VA Prioritizes Practical Tech Solutions to Advance Modernization

Budgetary concerns and technical debt have spurred leaders at the Department of Veterans Affairs to focus on human-centered design, streamlining operations and bolstering interoperability, VA leaders said at the AFCEA Health IT Summit this week.

“We really need to develop ways and frameworks for interoperability. Leveraging that data through data classification before technical implementation will allow agencies and industry to adopt interoperability standards which help prevent costly retrofit,” Sarah MacDawutey, acting deputy director of Innovations at the VA, said. “Additionally, starting small, in my opinion, is how we get our greatest wins.”

MacDawutey said the agency is no longer “building bridges to nowhere,” but looking for tools and solutions that solve problems VA stakeholders actually face.

She emphasized that human-centered design is a core component of the VA’s data and AI strategy. Understanding the needs of veterans and clinicians alike has encouraged the development of new innovations designed to optimize their workflows, rather than add additional processes.

“We want to build something that people are going to use. If they’re not going to use a product, it’s not worth building,” she said.

User Feedback Drives VA’s AI Investments

Dr. Evan Carey, acting director of the National Artificial Intelligence Institute, pointed to the VA’s AI implementation as a significant boon to optimizing clinician workloads. He said AI can be effective in streamlining clinical decision support, so that clinicians do not have to spend their time sifting through enormous amounts of unstructured data.

However, he said the VA’s budget does not leave much room for failed products, so the agency focuses on user feedback early and often to make sure resources are used efficiently.

“We’re at the edge of this hype cycle … The risk we have is that we do a bunch of interesting things that don’t help veterans at the end of the day, and there’s huge opportunity cost there,” Carey said. “If we roll out things that don’t work well and don’t actually help people, we distract clinicians, and we distract staff from getting the work done, and we decrease trust.”

As the agency pushes to modernize, Carey also warned the VA must not endanger critical systems or implement new operations that greatly increase technical debt.

“Managing that technical debt and being smart about where and when we incur it without breaking anything, and then making sure we can pay that debt now with resources, but still modernize as fast as we really need to, is, I think, one of the core challenges,” he said.

Using Data as a Strategic Asset

Lisa Rosenmerkel, chief data officer at the VA, emphasized that even with the strength of technology like AI, its power can only be truly realized if good data underlies each system.

“Having siloed data that we can’t bring together in any way, shape or form, doesn’t serve us at all. Using data as a strategic asset for us means figuring out how to bring those data feeds together, figure out how we can look for broader patterns outside of our own programs and our own staff offices, and really look to innovation in that way,” she said.

John Scott, interim chief data officer at the Veterans Health Administration, said during a discussion on data science that no matter how effective the science is, communication will continue to be a critical part of developing trust and understanding when it comes to integrating new technologies as part of a digital transformation.

“However much you communicate is often not enough,” Scott said during the panel. “That communication aspect is going to be vital to the successful transformation we’re talking about.”