- The Valenciennes hospital is experimenting with artificial intelligence software.
- The program predicts emergency patient flows to one week.
- The tool makes it possible to anticipate and thus avoid the congestion of the service.
Intelligent management. Since the end of 2019, the Valenciennes hospital center in the North has been experimenting with software that has revolutionized the organization of the emergency department, and even of the establishment as a whole. Thanks to artificial intelligence (AI), staff can anticipate patient flows up to a week with formidable reliability. Explanations.
One of the main missions of Rodolphe Bourret, the director general of the Valenciennes hospital, is to travel the world in order to hunt down innovations to be implemented in his establishment. It was in this context that he discovered Calyps, a Swiss company specializing in data processing for large retailers.
Data, lots of data
“To adapt their know-how to the health field, a sufficient combination of internal data was needed to enable foresight. This is the case in the emergency room, ”explains Rodolphe Bourret. And that's good because the problem is national: “These services are subject to random flows that are difficult to control because they are a function of many heterogeneous factors such as the weather, sporting or cultural events, epidemics, road traffic… “, Continues the director general.
But the problem is also part of the solution. “All of this data is available. They are aggregated with our internal emergency history data in software that predicts emergency patient flows at seven days with an 85% reliability rate, ”enthuses Rodolphe Bourret. Reliability rises to 95% at 48 hours from the prediction. The software information is sent directly to an application installed on the smartphones of emergency physicians. They can thus manage priorities in real time according to the planned activity.
Adjust the teams to avoid tension
According to hospital management, the benefits are many, particularly in terms of staff management. "This allows us to adjust the teams so that the service is not in tension as it often was before the experiment and it therefore also reduces the waiting time for patients", assures the director of the establishment . According to him, the algorithm also influences the overall reorganization of the hospital: availability of beds, stocks of drugs, places in operating theaters…
However, the software does not (yet) predict pathologies, even if some are starting to be integrated by AI: "We can already anticipate flows for cardio, gastro and even admissions of more patients 75 years old. On the other hand, there will always be this significant part of bobology that we cannot anticipate, ”deplores Rodolphe Bourret. For this, it relies more on the intelligence of people than on that of a program.
Health
World Cancer Day: A simple blood test to assess the radiosensitivity of patients
municipal
Municipales 2020 in Tourcoing: A list requests the introduction of “payment at source” of social benefits
- Artificial intelligence
- Emergencies
- Valenciennes
- Health