Key findings

Central message: Early 14-day ambulatory ECG monitoring substantially increased the detection of arrhythmias and led to more cardiac interventions, but did not significantly reduce recurrent syncope. A potentially important mortality difference was observed, although it cannot be considered definitive.

Background 

Syncope accounts for a substantial number of emergency department presentations. Approximately 50% of ED patients with syncope leave the initial assessment without an established cause.  Conventional short-duration monitoring has limited diagnostic yield, as arrhythmias may occur intermittently and be missed during short monitoring periods. Early monitoring may be particularly valuable because arrhythmias are more likely to be detected soon after the index event.

ASPIRED tested whether this plausible diagnostic and treatment pathway actually translates into better clinical outcomes.

Study design

Patient selection

Patients were eligible if they had experienced a transient loss of consciousness with loss of postural tone followed by immediate, complete spontaneous recovery, and no explanation had been established following initial assessment.

Importantly, patients who were admitted to hospital could still be recruited. This was not simply a study of low-risk patients discharged from the ED.

Approximately 39% of participants were admitted during their index presentation, and many admissions were classified as high-risk syncope.

This makes the population more clinically relevant than a study limited to otherwise well patients discharged home.

Intervention and comparator

Randomisation occurred within 72 hours of presentation ie., not monitpring commenced immediately in the ED.

The two groups were reasonably well matched and the group comprised a broad syncope population. About 10% had atrial fibrillation or coronary heart disease. A small proportion had heart disease. About 45% of patients had previously experienced syncope. This could increase the likelihood of detecting recurrent symptoms, but also may comprise a group of patients whose episodes may have a non-arrhythmic cause.

Critical appraisal: The population is representative of many ED syncope presentations, but this breadth potentially dilutes any benefit from monitoring in a narrower high-risk arrhythmic subgroup.

Primary outcome  

Did monitoring prevent recurrent syncope?

Mean patient-reported episodes over one year

Mean episodes per patient; standard deviations were 5.10 and 8.56, respectively.

 

The primary outcome did not reach statistical significance. The trial therefore does not establish that monitoring reduces recurrent syncope. However it also does not prove that there is no benefit.

Why might the primary outcome have been negative?

1. Most syncope is not caused by a treatable arrhythmia. Only 22% of the intervention group had a clinically significant arrhythmia identified.

2. Arrhythmias may be incidental, especially if not associated with a sy

3. The study may have included patients at relatively low risk of recurrent arrhythmic syncope.

4. Recurrent syncope is a difficult outcome to measure.

The mean number of episodes was highly variable. The standard deviations were much larger than the means, suggesting a skewed distribution with some patients experiencing many events.

There was little apparent difference in whether patients experienced another syncopal event.

Arrhythmia detection 

Percentage of patients with an identified clinically significant arrhythmia at one year.

Absolute increase

13.0%

Odds ratio

2.95

Number needed to monitor

~8

The odds ratio was 2.95 (95% CI 2.28–3.81). The simple reciprocal of the 13-percentage-point absolute difference gives approximately eight patients who need the monitoring strategy to identify one additional clinically significant arrhythmia compared with standard care.

Which arrhythmias were detected?

These figures require careful interpretation.

The largest difference was in nonsustained ventricular tachycardia (NSVT), (defined as at least four beats at a rate of at least 120 beats/minute, lasting less than 30 seconds) detected in 10% versus 1.2%. However we need to keep in mind that an episode of NSVT may not necessarily have caused a syncopal event.

Symptomatic arrhythmias

A symptomatic arrhythmia was detected in:

• 9.3% of the monitoring group.

• 5.9% of the standard-care group.

Absolute increase: 3.4 percentage points.

This is a much smaller difference than the 13-percentage-point increase in the broader category of clinically significant arrhythmias.

A simple number needed to monitor to identify one additional symptomatic arrhythmia is approximately 30.

We need to keep in mind that we need to establish a link between that arrhythmia and the syncopal episode. 

How quickly were arrhythmias detected?

The median time to detection of a clinically significant arrhythmia was:

Did monitoring change patient management?

Yes. The increased detection of arrhythmias translated into more interventions.

Pacemaker implantation occurred in 74 patients in the monitoring group compared with 50 in standard care.

Would those devices have prevented important outcomes? It is hard to tell.

Are we treating more patients due to incidental findings?

Mortality 

All-cause mortality at one year

 

Odds ratio = 0.50 95% CI (0.27–0.93)

Absolute difference −1.4%

There were 16 deaths among 1,095 patients in the monitoring group, compared with 31 deaths among 1,085 receiving standard care.

The unadjusted absolute difference corresponds to approximately one fewer death per 71 patients assigned to early monitoring.

We need to beware as the trial was not primarily powered to assess mortality (a secondary outcome), and the authors explicitly state that confidence intervals for secondary outcomes were not adjusted for multiple comparisons.

There were relatively few deaths. Forty-seven deaths occurred across both groups, making the estimate vulnerable to chance variation.

 Conclusion

Strengths

It includes a large number of patients across 45 hospitals and uses a clinically meaningful comparator.

The investigators also assessed outcomes extending beyond diagnosis, including syncope recurrence, treatment, admissions and mortality.

Limitations

1. Broad inclusion may dilute benefit. 

2. The primary outcome was patient-reported. 

3. There was missing data and incomplete adherence. 

4. Heterogeneous standard care. 

5. The high rate of NSVT detection demonstrates an increased dianostic yield. However we cannot determine the optimal period of monitoring. 

ASPIRED demonstrates that early 14-day ambulatory ECG monitoring in patients with unexplained syncope substantially improves the identification of clinically significant arrhythmias and increases the use of rhythm-directed interventions.

However, it did not significantly reduce recurrent syncope at one year.

The observed reduction in mortality if confirmed, is practice changing, however at the moment it is a secondary finding.

Monitoring detects more arrhythmias. Do patients benefit from this early detection?

For current practice, the trial strengthens the rationale for timely ambulatory monitoring in appropriately selected patients with unexplained syncope, while stopping short of establishing universal 14-day monitoring as a strategy that improves patient outcomes.