Managing provider data is a significant challenge for many payers, due to frequent changes and inaccuracies in directory listings.
Recently, generative AI and automation technologies have been identified as promising solutions. Research has shown these tools can meaningfully reduce payers' administrative and medical costs, while increasing revenue.
Becker's Healthcare recently spoke with Nate Fox, Chief Technology Officer at Ribbon Health, to learn how payer organizations are leveraging AI to create better member experiences and cut costs.
Poor-quality provider data creates risk and degrades the member experience
Fragmented and inaccurate provider data is a continuous challenge for payers. Incorrect provider directories limit members' ability to access appropriate care quickly. These hurdles often result in poor member experiences that can lead to decreased enrollment and retention.
Complying with consumer protection legislation, which mandates access to accurate and up-to-date provider information, presents additional challenges. Examples of these mandates include the federal No Surprises Act, as well as state laws like Illinois' Network Adequacy and Transparency Act and New York's Out-of-Network Consumer Protection Law.
Further, with network adequacy requirements, health plans must offer sufficient provider coverage across geographic areas to ensure members have timely access to care. It can be onerous for these organizations to analyze their networks, identify gaps in coverage, and proactively address them.
"Data management is inefficient for many payer organizations," Mr. Fox said. "They have big teams doing manual work, so they can check the compliance box, but they don't have the bandwidth to assess whether the information is actually accurate."
AI-powered platforms create efficiencies and impact the bottom line
Ribbon Health partners with payers, care navigators, primary care companies and other organizations to address fragmented and inaccurate provider data. Its industry-leading machine learning (ML) models predict the accuracy of provider data records, and the platform leverages AI to automate manual work.
"We determine where physicians are practicing based on a range of information we gather from across the healthcare ecosystem," Mr. Fox said. "We can predict the probability that directory records are actually true, so members can select providers with greater confidence."
Ribbon Health calibrates its AI and ML models using a truth set based on hundreds of thousands of phone calls to providers. This approach is aligned to how CMS defines the accuracy of directories.
Extensive research and development work is also underway to demonstrate the efficacy of large language models (LLMs) for unlocking further potential from provider data. Capabilities prioritized for development include automated provider outreach using voice recognition software that can confirm information and automate data quality checks.
"Sometimes provider data may look messy or fuzzy,” Mr. Fox said. "An intelligent LLM or AI agent can absorb that 'fuzzy' information and reconcile discrepancies or even validate data where there are ambiguities."
The right partner is key for AI success
Many healthcare organizations, including payers, are reluctant to adopt AI technologies because it can be daunting to satisfy the numerous security measures and requirements. One way to reduce this barrier is to work with an external technology partner that has taken a thoughtful approach to security protocol.
"You want a partner that has the right security certifications in place, like SOC 2 Type 2, HIPAA and HITRUST," Mr. Fox said. "Ribbon Health has invested in all of those things to ensure we are bringing AI innovations to healthcare organizations in a secure and compliant way."
It's also essential to find a partner with a proven track record in helping health plans reduce internal costs and improve the member experience. Mr. Fox pointed to the evolving AI vendor landscape as a challenge, as many companies lack expertise in how to best integrate tools with enterprise organizations.
However, if a health plan can find the right partner that is experienced in integrating into more established enterprise systems, that technology can offer significant ROI.
"If a company is spending $100 an hour on a specific job,, a software vendor might be able to do that work at a fraction of the price," Mr. Fox said. "Let's say a health plan has a team of 300 people doing gritty manual work that's operationally expensive. If we can supercharge that same work with AI and ML, and it's more automated and cost-effective, that can be very enabling for payers."