Drone defibrillator delivery could widen cardiac arrest coverage, Paris modelling study finds

More than 30,000 out-of-hospital cardiac arrests occur each year in the UK alone, according to the British Heart Foundation, and fewer than one in 10 people survive, with some survivors left with neurological problems. Every minute without CPR and defibrillation reduces the chance of survival by about 10%. "We have an objective as EMS organisations and EMS physicians to help cardiac arrest patients get defibrillated as early as possible," Heidet said.

To test whether drones could improve access to the devices, the team analysed the number and location of 28,349 out-of-hospital cardiac arrests recorded between 2011 and 2024 in administrative areas within the greater Paris region. Central Paris was excluded. The researchers cross-referenced these events with the locations of fixed automated external defibrillators and found that accessibility of the 1,893 devices varied considerably across the area. Dr Hillary Minka, an emergency physician at Lariboisiere hospital in Paris and first author of the work, said only 30% of all cases occurred within the target 500-metre network distance of the nearest recorded fixed AED.

The team calculated that an extra 910 fixed AEDs would be needed for 84.5% of the recorded cardiac arrests to fall within 500 metres of a device, and 1,712 extra devices would be required for full coverage. Using drones operating within a radius of 3,900 metres, coverage could be extended with far fewer devices. "As drone bases were introduced, the number of additional fixed AEDs required fell substantially," Minka said. With 100 drone bases and 26 additional fixed AEDs, more than 97% of cases would have been within 500 metres of a fixed device or roughly 4km of a drone base; with 200 drone bases and just four extra AEDs, more than 99% would have been covered.

The researchers also examined how many cases were close enough to an existing fixed AED for the device to be fetched and returned within five minutes. Only 30-40% of cases met that threshold using the ground transport network. Heidet said the figure was optimistic. "In France, only 8% of out-of-hospital cardiac arrest patients benefit from the application of a public AED before the arrival of the emergency services," he said, adding that a large proportion of AEDs are inaccessible because they are kept in enclosed private spaces that are not open at all times.

In a scenario involving 200 drone bases, 871 existing fixed AED sites and four additional fixed AED sites, Minka said drone delivery reached virtually all of the cardiac arrest cases covered by the model within five minutes. The models assumed drones would be housed at sites ranging from current AED locations to fire stations and mobile intensive care units, and did not examine real-world deployment of the technology.

Heidet said this is not the first time drones have been considered for delivering AEDs, and that the approach is already used in parts of Sweden, where it has saved lives. In existing systems, he said, drones are flown by trained pilots: once an emergency call is received and a cardiac arrest identified at the dispatch centre, the drone is activated by EMS dispatchers, takes off and follows an automated flight path, with the final approach and AED delivery handled by the pilot under supervision of the emergency medical service. Members of the public at the scene then use the device. Heidet added that the cost of drones needed to be built into any evaluation of the technology, and that there may be a balance to strike between the ideal model and a feasible economic one.

Editor's Summary The study models 28,349 out-of-hospital cardiac arrests in the greater Paris area against the locations of 1,893 fixed defibrillators, finding that only 30% of cases fell within 500 metres of one. Combining drone bases with existing and additional fixed devices would cover more than 99% of cases in the model, though the research has not been peer-reviewed and real-world deployment and costs remain to be assessed.