Abstract

Every year, a vast amount of money is spent across Canada for healthcare. In fact, Canada healthcare system has been spending more than numerous other countries, yet the outcome is not proportional to the efforts and budget that have been used. A big chunk of the pie of expense surplus comes from over-utilization of resources. Most notably, around 20 % to 50 % of this cost stems from the high-tech imaging modality, especially MRI scanners. This over-utilization in MRI procedures leads to higher costs for the system and patients, longer waiting time, and patients’ dissatisfaction. Researchers have developed methods to improve patient experience and the population’s health while reducing the per capita cost of healthcare. Motivated by the mentioned facts, the primary purpose of this thesis is to compare two-tiered and multi tiered approaches to MRI for the diagnostics of multiple sclerosis (MS) and prostate cancer (PC) diagnosis in a community-based practice in a large metropolitan area in Canada. The goal is to identify the best procedure protocol in terms of cost, efficiency and throughput. In particular, our contributions are two-fold. First, we compare different MRI procedures to MS diagnostics based on referring physician specialization (i.e. neurologists and non neurologists referrals). We further investigate the dependency between the test results and referral pattern via conducting a statistical hypothesis test. Our second contribution revolves around the comparison between two-tiered and multi-tiered MRI procedures for PC diagnostics in terms of cost and efficiency. Through a decision tree analysis, we demonstrate that two-tiered MS MRI approach and two-tiered MpMRI for PC turns out to be the most cost efficient methods with the lowest number of adverse events while improving patients experience through reducing the number of callbacks and venipuncture/contrast complications.