Privia Health, based in Arlington, Virginia, has a presence in 15 states and the District of Columbia. It builds scaled provider networks with primary-care centric medical groups, risk-bearing entities and a physician-led governance structure.
THE CHALLENGE
Privia Health’s clinics and IT setup faced several key challenges that impacted both its providers and operational efficiency across the national physician platform. The healthcare landscape is constantly evolving, and traditional methods are struggling to keep pace.
Chris Voigt is executive vice president and chief technology officer at Privia Health. He previously served as the vice president of business and product development and has 14 years of experience in the healthcare interoperability field.
Voigt offers a breakdown of the challenges Privia’s clinics and IT setup were facing.
“First there was provider burnout and administrative burden,” he said. “Provider burnout is a critical issue in healthcare. We found some of our efforts to optimize the system inadvertently added to this problem. Instead of alleviating burdens, some workflow strategies left providers to function as administrative reviewers.
“This task, while necessary, is time-consuming and doesn't contribute to the rewarding aspects of patient care,” he continued. “Providers were spending too much time on tasks that didn't directly involve using their medical expertise.”
Then there was the "Can't the computer just do that?" problem, Voigt explained.
“For years, providers have expressed frustration with tasks that seemed ripe for automation,” he noted. “The sentiment of ‘Can't the computer just do that?’ was common. We had accepted certain inefficiencies as inherent to the system.
“However, AI has emerged as a powerful tool, finally unlocking capabilities and providing solutions to problems we had previously considered intractable,” he continued. “Areas like reviewing extensive chart histories, accurately classifying and filing incoming documents, and identifying the correct patient insurance details were significant pain points.”
Data silos and inaccessibility were another pain point.
“One of the most significant challenges was making sense of the vast amounts of patient data, especially information residing outside of standard EHRs,” Voigt said. “Doctors often struggled to find the specific, relevant data they needed within this sea of information.
“This made it difficult to create meaningful interactions with patients and make the most accurate decisions,” he added. “The sheer volume of data was overwhelming, and the lack of efficient tools to process it hindered effective care delivery.”
Another challenge was back-office efficiency and automation.
“We consistently sought to enhance automation and efficiency in back-office operations,” Voigt recalled. “We had already taken steps such as optimizing our teams with a decentralized workforce, implementing robotic process automation, and engaging specialized vendors to address specific problems.
“However, we recognized the need to go further,” he noted. “We aimed to achieve a more productive workforce by leveraging more advanced technologies to streamline processes and reduce manual intervention.”
PROPOSAL
Artificial intelligence and other technologies newer to healthcare offer the opportunity to solve challenges in the healthcare ecosystem within the current EHR encounter workflow, practice optimization and overall workforce productivity, Voigt said. AI comes with a unique angle in that it might actually be so smart and revolutionary that it could “magically” solve problems in incredibly complicated business, he added.
However, finding the right technology can be a challenge in itself.
“Over the past two to three years, nearly every vendor has used the AI moniker on their wares,” Voigt said. “Too often, calling something AI seems like a sleight-of-hand trick, and it isn’t hard to see that they’re using more common technologies like advanced machine learning or have humans behind the scenes making the important AI decisions.
“With every vendor promoting some sort of AI integrated into their platform, it became hard to see through the apparent smoke and mirrors; prices were initially high, and it was hard to see how the ROI math worked out,” he continued. “Within the last 12 to 18 months, the smoke has cleared quite a bit; we can actively see how AI is impacting our provider experience and workforce productivity with demonstrated results.”
And the pricing of AI systems has come to better align with intended value, he added.
“We also see that anything labelled AI moves at a staggeringly fast pace,” he said. “It is hard to keep up with the advancements, and we even hear of some AI advancing just on its own, as it learns and improves its own internal models.”
MEETING THE CHALLENGE
Privia Health’s physician practices now benefit from a variety of AI tools embedded in its primary EHR platform from vendor athenahealth.
“We are seeing tangible benefits from inbound document classification, medication deduplication and even insurance selection,” Voigt explained. “We’re piloting patient experience tools that are showing promise both in improved experience and reduced provider documentation time.
“We also have had early, scaled success with our vendor Navina, a workflow system that reads and interprets complex data from various sources to create a concise, complete and timely view of patients' health status in the moment of care,” he continued. “This tool is integrated into our EHR platform and helps providers across all specialties comb through mountains of historic patient data to find previous diagnoses and suspected conditions.”
With this data pulled into provider workflows, Privia can help inform its providers of potential but not currently captured patient risks. This has been a game-changer that has both improved patient documentation and how providers take care of patients by ensuring past history doesn’t get buried in physician notes, he added.
“From an internal workforce perspective, all employees largely use AI and large language model tools to support their work,” he explained. “This has been transformative in nearly every department using AI to solve problems, draft policies, answer questions and write code. The organization as a whole has shifted from being wary of AI to seeing the tool as a critical part of carrying out the job.
“Despite recent technological advances bringing us further away from more human interaction, AI-driven technologies in doctors’ offices are having an opposite effect,” he continued. “We understand how to intentionally and strategically implement AI tools to automate administrative tasks and streamline data collection, allowing physicians the time to offer more personalized care.”
The integration of AI technologies creates a cohesive ecosystem that supports efficient, high-quality healthcare delivery, he added.
RESULTS
The Navina AI platform has offered Privia one of the most measurable successes, specifically within documentation and coding. Privia has achieved an 84% addressed rate for all coding suggestions given to a provider within a patient encounter to enhance documentation and specificity.
Additionally, there are nearly 90% active members within the provider network adopting this tool and seeing a reduction in time spent on chart prep by up to 2.5 hours per day. Doctors are able to focus more on patient care with minimal interference and administrative burden, Voigt said.
“With the addition of this tool, we are seeing front- and back-office optimizations of 10-30% and a reduction of insurance-related denials of around 7%,” he reported. “While these numbers may not seem significant at first glance, each individual improvement with a staff member at a practice unlocks downstream effectiveness on patient care or reimbursement.
“That one AI-corrected insurance classification at check-in saves significant work in the revenue cycle, reducing rework and accelerating payment,” he added.
Finally, Privia’s implementation of AI and advanced technologies as a part of its ambient scribe tools has resulted in substantial, measurable outcomes that have significantly alleviated physician burden, helped address care gaps and contributed to the improvement of patient outcomes, he said.
ADVICE FOR OTHERS
“Throughout every step in your AI evaluation, solicit provider input early and often,” Voigt advised. “These technologies have a lot to offer, and also change rapidly. Develop your early adopters into provider champions, leading the way across your organization to implement these changes.
“At Privia, we have a national IT advisory council that does just this, composed of multi-specialty physicians to review and guide our platform advancements, including AI opportunities,” he continued.
Additionally, be sure to establish a workgroup for developing and managing AI policies and evaluating vendor and service contracts, he said.
“If a technology seems too good to be true, it just might be,” he noted. “You have to look carefully to be sure your data isn’t being sold or used to inform other models. Don't ever think you are done with the above either: AI innovation moves incredibly quickly, and we've found it’s a perpetual cycle of tweaking policies, reviewing vendors and examining capabilities.”
Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: Chief AI Officers require a deep understanding of the technologies and clinical ops
