Centre for Excellence in Operations (CEO)

Forecasting Demand for Diagnostic Imaging Services

Detailed report
Final presentation

Over the last several years, the demand for and use of diagnostic imaging has been increasing at a rapid rate in the Capital Health Region. Some of this increase in demand can be attributed to a growing population and changing demographics; however, many of the other factors are less obvious and more difficult to measure. Our goal was to analyze these 'other' variables in order to predict their future effect on demand for CT, MRI and US in the Capital Health Region.

Our first stage consisted of both qualitative and quantitative data collection. The factors that we outlined as being the most significant to our project were:

  1. Changes in market share - Decreases in demand due to machines in other regions
  2. Supply-induced demand - Increases in demand due to increases in supply
  3. New technologies - Improvement in technology for CT, MRI and US machines
  4. New treatments - The development of new procedures and uses for CT, MRI and US machines
  5. Increasing affliction rates - If Albertans are seeing more injuries or disease per capita
  6. Changing physician behaviours - Changes in physician attitudes, standard of care, etc.

Due to time constraints and data availability, we were able to perform analysis on four of six factors. The effect of changes in market share was analysed by comparing the rate of Capital Health CT scans provided to residents of the Aspen Health Region before and after they acquired three CT machines. We found that this reduced the amount of Aspen Health Patients coming to Capital Health by close to 30%. A beta test to determine a relationship between change in waitlist times and change in demand failed to produce a clear conclusion, leading us to believe that the relationship is obscured. We calculated a comparison of the number of scans performed per patient over time, which indicated that the number of MRI scans per patient was decreasing while the number of CT scans per patient was increasing. These could be the result of new technology in that less repeat scans are required due to clearer images, but more abnormalities requiring a new scan are found due to higher quality. As well, an analysis of the use of different categories of scan as a percentage of total diagnoses of the corresponding type enabled us to determine an adoption pattern for Cardiac MRI. We found that it took approximately 5 years for it to reach is saturation point, and we think Cardiac CT and Breast MRI will have similar adoption patterns.

Through our qualitative and quantitative analysis, we were also able to rank each factor according to their perceived impact on growth of demand.