AED Drones Cut Cardiac Arrest Response to 5 Minutes

A geospatial simulation of 28,349 real out-of-hospital cardiac arrests in the Paris region found that drone-delivered automated external defibrillators (AEDs) could reach 95% of patients within five minutes, compared with just 30-40% for the region's existing network of fixed, wall-mounted AEDs. Dr. Hillary Minka, an emergency physician at Lariboisière Hospital, presented the findings on September 26 at the European Emergency Medicine Congress (EUSEM) in Paris.

Background

Out-of-hospital cardiac arrest (OHCA) happens when the heart suddenly stops beating outside a medical facility, and survival drops roughly 10% for every minute defibrillation is delayed. An AED (automated external defibrillator) is a portable device that analyzes heart rhythm and delivers a shock to restart a normal beat; getting one to a patient fast is the single biggest lever bystanders and emergency services have before paramedics arrive.

Most European cities, including Paris, rely on fixed AEDs bolted to walls in train stations, offices, and public buildings. The study's underlying problem: only 30% of the 28,349 cardiac arrests analyzed occurred within 500 meters of an existing fixed AED, and devices inside locked private buildings are often unreachable when a bystander needs one most. Dr. Matthieu Heidet, another EUSEM presenter, noted that in France only 8% of cardiac arrest patients currently receive a public AED before the ambulance arrives.

How the study modeled drone delivery

Minka's team, working from an abstract titled "Optimizing automated external defibrillator access using drones: a geospatial analysis in the Paris region," modeled seven departments of the Île-de-France region, excluding central Paris, using historical OHCA data collected between 2011 and 2024. The team compared three strategies for closing the coverage gap:

The hybrid model cut the average time to AED arrival by 3.76 minutes across the study area, with savings exceeding 10 minutes in some congested or peripheral neighborhoods where road access is slowest.

The study has a significant limitation the authors flagged themselves: it is a straight-line simulation, not a flight trial. No drones actually flew during the research, and the model does not account for airspace restrictions, weather holds, or the regulatory approvals a real BVLOS (Beyond Visual Line of Sight) drone network would need in France. The abstract, numbered OA066, has not been peer-reviewed and received no external funding.

Real-world AED drone programs already flying

The Paris findings arrive as several operational AED drone programs are already proving the concept outside a simulation. Everdrone has been flying AED deliveries for France's SAMU 76 emergency service in Normandy since December 2025. Sweden's national AED drone network, also run by Everdrone, has been operating since 2020 and remains the longest-running program of its kind. In the United States, Duke Health began AED drone flights in July 2026, and James City County, Virginia launched a 911-dispatched AED drone pilot on August 1, 2026 that has already beaten fire crews to cardiac arrest calls by more than five minutes in early tests.

Those live deployments matter because they answer exactly the question the Paris study can't: whether a drone can legally and reliably fly the route a simulation only draws on a map. For a broader look at how the underlying logistics work, see our drone delivery guide.

What This Means for Drone Pilots

For pilots and operators working in or entering the medical logistics space, this study is a data point that regulators and hospital systems will likely cite as they weigh new AED drone approvals — but it is not itself evidence a network works. Operators pitching AED drone programs to municipalities or health systems should expect officials to ask for exactly what Minka's team didn't have: flight-tested response times, not modeled ones. The gap between 95% coverage on a map and 95% coverage in the air is where BVLOS waivers, weather reliability, and dock placement decisions will actually be made. Programs like Everdrone's in Normandy and Sweden, and the US pilots at Duke Health and James City County, are the ones generating the flight data that will determine whether findings like this one translate into funded, scaled networks.

Sources: EUSEM Press Release | DroneXL