Document Type


Degree Name

Doctor of Philosophy (PhD)



Program Name/Specialization

Operations and Supply Chain Management


Lazaridis School of Business and Economics

First Advisor

Michael Pavlin

Advisor Role



This dissertation studies how patient access to specialized services in referral networks can be improved. The first study focused on optimizing and coordinating referral and scheduling decisions in a centralized referral network. I proposed a bi-level optimization model which enables the referrer to make optimal decisions for different scenarios based on available capacity in the network and operational competency levels of surgeons. First, I derived optimal scheduling policies for each surgeon in the network. Next, optimal referral decisions for the central referrer were derived for each capacity scenario. Finally, I studied how incorporating fairness in referral decisions can impact patient access to surgeons.

The second study applies deep reinforcement learning (DRL) algorithms in centralized referral networks that help referrers make optimal decisions during the patient referral process while considering different challenges such as distance of the patient from the specialist and wait time. First, I studied the potential impact of using these algorithms in a single centralized referral network. Next, I defined a general framework under which two adjacent centralized referral networks that are applying DRL algorithms can collaborate. Finally, I studied how governments can motivate networks to collaborate and what the impact would be of this collaboration on patient access to surgeons.

The third study focuses on the patient referral process in the Waterloo Cataract referral network. First, I analyzed the data and three different ways that are practiced by the network to refer patients to surgeons. Next, I simulated the whole network and studied how changing current referral policies or adding more surgeons to the network can impact patient access to surgeons.

Convocation Year