EU cardiovascular health checks: who will read all the ECGs?
EU cardiovascular health checks: who will read all the ECGs?
November 13, 2026
On 28 September 2026, the European Commission proposed regular cardiovascular health checks for every adult in the EU, with ECG and heart-rhythm checks for people aged 65 and over. If member states follow it, millions of extra ECGs will need to be read, interpreted and signed off each year. AI-assisted ECG interpretation, with a clinician making the final call, is one practical way to keep up.
Key takeaways
The Commission's proposal for a Council Recommendation sets out health checks by age: at least once under 35, every five years from 35 to 64, and every three to five years from 65.
For people over 65, checks may include an electrocardiogram and pulse checks for heart rate and rhythm.
Cardiovascular disease causes 1.7 million deaths a year in the EU and costs around €282 billion, yet fewer than half of member states run health checks today.
The proposal supports AI-based tools for screening, as long as they are evidence-based and used with proper human oversight.
The bottleneck will not be recording ECGs. It will be reading them quickly and consistently across many sites and devices.
What does the EU proposal on cardiovascular health checks recommend?
The proposal turns the EU Safe Hearts Plan, adopted in December 2025, into concrete guidance on early detection. The Commission adopted it on 28 September 2026, at the start of the first EU Screening Week. Member states still need to adopt it in the Council, and each country decides how to put it in place.
The checks are organised by age group:
Proposed cardiovascular health checks by age group (European Commission, 28 September 2026)
Age group
How often
What is checked
Under 35
At least once
Blood pressure, blood sugar and cholesterol; more individual screening with a family history of early heart disease or sudden cardiac death
35 to 64
At least every 5 years, more often with risk factors
A wider set: lifestyle, kidney health, body mass index and waist, respiratory and mental health, medical and family history
65 and over
Every 3 to 5 years
Additional heart assessments, which may include an electrocardiogram and pulse checks for heart rate and rhythm
The proposal also suggests offering checks outside hospitals: in primary care, pharmacies, community centres and mobile units in places such as marketplaces and shopping centres.
Why ECG reading becomes the bottleneck
Around 99 million people in the EU were aged 65 or over on 1 January 2025, about 22% of the population, according to Eurostat. If every one of them had an ECG every three to five years, that would mean roughly 20 to 33 million additional ECGs a year. Real uptake will be lower, but even a fraction of that is a large new workload.
Recording an ECG takes minutes and is already routine in primary care and pharmacies. The harder part comes after the recording:
Interpretation: each 12-lead ECG needs measurements and a rhythm assessment.
Review and sign-off: a qualified clinician has to check the findings and take responsibility for the report.
Follow-up: an irregular heart rhythm or abnormal interval needs a clear referral path, sometimes to a longer Holter recording.
Consistency: checks will run in many settings, on ECG devices from many manufacturers, with reports that must look the same to the doctor who receives them.
Time from cardiologists and trained ECG readers is limited in most health systems. Screening at this scale only works if each ECG reaches a clinician already measured, organised and ready to review.
How AI ECG interpretation helps clinicians keep up
AI ECG interpretation does the first read, so the clinician starts from a measured, structured ECG rather than a blank trace. The clinician still decides. This is how Cardiolyse is built.
Cardiolyse is a CE-certified, device-agnostic AI platform for ECG interpretation and reporting. In a health-check programme, it supports each step after the recording:
AI first read. Every 12-lead resting ECG arrives with automated measurements and proposed findings. The platform covers 30+ arrhythmias and 300+ ECG parameters, including heart rate and rhythm.
Review in one place. In Cardiolyse ECG Workspace, clinicians view all 12 leads, measure waveforms and edit findings alongside the original recording.
Sign-off and reporting. The clinician approves an annotated, editable report, exported as PDF or JSON for the patient record.
Organised reading. A central worklist brings ECGs from many sites and devices into one queue, so reading teams and remote ECG reading services can share the load.
Follow-up when needed. When a check raises a rhythm question, Holter ECG analysis supports the longer recording that often comes next.
From raw signal to clinician-ready report takes under 30 seconds. The time saved goes to the part only a clinician can do: judging the result and deciding what happens next for the patient.
Health checks in primary care, pharmacies and mobile units
The Commission wants checks close to where people live, not only in hospitals. That means ECGs will be recorded on many different devices, by staff with different levels of ECG training, and often without a cardiologist on site.
Two things make that workable:
Device-agnostic analysis. Cardiolyse takes signals from any digital ECG source, whether a 12-lead resting ECG, Holter, patch or wearable, and returns the same structured measurements and report format. A pharmacy and a hospital clinic can use different devices and still send comparable results.
Integration through an API. The Cardiolyse ECG API connects ingestion, AI interpretation and reporting to existing systems, so results flow into electronic health records rather than staying on the device.
For programmes that already hold older ECGs on paper or as scans, ECG digitization turns them into analysable signals, so earlier recordings can be compared with new ones.
Why human oversight of AI matters in screening
The proposal is clear that digital and AI-based tools must be evidence-based, used with proper human oversight and in line with EU law. For ECG screening, that means AI proposes and the clinician decides.
“AI should give clinicians a better starting point, not take away their final judgement.”
Normunds Daudiss, CEO of Cardiolyse
Cardiolyse is built around that principle:
Clinician sign-off on every report. Findings stay editable until a clinician approves them.
Regulated medical software. Cardiolyse is CE Class IIa certified medical device software, developed under ISO 13485.
Data protection. ECG data is processed in line with GDPR.
Experience. More than a decade of clinical R&D and over 1 million ECGs analysed.
Frequently asked questions
What is the EU Safe Hearts Plan?
The EU Safe Hearts Plan is the European Commission's plan on cardiovascular health, adopted in December 2025. In September 2026 the Commission followed it with a proposal for a Council Recommendation on cardiovascular health checks.
Is the EU cardiovascular health check recommendation binding?
No. It is a proposal for a Council Recommendation, which member states must still adopt. Each country decides how to organise its own checks.
Who should get an ECG under the EU proposal?
The proposal says checks for people aged 65 and over, every three to five years, may include an electrocardiogram and pulse checks for heart rate and rhythm.
How often are heart health checks recommended?
At least once before age 35, at least every five years from 35 to 64, and every three to five years from 65 onwards.
Can AI read ECGs in a screening programme?
AI can do the first read of an ECG, with automated measurements and proposed findings. Under the proposal, AI tools should be used with proper human oversight, so a clinician reviews and signs off each report.
Does Cardiolyse work with any ECG device?
Yes. Cardiolyse is device-agnostic and analyses signals from 12-lead resting ECGs, Holter monitors, patches and wearables through one API and web dashboard.
Planning ECG reading for a health-check programme?
If you are designing cardiovascular health checks in primary care, pharmacies or a remote ECG reading service, we would be glad to show how AI-assisted ECG interpretation fits your workflow. Book a demo or read about ECG Workspace.