Business Analytics in Diagnostic Imaging

January 18, 2018 | Blog post by Brent Maranzan

A little while ago, I came across this short white paper written by a provider of a cloud-hosted diagnostic imaging software service. Platform aside, what caught my eye in particular was the “six measures of radiology operational performance” discussed in the paper.

Diagnostic Imaging (“DI”) and lab departments comprise roughly 20% of a typical Hospital’s total expenses and the operations of these departments can have a dramatic impact on patients’ quality of care, perception of care and length of stay. Overutilization of DI can be a significant drag on operational results. In Ontario, Hospitals funded through the Health Based Allocation Model essentially get paid for the average (expected) level of use of “investigative technologies” in their Emergency Department funding allocation. Hospitals that use higher than expected levels of DI will have higher costs, which will result in a reduced overall funding allocation. Analysis of DI operations should be a significant part of any Hospital’s performance management framework.

Unfortunately, these databases are often large and complex and they can be very difficult for Hospital Management to analyze. In the Northwest, Thunder Bay Regional Health Sciences Centre for example, performs over 260,000 medical imaging procedures in a single year.

Microsoft Excel (still the business analyst’s tool of choice) has a record limit of just over 1 million records and files of more than two or three hundred thousand records can be difficult to work with without advanced techniques. As a result, these databases are often incompletely summarized, poorly extrapolated from samples and anecdotes, or even ignored altogether.

These types of databases are perfect for Business Intelligence (“BI”) projects. BI systems take the “grunt-work” out of mining, transforming and presenting business information for analysis. Databases with millions of records are easy for systems such as Microsoft SQL Enterprise to handle.

We at the Northwest Health Alliance are currently in the midst of a regional BI project using Microsoft SQL Enterprise and the suite of Microsoft BI tools, such as Sharepoint Reporting Services and Power BI. The current project plan includes a DI database as one of the first available databases. According to the current project plan, by the end of the fiscal year (March 31, 2018) we will have a fully-functioning, regional, DI enterprise data warehouse producing automatic reporting to address three of the “six measures of radiology operational performance” discussed in this paper:

  • Measuring and optimizing turnaround times
  • Physician ordering analysis, and
  • Capacity management

Hopefully, this will contribute to efforts to improve care at Northwestern Ontario Hospitals and at the same time favourably bend the cost curve downward. In the meantime, we continue to assist Hospitals with specific projects using manual data extracts and more advanced MS Excel techniques such as PowerPivot and PowerView to report on measures such as turn-around times for radiology, orders by physician and other topics on demand.

If you would like to know more about this topic or how we currently assist Hospitals with ad-hoc reports don’t hesitate to contact us!



Contact us today and see how we can help you access and make sense of your healthcare organization’s data!