24-Hour Holter ECG: How to Prepare, How to Wear It and Whether a Smartwatch Can Replace It

A Holter is most often ordered for palpitations, fainting and near-fainting, dizziness of unclear cause, chest pain or suspected rhythm disturbance, and also to assess the effectiveness of treatment for arrhythmias. The recorder records rhythm throughout the day, and software analyses the trace, indicating among other things the number of ventricular and supraventricular extra beats, episodes of tachycardia, pauses, bradycardia and atrial fibrillation. The doctor combines this with the times at which the patient noted symptoms.

An important limitation: a Holter is a 'time-window' test. If there were no symptoms and arrhythmias during that day, the result may be normal despite a genuine problem that occurs once a week or once a month. Hence the question of extending monitoring, which we address below. A normal Holter in a person with symptoms therefore does not end the inquiry if the clinical situation is worrying.

Preparation is simple, but it affects the quality of the recording. The electrodes are stuck to the chest, and good contact with the skin is a condition for a clean signal. Interference (artefacts) from poorly adhering electrodes, sweat or movement can hamper analysis and imitate arrhythmias.

A standard recorder is not waterproof, so you usually must not bathe or shower throughout the recording — this is the most common instruction, but waterproof devices or patches are also available, so it is worth asking about the specific model. Strong magnetic and electromagnetic fields directly next to the recorder should be avoided, as well as electric blankets during sleep, if the centre advises it.

Activity during the recording should correspond to an ordinary day. There is no need, and no point, to 'test the heart' with extreme exertion, but neither should you 'save' the recording by lying in bed, because then nothing will be captured. It is worth keeping the recorder in a position that does not pull on the leads; at night it can be placed beside you or under the pillow so that it does not tension the cables. If an electrode comes off or a lead disconnects, tell the unit — a briefly interrupted recording can often be completed, but a long gap limits the value of the test.

Detection depends on how often arrhythmias occur and how long the recording lasts. Two studies illustrate this well, although each from a different perspective.

Cautious conclusions follow from both studies. First, in the retrospective study of emergency department patients 24 hours gave a similar diagnostic yield to 48 hours, so extending a standard recording by a day does not always increase detection. Second, for episodes that occur less often than once a day a longer recording (for example a 14-day patch) detects more events — which is obvious, because it observes for longer. It does not mean the patch is always better: it has one lead, and its clinical value compared with a classic Holter depends on the purpose of the test. The choice of method (24 h, 48 h, patch, event recorder, implantable loop recorder) belongs to the doctor and depends on how often the symptoms occur.

Smartwatches with an ECG function (for example the Apple Watch) record a short single-lead trace, usually about half a minute, when the user touches the crown or sensor. Models with an optical sensor (photoplethysmography) can also alert to an irregular pulse. For atrial fibrillation their accuracy has been well studied, but the question 'will they replace a Holter' must be answered separately.

These results are good, but they answer a different question from the one a Holter poses. The meta-analyses assess whether a short trace from a watch correctly identifies atrial fibrillation when the user takes it, and compare it with a 12-lead ECG. They do not show that a watch replaces continuous round-the-clock monitoring, which (1) records rhythm all the time, including during sleep and asymptomatic episodes, (2) allows arrhythmia burden to be counted (number of extra beats, length of pauses), and (3) links the recording to a symptom diary. In the abstracts of the studies cited we found no data on detecting arrhythmias other than atrial fibrillation. A 2024 narrative review also notes that a limitation of consumer devices is high false-positive rates in some populations.

When the recording ends, the recorder and diary are returned to the unit. The trace is analysed by computer and then verified by a doctor; the result is often available after a few days. The report usually contains the mean, minimum and maximum heart rate, the number and type of extra beats, episodes of tachycardia, pauses and the correlation of symptoms with rhythm. Frequent but isolated extra beats in people without heart disease are usually harmless, and the significance of the result depends on context — we write about that in the article on extra beats.

A good Holter is one in which the ECG trace and the symptom diary tell the same story. A few simple habits — prepared skin, a normal day, noting the times of complaints — raise the chance that the test will answer the question for which it was ordered. Watches and patches are a valuable supplement and, for rare symptoms, even a better choice than a single day, but the decision on which device and for how long should come from the symptoms and a conversation with the doctor, not from what we have on our wrist.