WakeMed Health & Hospitals, a health system based in Raleigh, North Carolina, faced a significant challenge regarding its electronic health record. Dollars and quality points matter deeply to health systems, and the EHR is central to generating both revenue and quality scores.
THE CHALLENGE
However, improving the accuracy of documentation required asking physicians to spend more time in the EHR, which was already a major source of frustration and burnout. Research shows clinicians may need as much as two hours in electronic data entry beyond the exam room for every hour of direct patient contact – and physicians with insufficient time for documentation are 2.8 times more likely to report symptoms of burnout.
This created a “rock and a hard place” situation for WakeMed because it needed better documentation for financial and quality outcomes but pushing harder in the EHR risked worsening physician wellbeing and job satisfaction. The more the organization pushed physicians to include detailed documentation, the more it felt like a burden.
“I heard from many of our clinicians, ‘This isn’t what I trained for – I trained to care for patients, not to code charts,'” said Dr. David Kirk, chief clinical integration officer at WakeMed. In this role, he leads and coordinates transformational innovation and analytics initiatives to support the delivery of high-quality, cost-effective patient care. “But the reality is, if we don’t document correctly, we miss out on the revenue we’re owed and the quality scores we need. It’s all connected.
“Unfortunately, these tasks aren’t where most physicians or advanced practice providers thrive, especially coming out of COVID, when many are already exhausted,” he continued. “We had to find a way to make the chart better while taking work off the clinician’s plates, which led us to explore AI technology.”
PROPOSAL
To solve the documentation accuracy challenges, WakeMed used a three-pronged approach that included data, people and technology components.
From its extensive data work, the health system knew most of its opportunity was with patients seen by the hospitalist team. It installed an AI-powered clinical insights platform from vendor Regard and trained those physicians on its use.
“We provided a small amount of protected time to a hospitalist leader to help educate the group on the importance of the overall project and in our specific opportunities around clinical conditions,” Kirk explained.
“The Regard application helps providers by suggesting diagnoses and related supporting evidence after reviewing 100% of a patient’s electronic record,” he added. “The application’s at-the-elbow support was key in initial adoption and satisfaction.”
MEETING THE CHALLENGE
The AI-powered clinical insights technology helps WakeMed to directly address the documentation challenges physicians face within the hospital. The technology integrates with the EHR and works by reviewing and analyzing a chart to surface relevant clinical insight, summarize the patient’s record, and enhance diagnoses with supporting evidence that would take significant time and effort for physicians to manually compile.
By integrating the AI platform with the EHR and streamlining the charting process, WakeMed has been able to improve documentation accuracy and enhance coding and revenue capture – helping to address both operational and burnout-related challenges.
For example, a physician might diagnose a patient with sepsis and begin treatment. The AI application, however, takes it further by pulling in multiple layers of clinical evidence to support that diagnosis, detailing the severity and contributing factors.
“This not only strengthens the documentation but also ensures the patient’s condition is accurately represented, which is critical for appropriate reimbursement, quality scoring, and, most important, patient care,” Kirk said.
“Currently, we’ve implemented Regard primarily with our hospitalist team, since they handle the majority of inpatient documentation,” he continued. “These clinicians still review and finalize the information, but the AI application dramatically reduces the time needed to gather and organize data, enabling them to focus more on clinical decision making.”
There are plans to expand into other service lines like critical care. Eventually WakeMed will roll the application out to all physicians and advanced practice providers.
RESULTS
Despite being early in its documentary accuracy journey, WakeMed’s results have been impressive.
“Using the AI application has led to $9.3 million in claims paid that might have been denied,” Kirk reported. “Beyond reducing lost revenue, administrative burden and delayed payments, reducing denials helps reduce delays in care, ensuring patients receive their recommended treatment plans.
“Additionally, this has resulted in $871,000 in new revenue for Medicare Severity Diagnosis-Related Group, a system used by Medicare to determine how much to pay hospitals for inpatient stays,” he continued. “This increase in reimbursement occurs because the hospital is demonstrating it treated a more complex, resource-intensive case. $871,000 in new revenue means better documentation allowed the hospital to more accurately reflect patient complexity and be paid accordingly.”
There also has been a 3% improvement in severity of illness, helping clinical teams recognize the full clinical picture and leading to more tailored, appropriate care plans, he added. This also facilitates better handoffs and transitions between providers by clearly communicating complexity.
“The technology also has led to a 3.6% improvement in CC/MCC capture rate, which contributes to more accurate Hospital Compare and CMS star ratings, which can affect public perception and reimbursement,” he said.
ADVICE FOR OTHERS
One of the biggest hurdles Kirk has seen when considering the adoption of AI-powered clinical insights technology isn’t technical – it’s cultural.
“Many physicians still view the chart primarily as a tool to tell the patient’s story, and they may resist using tools that focus on documentation for risk adjustment, coding or revenue purposes,” he observed. “As AI software improves and allows physicians to maintain their own unique styles in documentation, adoption will increase.
“Explaining why documentation improvement is important is key in adoption,” he continued. “Although we know that documentation improvement drives revenue, clinicians are much more interested and excited about the improvements in risk-adjusted quality scores. Documentation accuracy allows the team to get credit in the various quality benchmarks for the amazing quality work they do.”
Kirk encourages provider organizations to involve physicians early on in the process, show them how the technology works, and demonstrate clearly and directly that it works to improve patient care and not just help the hospital.
“Accurate documentation isn’t just about reimbursement, it’s about ensuring continuity of care and giving the next clinician the full picture of the patient’s condition,” he concluded. “Focus on building trust, showing value and integrating the AI technology in a way that doesn’t replace but supports the physician.”
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: [email protected]
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: How to launch a healthcare AI project, per the VA AI chief
