9 steps for successfully using CDS to decrease imaging utilization and join the value era

Most healthcare sites have some form of clinical decision support in place to guide imaging order entries. But many hospitals have struggled to progress their systems since the 2014 Protecting Access to Medicare Act required physicians to consult CDS tools before ordering advanced studies.

Noted imaging guru Christopher Roth, MD, MMCI, vice chair of Radiology for Clinical Informatics and IT at Duke University Medical Center, on Monday shared his own roadmap for using order entry CDS to decrease imaging utilization, fit into value-based care plans and take on risk.

“CDS is not perfect but it is the best thing we have,” Roth, who also volunteers on the American College of Radiology’s Informatics Commission, explained during the virtual session. “You absolutely can use it successfully at your organization.”

Below are takeaways from his presentation.

1. Get the right people together: Roth first found allies in electronic health record experts and revenue cycle managers, among many others. The more he preached about this project, the more support he gained from others, including orthopedic surgeons and Duke Health’s executive vice president.

2. Learn the process front to back: Understanding how decision support guidelines are created is key, and the ACR and Society of Nuclear Medicine and Molecular Imaging are crucial resources. The physician groups approving policies at insurance companies are quite reasonable and knowledgeable, Roth added; it’s the implementation of these rules where the trouble starts. Front-line, non-clinical representatives are required to follow rules to the tee and often don't have all the information they need.

“You really are turning a cruise liner to make change,” he added. “Payers are inefficient just like hospitals are.”

3. Decide goals (and non-goals): Like any big enterprise imaging project you must decide what you’re going to do and what you’re not going to do. Roth and his team prioritized complying with PAMA and reducing the costs and nuisances of preauthorization, among other goals.

4. Plan to educate: Budget time to educate, he urged. “It takes a lot of time to stand up and be successful in doing this,” Roth added.

5. Understand and build analytics: Making the case for why this is important may be difficult, but “you’re going to find interest here,” he said. Gathering red rate data (guideline-discordant ordering) and total orders per provider plotted against days or clinical encounters are two necessary pieces of information.

6. Build accountability: All Duke physicians have a red rate on their balance scorecard along with several other data points depending on the specialist.

7. Gather feedback: As the program gains steam, feedback will roll in and there’s no turnkey solution to manage this deluge. Roth’s group implemented a feedback survey o give users an opportunity to comment at the exact point they’re struggling.

8. Optimize aggressively: Use this feedback to tailor reasons for exams for providers, residents and staff. For example, if you have an urgent stroke code order for head CT, you want one reason for the exam—“just make it simple.”

9. Recognize risks: “CDS systems when rolled out sub-optimally can hurt people,” Roth explained. Doctors may pick reasons for exams that are quick and available, but aren’t totally accurate.

Around the web

Its testing slate will kick off with the diagnostic and interventional radiology subspecialty exams in nuclear imaging on Jan. 10. 

EZaccessMD offers teleconsults with board-certified physicians coupled with mobile diagnostics, hoping to help health plan holders limit costs on facility-based visits. 

A new survey of ortho experts offers key clues to help improve musculoskeletal radiology services and interdisciplinary collaboration. 

Psychology researchers have used machine learning to wring useful two-year dementia trajectory predictions from more than 500 potentially contributing risk factors.

Trimed Popup