debate, and probably a few more follow up studies. The wording in the discussion about choice of primary outcome is interesting and suggests there may have been discussions about this during the trial design stage (my interpretation, I don’t know this for sure).
If I had been designing this trial myself I would have been pushing for arrhythmia detection or changes to treatment as the primary outcome, but it’s not my study!
What did they find?
The primary outcome was negative. At one year, patients in the monitoring group reported a mean of 1.37 syncopal episodes compared with 1.58 with standard care. The incidence rate ratio was 0.89 (95% CI 0.68–1.18; P=0.42) so neither clinically nor statistically important. Recurrent syncope occurred in 27.9% and 28.6% respectively. So pretty conclusive that there is no difference, and as this is the primary outcome of the trial we should consider it negative. We could and technically probably should conclude that early monitoring didn’t work, and anything else is hypothesis generating (but read on……).
However, the secondary outcomes are really interesting. Clinically significant arrhythmias were detected in 22.0% of patients receiving early monitoring compared with 9.0% receiving standard care. They were also detected considerably earlier, at a median of 22 days compared with 54.5 days. So as a device it does what it says it will do. We will get more information regarding patient’s underlying conditions and by quite a large margin. Symptomatic arrhythmias were identified in 9.3% versus 5.9%. More patients subsequently received antiarrhythmic treatment, 10.8% versus 7.3%, and more received pacemakers, 6.8% versus 4.6%.
The monitor was therefore not just producing more ECG diagnoses, it led to fairly significant changes to treatment. There was also a really interesting mortality difference. At one year, 16 of 1095 patients in the monitoring group had died compared with 31 of 1085 receiving standard care, giving an odds ratio of 0.50 (95% CI 0.27–0.93).
Is this really a negative trial?
The primary outcome was negative and we shouldn’t get into the habit of searching through secondary outcomes until we find something we like. However, recurrent syncope is an interesting outcome for this particular intervention. Syncope is a symptom with multiple possible causes. An ECG monitor can only reasonably be expected to prevent another episode in the subset whose syncope is caused by an arrhythmia that is detected and can then be effectively treated. Even in the intervention group, clinically significant arrhythmias were found in 22%. Any effect on recurrent syncope across the entire population is therefore likely to be diluted.
There is also a potential ascertainment problem. Giving someone a device, asking them to mark symptoms and repeatedly asking about their episodes may make them more likely to notice and report events. None of this makes the primary outcome unimportant but it does mean that the simple description of ASPIRED as a negative trial doesn’t really capture everything that happened.
What about the mortality result?
It’s interesting, but again we should be cautious. A halving of mortality immediately gets your attention, particularly when there is a plausible mechanism. Early monitoring found more clinically significant arrhythmias. Those patients subsequently received more specific treatment, including pacemakers and antiarrhythmics. Some of those interventions could plausibly prevent death. However, mortality was a secondary outcome. The study was not primarily powered to demonstrate a mortality difference and there was no adjustment for multiple analyses across the secondary outcomes.
So the mortality finding is really interesting rather than definitive, and we certainly need to know more. I do wonder if there will be some follow up publications that might dive into this more (I reckon so).
Who should actually get a monitor?
In the NEJM, there is also a recent Clinical Practice review on Syncope by Kenny. In that they reiterate that the initial approach is still about basic clinical skills. Take a history, examine the patient, supine and standing blood pressure (though they are somewhat questionable, but we do them ) and the ECG. The history should include previous episodes and, where possible, an eyewitness account, or even video from bystanders or family which is something that I increasingly see (and is incredibly useful).
The review also gives us useful pointers towards patients in whom a cardiac cause is more likely. These include age over 60, known cardiac disease, male sex, a brief prodrome, palpitations or chest pain before the episode, syncope during exertion or while seated or supine, abnormal cardiac examination and a concerning family history.
If we integrate that with the APIRED results then the message is clearly not that everybody who faints should leave the ED wearing an ECG patch. If there are other causes identified, or if the patient is clearly in a very low risk group then it’s probably not indicated, but as the risk factors increase then so should our enthusiasm for monitoring. The Kenny review already recommends immediate ECG monitoring when arrhythmic syncope is suspected and suggests that the duration of longer-term monitoring should reflect the frequency of symptoms and the likelihood of detecting an arrhythmia. It’s not definitive, and does not tell us exactly what to do, but for me the accumulation of risk factors will make me more likely to recommend monitoring. ASPIRED suggests, that if we do that then we will pick up clinically significant diagnoses, that lead to treatment changes.
Do we need to admit these patients?
This is an area where I see a lot of practice variation Many clinicians will diagnostic uncertainty by admitting the patient to hospital, but for what? It only makes sense to do this if we are going to do something that we cannot do elsewhere, or when it is too difficult/dangerous to discharge. If you’ve already ruled out obvious causes, then admission is unlikely to be of much value. The Kenny review describes an alternative approach. Intermediate-risk patients with an unclear cause can be managed using structured ED observation with 6–12 hours of cardiac monitoring, which can reduce hospital admission without an apparent increase in serious outcomes. I think we do that fairly frequently, but not always in structured way.
