What is Artificial Intelligence Doing on your Emergency Call in France?
Every day, the heroes of French emergency departments help thousands of people on the worst day of their lives, but by the end of 2019, more than a million emergency calls had gone unanswered leaving patients in despair.
Both fighting valiantly to catch up to the ever-increasing volume of emergency calls, French Fire and Emergency Services, SDIS and SAMU were troubled with too few resources, lack of trained personnel and outdated software tools. It proved to be an unfair battle.
Blindly triaging a patient with limited time, under high pressure and without ever seeing or feeling the patient, is an extremely daunting challenge, even for the highly educated physician, and that is exactly what the daily life of a call-taker consists of. France is reaching a yearly increase of 20% in emergency calls where call-takers are being pushed to their limits, with too little continuous training and diminishing resources for quality assurance, leading to a suboptimal working environment. An emergency call-taker is one of the most important roles not only in the healthcare sector but in our society, and as research has shown from multiple different sectors, high levels of stress and pressure will inevitably compromise decision-making abilities, which is unacceptable when it comes to critical decision making, like in the case of triaging an overdose or a cardiac arrest over the phone.
France has seen several tragic and touching incidents, where a more technologically supportive environment for call-takers might have made a life-changing difference. In 2018, the Naomi Musenga scandal in Strasbourg exposed how exhausted call-takers can be. A young woman received help too late after having to call two different medical emergency services, getting less than optimal feedback from an overworked call-taker. In the recent Mulhouse case, all of France grieved for a 60-year-old woman who called the emergency medical service suffering a cardiac arrest but sadly, she didn’t either receive the assistance she needed. She was later found deceased from her neighbour 10-days after calling. The responsibility was pointed at the call-takers working 12-hour shifts under demanding conditions, using tools that weren’t worthy of the unsung heroes of the emergency sector. In Nancy, France a patient suffering cardiac-arrest called the emergency line 3 times, but every time it was concluded that no ambulance was needed. It took the patient a final fourth call to receive ambulance assistance, but it was too late.
These stories are no longer one-time cases and the burndown of call-takers and the absence of resources to train and enable them is affecting the prehospital sector. Thousands have been protesting during the last couple of months against the lack of funding and improper working conditions, but not much seemed to change to tip the reality in their favour
Noticing importance for a change, a group of SDIS and SAMU visionaries began looking for new solutions that would enable call-takers during the hardest calls that they would face. Working with the French company Systel, the group envisioned a product that would help streamline the process of blindly triaging patients over the phone, without ever getting in the way of the call-takers, but offering them a safety net during the most troublesome circumstance.
Far to the north in Scandinavia, they found the Copenhagen-based startup Corti. Corti, named after a small organ in the ear that converts audio waves to brain impulses, was introducing a new paradigm in patient triaging. The team at Corti had built a patent-pending technology, in many respects similar to Apple’s SIRI, capable of understanding patient interviews in real-time, finding subtle signals in the audio that could help the nurse or doctor, to triage and diagnose the patient correctly. Corti began its journey looking for ways to provide emergency call-takers in Denmark and the United States with their real-time AI tool in the war against fatal mistakes and time-constraints, and they had made cardiac arrests their expertise. Shortly after Corti’s founding in 2017, one of the biggest AI research projects on emergency calls ever initiated began, to build what could be described as an intelligent co-pilot for call-takers, working with Copenhagen EMS and Seattle Medic One, two of the premier emergency medicine departments.
“Every emergency call listened by the AI is a step further towards an even more intelligent assistant diagnosing and instructing the proper response faster and more accurately, significantly reducing the non-detected conditions and non-assisted patients”. Lars Maaloe, co-founder and CTO of Corti.
Believing in the technology potential, the French department of Haute Savoie, the SAMU-SDIS 74, took part in a project between Corti and EENA, demonstrating how AI could work alongside call-takers to improve their responses. Following the project results being published, research with France has begun and the first deployments have been scheduled. Alongside the first pilot sites, every innovative department working with SAMU or SDIS can leap into the future and join the project. Corti maintains scalability for additional partnerships across France to keep training Corti’s AI models and to have them learn from the best, developing additional detections and applications.
The new partnership between the SAMU/SDIS of France and Corti is seeing positive results, motivating all parties to continue to drive forward and pave the way for an exceptionally promising future. Having unlimited assistance to navigate a call-taker through the phone call will not only establish a new standard where our everyday heroes receive the most superb working environment in their profession but a new standard in France where superior emergency assistance will be available for everyone.
