HosmartAI. Artificial Intelligence based appointment scheduling to increase patient and healthcare professional satisfaction - ITCL

The Scheduler logistics tool emerged as a solution to an existing problem in hospitals such as CHUL, where a change in a patient’s appointment entails a lot of work by medical staff.

It is the medical professionals who must be aware of what this change implies due to the treatment that the patient receives. In some cases it takes up to 20 minutes to make the change. These changes not only come from the hand of the patient but also from emergency admissions that require rapid action. Any modification not only affects the rest of the patient’s appointments but when the appointment becomes available, it can be used by another patient from this or another area of the hospital.

For this reason, it is critical to define the objective function, which can be as complex an equation as necessary.   For example, the solution will help prioritize patient treatment, while reducing the time between their appointments as well as downtime of human resources and machines. To address this, an optimization algorithm was chosen among several other possibilities because its search time for an optimal solution as well as other variables that determine the adequacy of the resulting solution can be parameterized.

In addition, this pilot wants to neutralize the effect of stress experienced by patients who, for medical reasons, need a continuous change of appointments. To solve this problem, a protocol for changing appointments has been set up via a Chatbot. The latter will propose to the patient to accept or refuse three appointments. In the event that none of them suits the patient, he/she will be contacted directly by the hospital staff to set the appointment that suits him/her best while respecting the guidelines of his/her treatment. The AI-based software tool aims to improve the flow of patients in large hospitals where some units, such as radiotherapy, have to carry out heavy work of constant appointment changes.

Establishing a new appointment for a patient can sometimes take up to 20 min, as this should in no way affect the outcome of his/her treatment or that of other patients.

An intelligent system that allows selecting the optimal appointments in minutes from endless possibilities and combinations while satisfying patients’ preferences regarding the time of their appointments is becoming more and more relevant.