ASPIRED potentially extends that idea further as ambukatory monitoring through a hospital at home scheme allows us to conduct monitoring over a much more clinically meaningful time period. And they clearly shiwed a significant diagnostic yield in ASPIRED. Clinically significant arrhythmias were identified in around one in five monitored patients, and approximately seven patients needed monitoring to detect one additional clinically significant arrhythmia. Most were detected relatively quickly, with more than 70% appearing within the first seven days.
What does this paper add?
There has been a gradual shift in syncope care away from the idea that everyone with unexplained syncope needs admission until we have somehow proved that they’re safe (which we cannot do btw). What this paper and the review tell us is that with a more structured approach, perhspas with the addition of ambulatory monitoring we can discharge patients from the ED earlier AND get a higher diagnostic yield. That’s surely a win-win.
Should we change practice?
I think this study should make us look seriously at how we manage unexplained syncope. It doesn’t demonstrate that 14-day monitoring prevents recurrent syncope, and it certainly doesn’t prove that it reduces mortality (though it might). It does provide strong randomised evidence that starting prolonged ECG monitoring early identifies substantially more clinically important arrhythmias, identifies them earlier and results in more specific treatment.
What we need now are ideas on how we would practically integrate that into NHS practice. Who selects the patients (EM or medicine)? Who reviews the recordings? How quickly are significant findings acted upon? What does the pathway cost? Does easy access to ambulatory monitoring reduce admission, or does it simply add another investigation to everything we already do? These questions need sorting before we can understand where the technology needs to sit within good clinical assessment, sensible risk stratification and a system capable of responding rapidly when something important is detected.
The bottom line
ASPIRED missed its primary outcome. Fourteen days of early ambulatory ECG monitoring did not reduce recurrent syncope at one year. If you’re an EBM purist (I am a bit) you might want to stop reading right here.
However, early monitoring more than doubled the detection of clinically significant arrhythmias, found them sooner and resulted in more targeted treatment. There was also a potentially important mortality signal that requires further studies to explore.
So……. I would argue that there is an argument that ASPIRED shows a benefit to patients, and if I had an unexplained collapse tomorrow I’d be quite keen to go home with one if these devices on. Indeed I have worn a different brand (Zio) for 2 weeks post cardiac valve surgery on two occasions, and they are not especially intrusive. There is a substudy (ref 9) on wearability from ASPIRED in the references below. I’d happily wear one again to see what’s going on……
vb
S
References
- Reed MJ, Goodacre S, Weir CJ, et al. Immediate Ambulatory Electrocardiographic Monitoring in Syncope. N Engl J Med. 2026.
- Kenny RA. Syncope. N Engl J Med. 2026;395:582–591. doi:10.1056/NEJMcp2517255.
- Brignole M, Moya A, de Lange FJ, et al. 2018 ESC Guidelines for the diagnosis and management of syncope. Eur Heart J. 2018;39:1883–1948.
- Sun BC, McCreath H, Liang LJ, et al. Randomized clinical trial of an emergency department observation syncope protocol versus routine inpatient admission. Ann Emerg Med. 2014;64:167–175. NEJMcp2517255.pdfPDF
- Reed MJ. Approach to syncope in the emergency department. Emerg Med J. 2019 Feb;36(2):108-116. doi: 10.1136/emermed-2018-207767. Epub 2018 Nov 23. PMID: 30470687.
- Reed MJ, Gibson L. The effect of syncope on brain natriuretic peptide. Emerg Med J. 2011 Dec;28(12):1066-7. doi: 10.1136/emj.2010.106955. Epub 2011 Feb 3. PMID: 21292790.
- Reed MJ, Newby DE, Coull AJ, Prescott RJ, Jacques KG, Gray AJ. The ROSE (risk stratification of syncope in the emergency department) study. J Am Coll Cardiol. 2010 Feb 23;55(8):713-21. doi: 10.1016/j.jacc.2009.09.049. PMID: 20170806.
- Reed MJ, Grubb NR, Lang CC, O’Brien R, Simpson K, Padarenga M, Grant A, Tuck S. Multi-centre randomised controlled trial of a smart phone-based event recorder alongside standard care versus standard care for patients presenting to the Emergency Department with palpitations and pre-syncope – the IPED (Investigation of Palpitations in the ED) study: study protocol for a randomised controlled trial. Trials. 2018 Dec 29;19(1):711. doi: 10.1186/s13063-018-3098-1. PMID: 30594256; PMCID: PMC6311046.
- Pearsons A, Hanson CL, Neubeck L, Blackstock C, Clarke E, Reed MJ. Usability and acceptability of ambulatory monitoring in undiagnosed syncope: insights from the ASPIRED-Q qualitative study. BMJ Open. 2025 Apr 8;15(4):e095927. doi: 10.1136/bmjopen-2024-095927. PMID: 40204326; PMCID: PMC11979494.
