Smartphone ECG monitoring found a cardiac cause in a small but clinically important minority of ED palpitations
Transient palpitations are a common emergency department complaint, yet they rarely announce themselves while a clinician is holding a standard 12-lead ECG. Symptoms may disappear before any rhythm strip is obtained, leaving clinicians to choose between reassurance, downstream testing, and prolonged uncertainty. This five-year service evaluation from Edinburgh examined a smartphone-based ambulatory ECG pathway built into routine care after the IPED trial, and it shows how a simple event-triggered recorder can change the diagnostic trajectory for selected patients with intermittent symptoms.
Across 1750 patients referred over five years, the cohort was relatively young, with a mean age of 44.3 years, and women accounted for 58.2%. A total of 1176 patients underwent monitoring using either the single-lead Kardia Mobile device or a smartwatch with ECG capability. The key clinical truth is blunt: most palpitations were not due to a cardiac arrhythmia, but a meaningful minority were. Among those monitored, a cardiac arrhythmia was directly captured in 67 patients, or 5.7% of the monitored group.
Direct rhythm capture changed the diagnostic pathway
The rhythm diagnoses were the kinds that matter at the bedside because they explain symptoms, influence follow-up, and often trigger treatment. Atrial fibrillation was identified in 39 patients, supraventricular tachycardia in 13, atrial flutter in 9, atrial tachycardia in 5, and ventricular tachycardia in 1. Those recordings are valuable precisely because they were obtained during the symptomatic window, when the arrhythmia was present and the symptom label could be matched to a physiologic event. In ambulatory arrhythmia care, that temporal link is often the decisive piece of evidence.
One sentence captures the service’s practical value: the pathway found the arrhythmia while symptoms were still happening.
By one year, 84 patients who underwent monitoring had received a cardiac diagnosis, corresponding to 7.1% of those monitored. Of those diagnoses, 79.8% were established directly through clinic monitoring, which speaks to the efficiency of an event-driven approach in a population with intermittent symptoms. The remaining diagnoses presumably emerged through other clinical workup, but the core point is that the smartphone pathway was not merely a convenience; it was the main diagnostic engine for most of the patients who ultimately proved to have a cardiac explanation.
Most patients did not have a cardiac cause, and that shaped management
Within the full cohort of 1750 patients, 114 individuals, or 6.5%, had a cardiac cause identified at one year. Far more often, the symptoms were ultimately classified as non-cardiac: 1472 patients, or 84.2%, received that explanation. In practice, that matters because palpitations are anxiety-provoking, and the absence of a cardiac diagnosis can still be a useful outcome when it is supported by symptom-rhythm correlation and follow-up rather than vague reassurance.
The service also translated diagnosis into action. Medication changes were initiated in 106 patients, 6.1% of the total cohort. The article does not provide a breakdown of the specific medication adjustments, so the safest interpretation is limited to the fact that rhythm findings or associated clinical review led to treatment modification in a non-trivial subset. For emergency and acute medicine teams, that is the bridge between monitoring and management: a diagnosis that changes neither therapy nor risk stratification has limited value, but one that prompts medication change has immediate utility.
A second blunt clinical reality follows from the data: nearly nine out of ten patients were discharged back to primary care.
Who is most likely to have a cardiac diagnosis
Predicting which palpitations deserve closer cardiac workup is where service design becomes more efficient. Multivariable analysis identified three independent predictors of a one-year cardiac diagnosis. Increasing age carried an adjusted odds ratio of 1.04 per one-year increase, prior atrial fibrillation/flutter carried an adjusted odds ratio of 5.57, and a presenting symptom described as irregular beating carried an adjusted odds ratio of 2.65. Those variables are clinically plausible because arrhythmia risk rises with age and with established atrial disease, while the patient’s own symptom description may reflect irregularly irregular rhythm rather than a fast but regular sensation.
These predictors do not replace monitoring, but they can help triage access. A patient with a history of atrial arrhythmia and a description of irregular beats enters the pathway with a substantially higher pretest probability than a younger patient describing isolated skipped beats or a fleeting sensation of pounding. In a service built around event capture, pretest probability matters because it determines who is most likely to benefit from rapid monitoring and who may be better served by a different diagnostic route.
Why the pathway makes physiological sense
Intermittent arrhythmias are poorly matched to single-point testing. A smartphone-based event recorder works because it is patient-activated, portable, and present when the symptom happens, which is exactly when the heart rhythm can be converted from a subjective complaint into a documented signal. That is especially useful for paroxysmal supraventricular tachycardias and episodic atrial fibrillation, where the rhythm can normalize before the patient reaches the hospital. The workflow also fits emergency care, where throughput, bed pressure, and outpatient follow-up all compete for attention.
The one ventricular tachycardia diagnosis in this cohort is small in absolute terms, but it is the sort of finding that justifies the safety logic of the pathway. Even rare high-risk rhythms can be identified when the right patient is monitored at the right time. At the same time, the high rate of non-cardiac diagnoses suggests that the pathway can reduce diagnostic drift and avoid unnecessary escalation for many patients whose symptoms are real but not arrhythmic.
Service evaluation strengths and the limits of a preprint
This was a single-centre service evaluation, not a randomized trial, so the findings speak to implementation rather than comparative efficacy. The data also come from a preprint that has not undergone peer review, which leaves methodological details and external validity open to later scrutiny. Even so, the cohort is large, the follow-up window is meaningful, and the service operated over five years rather than during a brief pilot period. That combination gives the evaluation real weight as a description of what happened when the pathway was integrated into routine NHS care.
The study also leaves some practical questions unanswered. The report does not describe the full downstream workup for every patient, the exact nature of the medication changes, or whether the pathway changed health resource use beyond discharge disposition. It also does not establish whether the same diagnostic yield would hold in other emergency departments with different referral patterns, patient mixes, or access to ambulatory rhythm services. What it does show is straightforward: smartphone ambulatory ECG can capture symptomatic arrhythmias, identify a cardiac cause in a minority of palpitations, and support safe discharge for most patients whose symptoms prove non-cardiac.