BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
CHICAGO -- While some radiology practices and groups still make solid profits, the writing is on the wall at RSNA 2013: What was once a cash cow profession is turning into a cost center for health systems.
As the fee-for-service era sunsets, both public and commercial payers are moving to outcomes-based models that reward healthcare providers that slash expenses. Even if those ideas take several years to take root in U.S. healthcare, reimbursements are decreasing in general, specifically in the Medicare population, which is radiology's bread-and-butter patient base.
Business intelligence is really the use of information to make smarter decisions. [Just] knowing is not enough.
Matthew Morgan, chief of imaging informatics, University of Utah
That, along with other "disruptive forces" such as HIPAA and meaningful use requirements that are changing the way providers are getting paid, puts radiologists in the hot seat, said Matthew Morgan, M.D., chief of imaging informatics for the University of Utah. Getting ahead of the game -- understanding where inefficiencies exist in a radiology operation and boosting quality while cutting costs -- has become a mandate for radiologists who want to survive the disruptions that are reshaping healthcare economics. His answer: Radiology business intelligence (BI).
Morgan, who is developing BI processes and software for the University of Utah, said radiologists can't afford to ignore this emerging field. They also can't afford to wait for their CIOs to implement some BI system that will lord over their department because it might be too little, too late.
"You can wait for something to be forced on physicians that maybe you don't think is advantageous," Morgan told SearchHealthIT after presenting on the topic at the Radiological Society of North America's (RSNA) 99th Scientific Assembly and Annual Meeting. "Or we can get people who are willing to jump into this. Maybe that's not the endgame, because I think eventually you'll need alignment [of BI in all departments] through enterprise IT, but I don't think that's in the next five years."
What to track
Morgan, during his presentation at RSNA 2013, laid out several potential examples of where some radiology practices have found success in using business intelligence. He broke them into three categories:
- Clinical: Helping prioritize which tests are more urgent (such as those assigned to the emergency department, intensive care unit or inpatient versus outpatient tests) and supporting out-of-band tasks such as detecting duplicate imaging studies before they're performed and alerting practitioners to gaffes such as not signing reports.
- Operational: More deeply understanding utilization of different radiology modalities to determine when you're busy and when you're not. Such "trends and volumes" intelligence can directly help bottom lines and make a more effective staffing model.
- IT: Setting up what Morgan calls "pre-emptive and proactive" alerting that something may be amiss in a radiology software system before it crashes to prevent a worse outcome, such as taking a practitioner or group offline.
"Business intelligence is really the use of information to make smarter decisions," Morgan said.
He added that some radiology groups might be tempted to rest once they build the systems and start dashboarding legitimate, normalized data. "[Just] knowing is not enough."
Assembling a BI team
There's no such thing as an out-of-the-box radiology BI software application that will gather all relevant data points, effectively dashboard them, and yield actionable data to support improved business decisions, Morgan said.
Radiology departments need to assemble teams to tackle BI, starting with a champion sponsor in the C-suite who sees the value of the project. Reporting to him or her is the manager of the program -- someone with clinical informatics experience. This person can oversee the efforts of the IT analysts and radiology department technology experts who execute the plan by tapping into the data systems on the network.
Those four parties together will identify and set the indicators that are relevant for tracking in the practice. They also will figure out how to normalize data across what will likely be several siloed data systems from multiple locations, and how to break that normalized data down into easy-to-digest dashboard views.
All that, and purchase the software to help accomplish it. But there's only so much software to automate this process, Morgan said. "I think you're fooling yourself if you think you don't need people at your institution that both appreciate the value and the need for this … and are willing to take the time to work on it so you're presenting the right data to the right people at the right time."