How data analytics supports Virginia's Medicaid services
Connecting state and local government leaders
The Virginia Department of Medical Assistance Services is bringing data analytics to the state's Medicaid services.
The Virginia Department of Medical Assistance Services created the Office of Data Analytics in 2014 to develop a data governance policy. Bhaskar Mukherjee, who serves as director of the analytics office, spoke at the Sept. 19 SAS Analytics Experience conference about how DMAS is taking data in its “rudimentary form” and making it usable for supporting decisions on Medicaid services.
Eligibility for Medicaid depends on a number of factors that DMAS must analyze to provide the best possible services, he explained. “When we try to make decisions, we want to know how we can better serve [our residents] and part of that is looking at data as entities rather than transactions,” Mukherjee said.
When looking into metadata management, Mukherjee emphasized the need to define data elements to create organizational knowledge.
“You need to create accountability for where the data is coming in and governance tries to define that through functional, technical and business processes,” he said.
Other elements of the DMAS data analytics strategy include data acquisition, a master data management plan and an analytics platform designed to meet the needs of analysts with different skill levels.
DMAS uses SAS Office Analytics and Visual Analytics platforms to interpret data. “The platforms give us the ability to not just process data but also put it into a visual and technical form,” Mukherjee said.
DMAS is dealing with a large amount of data that it maximizes for operational efficiencies. For instance, each claim submitted has 1,500 data elements and DMAS gets 60 million claims every year to be processed.
Data analytics is also helping DMAS to connect Virginia residents to opioid addiction services. By connecting SAS with the search engine Bing’s mapping technology, analysts can determine areas where more coverage should be added.
“Our chief medical officer has been able to set up programs and recruit providers,” Mukherjee said. “This [analytical mapping] allowed us to determine our partners and bridge the gaps.”
DMAS also has been able to identify high-use, high-risk Medicaid recipients. By targeting the people who use DMAS services the most, Mukherjee said the agency is able to “effectively prevent some things from happening.”
After his presentation, Mukherjee told GCN that DMAS decided to move forward with SAS Visual Analytics capabilities after 15 years of experience with the software on the desktop.
“These new tools bring operational efficiency and automation so you don’t need to press a button every Monday morning to run something,” he said. “You can automate some of these tasks so a person can take on more of a data load than just 40 hours a week.”
With data visualizations, Mukherjee sees the push toward more interactive data analytics allowing analysts to push out more information that people can use. “From a strategy point of view, I’m always looking for innovations that can keep the shop running at a reasonable price,” he said. “Affordability, adaptability and continuity are fundamental to our success.”
When it comes to the future tools for analytics, DMAS is interested in ways to predict Medicaid outcomes. Work is also being done to create public-facing dashboards, but the agency must first determine what data can be made available to the public.
Note: This article was updated on Sept. 21 to correct the number of data elements in each DMAS claim.
NEXT STORY: Leveraging big data for pediatric health