A Concordia-led research team has developed a planning tool that could help hospitals book their operating rooms more efficiently, shorten wait times and better cope with last‑minute emergencies.
The researchers developed their model using artificial intelligence tools to plan which operating rooms to open on each day, when each surgery should start and which cases may need to be delayed, all in a single, integrated framework. Their model uses far fewer variables than a widely used previous approach, making it faster and more practical for real hospital conditions, especially when dealing with dozens or even hundreds of operations in a week.
The model can re-plan the schedule day by day to insert true emergency surgeries or a patient whose condition has suddenly become more urgent, while keeping disruptions and postponements for other patients as low as possible.
In tests using both simulated data and real schedules from a hospital in Naples, Italy, the system absorbed same-day emergency arrivals with only modest changes from the original plan, using tools like limited overtime, opening extra rooms or carefully deferring a small number of elective cases.
The researchers, led by Hossein Hashemi Doulabi, an associate professor in the Department of Mechanical, Industrial and Aerospace Engineering, believe the tool could contribute to better cost control and supports hospital teams as they work to shorten surgical wait lists. They also say it could reduce day-of-surgery cancellations and respond more smoothly when emergency cases place extra demands on the system.
The study was published in the International Journal of Production Research.
Concordia PhD candidate Mahdi Dolatkhah is the paper's lead author. Walter Rei at Université du Québec à Montréal and Michel Gendreau at Polytechnique Montréal also contributed.
Read the cited paper: "A reinforcement-learning-based column generation algorithm for integrated operating room planning and scheduling"
