In France, one person is affected every ten minutes by sudden death syndrome. The heartbeat soars and the heart stops beating within an hour. To try to understand the phenomenon, the team of Xavier Jouven, cardiologist at the European Georges-Pompidou Hospital in Paris, and founder of the Sudden Death Expertise Center, has been working on the subject for more than 30 years, indicates BFM -TV which relays his research.
A database of 350,000 cases
A team of engineers and doctors are working on data from sudden death cases in and around Paris (inner suburbs). To build a reliable and solid basis, the experts collect information from the National Health Insurance Fund (CNAM). Medical histories, examinations, and diagnoses up to ten years before the patient’s death are thus collected. The team also works with all the data from the Samu and the intensive care units. In total, 350,000 cases are listed in this database approved by the CNIL.
A veritable mine of information which, once processed, can be used by artificial intelligence to try to find common denominators in these unexpected deaths. In this respect, mass treatment makes it possible to detect certain fairly rare risk factors. This is enough to get closer to a real prediction of the syndrome, which could be used to make a “personalized risk equation” for each patient, in the hope of preventing the occurrence of a case. AI would also aim to detect undetectable micro-signals, especially among the youngest.
A confirmation of the first results necessary
Currently, the AI used by the Sudden Death Center of Expertise is able to identify people with a greater than 90% chance of experiencing sudden death within the next 12 months. To validate more fully the validity of these initial results, the center will launch an observational study, with data from Health Insurance, in order to identify people at risk and confirm the veracity of the model. If successful, an interventional study will be launched, explains the professor to BFM.
“In the next step, we would like to do individualized prevention. That is to say that we will remove the person’s first risk factor, then his second, his third etc… But all this still has to be proven because, for the moment, we have shown association but causality has not yet been proven“, tempered Xavier Jouven in early January on Europe 1.