With the Medicare Access and CHIP Reauthorization Act (MACRA) and Merit-Based Incentive Payment System (MIPS) going live in 2017, a substantial amount of reimbursement dollars will be tied to quality metrics. Radiology practices must leverage performance data and EHRs to receive the biggest possible reimbursement.
According to an RSNA 2016 presentation by Michael E. Zalis, MD, an associate professor of radiology at Massachusetts General Hospital in Boston, utilizing clinical decision support systems, more in-depth EHR integration, and population health tactics can improve operational efficiency and reporting.
However, most current health IT tools are only able to handle structured data, said Zalis.
“They’re unable to fully leverage the narrative text that constitutes the majority of the information in the record,” he said. “When you think about it, when people don’t feel well they come to tell a story to their clinician, and our colleagues hear stories like this and recount them in their own language.”
In addition, EHRs aren’t designed with MACRA in mind. Vital tasks such as extracting information or making and submitting quality measurements can pile onto a clinician’s already busy workload, increasing costs while an advance clinical practitioner is taking time to generate these reports.
Zalis proposes an assisted clinical reasoning system: software that utilizes all kinds of data including unstructured reports written by physicians, imaging data and patient-reported information. A program like this would extract all relevant information and present it to the clinician in a clean user interface.
“You’d also like to be able to deploy machine intelligence so the system could learn based on user feedback, based on your particular use of language and the specific demands of your practice,” added Zalis.
This type of system was implemented in a leading academic institution to great success, according to Zalis. The system was used to generate CMS reports at nine sites in 2014, substantially reducing labor and increasing accuracy and compliance by around 30 percent. The money saved was significant, resulting in a $5 billion upside accrued from bonuses and reduced penalties.
“In another example, 2015 CMS quality measures were delivered using such an advanced clinician reasoning system, resulting in a two-thirds reduction in labor and a 47 percent reduction in the cost of generating these metrics,” said Zalis.
Another arena that benefits from a system like this is the prior authorization process when ordering advanced imaging studies.
“There’s too much labor and hassle in it, and typically the clinician is disengaged from this process unless there’s a denial,” said Zalis.
To realize this process into a clinical reasoning system, Zalis and his colleagues implemented the prior authorization decision tree, and instructed the system to pull and display information related to questions in the decision tree, using a neat software trick called conceptual queries. Both clinicians and patients engaged well with the system, according to Zalis.
“It reduced the burden and cost for the clinician, brought the patient into the conversation by presenting them their guidelines and their own data,” said Zalis. “The other key thing was that local insurance carriers began to accept this as a replacement for a traditional prior authorization.”