Big Data in Radiology: Use it Right

In recent years, there’s been an ongoing discussion about whether big data exists in radiology, and if it does, how useful can it be? The answer, so far, is a resounding, “It depends.”

According to Jon Hernandez, a health care software developer and database architect, yes, big data is present in radiology, but, he said at RSNA 2015, just how effective it is depends on who’s using it and how.

“There are beginner, intermediate, and expert users that already think about data outside their four walls and using it internally to help influence change and drive improvements around outcomes to reduce cost,” he said.

When used properly, this treasure-trove of information can also improve outcomes and increase value, he said. But, organizations must know how to capture it and look at it through a variety of lenses. In fact, in today’s health care environment where patient care is being provided through huge wellness campuses that include primary care, orthopedic, cardiac, and imaging services, as well as gyms, pools, and other healthy amenities, being big data-literate is paramount, he said.

In many cases, using big data successfully can be accomplished through a two-phase process, he said.[[{"type":"media","view_mode":"media_crop","fid":"44520","attributes":{"alt":"Jon Hernandez","class":"media-image media-image-right","id":"media_crop_5033867070146","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"4993","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 180px; width: 180px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"Jon Hernandez","typeof":"foaf:Image"}}]]

Phase 1: Before jumping into big data with both feet, he said, it’s important to determine what measurements you want to track and understand. How do you want the data to influence your workflow and decision-making process? Generate ideas about the tools that will help you capture and use this data internally.

Once you’ve decided what data you will capture, determine how best to visualize it. Map out a storyboard that will present information to patients in ways other than pie charts and bar graphs. Investigating new ways to present information could also spark other ideas for how big data can be used for other business intelligence solutions within your practice or department.

Even at this initial stage, gathering this data could give you significant information about your reimbursement that can be used in contract negotiations, Hernandez said. It could change how you get paid. In addition, bring together clinical and outcomes data from your hospitals and ambulatory clinics.

Next, search for and find an analytic partner who can help deliver these solutions for you, and work with them to better understand how any analysis can influence your daily operations.

Phase 2: Look for ways to expand and build upon your space in the radiology market to plug into an even broader set of analytical solutions. Be sure to educate everyone on your team so they are all working toward the same goal. Introduce the concept incrementally, he said, and bolster it with as many case studies as you can.

Be aware, your patients will have ever-growing access to any data you begin to track in upcoming years. In fact, the Fast Health Interoperability Resources (FHIR) standard, part of HL7, not only connects patients with their personal data more often, but it also notifies patients and allows them to accept any alerts that their health information has been transferred to a third party.

The most important thing when you begin to use big data is to ensure your organization is properly positioned for change, Hernandez said. Have a clear, clinical mission of how to collect disparate data and use it to implement the different business intelligence solutions your analytical partners might design.

“The key is to make sure your care providers know and understand any data they work with intimately before patients start to see what’s going on with it,” Hernandez said. “Make sure any data given to patients is highly relevant, strong, and paints the right picture.”