How AI helped 2 health systems automate case reviews & reduce claim denials

Healthcare lags behind other industries when it comes to adopting technologies, and artificial intelligence tools are no exception, according to XSOLIS Chief Medical Officer Heather Bassett, MD.

During an April 30 webinar sponsored by XSOLIS and hosted by Becker's Hospital Review, Dr. Bassett discussed how organizations can unlock AI's potential, with prime examples from Charlotte, N.C.-based Atrium Health's Director of Clinical Care Management Tonya Harrison and Knoxville, Tenn.-based Covenant Health's Revenue Integrity and Utilization Management Officer Sherri Ernst.

A 2018 survey of 200 healthcare decision-makers revealed that a mere 37 percent of respondents used AI tools — ones that rapidly process and learn from large amounts of data — at their organizations. The problem continues as most of the 37 percent only use AI to a limited extent. Rather than harnessing AI to reduce administrative burden and drive operational efficiencies in real time, providers use it for simple, task-based purposes only, according to Dr. Bassett. 

"Utilization review is still a very manual — even paper-driven — process, which is shocking if you consider the volume of constantly changing healthcare data associated with just a single patient visit, let alone the census that a UR nurse or a case manager is responsible for," Dr. Bassett said.

When organizations leverage AI to the extent of its capabilities, it is a technology that empowers UR nurses to practice at the top of their license and make informed decisions directly tied to revenue cycle success, she said.

Wielding AI tools better manage care

XSOLIS' AI-supported smart utilization technology can make highly accurate electronic predictions about a patient's status using data directly from the EMR, Dr. Bassett said.

Without requiring user input, XSOLIS' software generates care level scores for patients. A score greater than 75 indicates the patient's ideal care setting is likely inpatient. A significantly lower score is a strong indication the patient is appropriate for outpatient treatment with observation.

These scores empower organizations like Atrium Health — one of the platform's early adopters — to assess and stratify their patient populations, Dr. Bassett said. Using that information, nurses can make informed decisions about what care settings patients should be treated in and which patients should be prioritized for review, ultimately freeing up more time for patient interaction.

XSOLIS allowed Atrium's UR nurses to apply their training and expertise, and it helped the system better manage extended observation cases, Ms. Harrison said.

"We realized the efficiencies in XSOLIS with some of our inpatients," Ms. Harrison said. "There were cases that really did not need a UR touch by a nurse, based on the care-level score. So, we worked with XSOLIS to answer the question 'If this patient meets inpatient-criteria level of care coming through the door, is there truly a reason for a UR nurse to look at that entire medical record?'"

The outcomes were exciting: XSOLIS helped Atrium determine with 99 percent certainty that patients with a care level score of 120 were solid candidates for inpatient-level care, allowing them to automate reviews on that subset of patients across select payers.

Improving payer-provider relationships through AI

XSOLIS' predictive analytics technology also helps organizations target another major pain point: reimbursement disagreements between payers and providers. Among other benefits, the software establishes a common framework with shared analytics between the two entities to help determine medical necessity in reimbursement disputes.

That seemed like a promising solution to Ms. Ernst, whose nine-hospital health system was at a crossroads.

"For Covenant, we [realized we] needed to look forward and not go backward because that was really the only way our process was going to be viable," she said.

Accurate status decisions made possible through XSOLIS technology helped increase clinicians' productivity, according to Ms. Ernst, but the platform also helped payers validate claims. As a result, Covenant was able to decrease its volume of calls, faxes, and ultimately, denials "almost immediately," allowing staff members to devote more time and resources to sicker patients, Ms. Ernst said.

The reduction didn't happen because physicians were suddenly documenting everything perfectly — it was because payers could now access all the information they needed to support medical necessity determinations within the XSOLIS platform, Ms. Ernst said. That collaborative process in turn helped Covenant achieve one of its primary goals: reducing administrative burden.

"We have an obligation to our patients and our communities to look for any and all ways to decrease administrative burden so we can focus those dollars back on patient care," she said. "Everyone needs to ask their payer partners to participate in conversations with XSOLIS because … the more of us who actually ask, the more leverage we have in making this happen."

More articles about AI:
Study: Deep learning can predict reactions to thrombolysis
Machine learning used to detect tuberculosis strains' drug resistance
AI-powered kidney disease diagnostic tool receives FDA 'Breakthrough Device' designation

© Copyright ASC COMMUNICATIONS 2019. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.

 

Top 40 Articles from the Past 6 Months