How many patients are in the emergency room today? How long have they been waiting? Why do they consult? All this information — and much more — is compiled daily in Quebec hospitals. What if they were used to optimize the operation of emergencies? This is the goal set by a management researcher and a family doctor in Rouyn-Noranda.
“Since I became head of the emergency room, I only spend about ten hours each month doing schedules,” says Doctor Justin Langevin, sighing from his side of the screen. He must also combine the work of administering his service with his obligations as a full-time doctor. “These are excess tasks that I have to do in the evening or on the weekend. And it’s not my expertise: I didn’t have management training! »
It was during a chance meeting with professor of management sciences at the University of Quebec in Abitibi-Témiscamingue Chahid Ahabchane that an idea was born. Why not entrust the task of creating schedules to artificial intelligence? The software could take into account employee availability and preferences to generate the ideal schedule — and relieve the Dr Langevin. “This is my main management task. If I free up that time, I will be able to concentrate on something else,” emphasizes the doctor.
Better manage your time
Artificial intelligence to generate… Schedules? Isn’t this too simple a task for such a specialized tool? “Emergency schedules are very complex,” replies the Dr Langevin. Some doctors work full time, others part time. Sometimes they have related tasks, so we cannot make them work evenings or nights. We have to balance weekday and weekend work. » Not to mention the difference in performance between newly arrived doctors, who will see fewer patients during a shift, and their more experienced colleagues.
But the ambition of the two collaborators does not stop there. In addition to producing schedules based on staff availability, why not also add an indicator: emergency room traffic. “We said to ourselves that we could take advantage of historical emergency data to create tailor-made schedules,” explains Mr. Ahabchane. We see, for example, that the emergency room is always busiest on Monday and Tuesday mornings. » The tool could take this information into account and assign more experienced doctors to this shift to minimize wait times for patients.
Quebec has a gold mine of data: the Med-Urge software, implemented since the early 2000s in emergencies across the province. Some basic information is noted about each patient, including the reason for the consultation, in addition to the arrival and departure time, the time spent in the emergency room and the priority for triage. “We have a lot of data, but it is under-exploited,” says Mr. Ahabchane, who sees immense potential. This information was used to feed a machine learning tool.
Result: since November 2023, the emergency room schedules at the Rouyn-Noranda hospital have been designed by artificial intelligence — under the close supervision of the Dr Langevin. “The algorithm has a different style from mine,” he notes with a smile. An example: “I usually give one or two shifts and then two days off. Instead, the algorithm makes people work for four or five days, but gives them a longer rest period. » A survey was carried out among emergency staff to measure their satisfaction with this new way of working and optimize the tool accordingly. Other improvements could also be made. “I would like people to be able to enter their preferences, such as the number of days in a row they want to work, and for the schedule created to take this into account,” says the doctor.
Support doctors
Their project may only be in its infancy, but the two colleagues are already thinking about what comes next. “If we manage to develop a standardized program, we could export that elsewhere in Quebec,” emphasizes Justin Langevin. Before achieving this, confidentiality issues will need to be resolved. Fortunately, very little information regarding patients’ health status is noted in Med-Urge. “Patient data is in the files in paper format,” explains the head of the emergency room.
One thing is clear: doctors need all the support they can get. “Emergency medicine is notoriously difficult, and atypical hours contribute to it,” recalls Dr.r Langevin. If we can optimize this aspect to improve the quality of life of doctors, we may be able to increase retention. »
This content was produced by the Special Publications team at Duty, relating to marketing. The writing of the Duty did not take part.