Four Takeaways for Smarter Provider Data Systems

Accuracy Is Now a Regulatory and Reputational Imperative

In a recent Fierce Healthcare webinar with CertifyOS and Oscar Health, one central truth surfaced across every discussion: provider data accuracy and modern infrastructure must be solved together. The two problems reinforce each other — and neither can wait.

For years, provider directory inaccuracies were treated as an operational inconvenience. Regulatory pressure has changed that calculus.

Candor Health CEO Sury Agarwal framed the stakes directly: “The No Surprises Act required directory updates every 90 days, but the Real Health Provider Act takes it further with civil penalties for inaccuracies.”

CMS now allows Medicare Advantage members to switch plans mid-year when directories are found to be inaccurate — a provision that translates directory errors into direct membership exposure.

Mitch Gorodokin, Senior Vice President of Business Development at CertifyOS, added a structural caution that applies regardless of how much automation a health plan deploys.

The Data Problem Is Structural, Not Just Technical

Provider data doesn't fail because health plans lack tools. It fails because the same data moves through multiple organizations — payers, MSOs, delegated groups — each operating on different definitions, formats, and update schedules.

That structural fragmentation means a single change propagates unevenly. Agarwal described the core challenge: “Even if accurate data exists somewhere, maintaining and distributing it across systems is incredibly hard without the right infrastructure in place.”

Ownership is part of the answer. Health plans cannot assume clean data will arrive from delegated groups or provider organizations.

“We cannot rely solely on delegate or provider groups to send us clean rosters,” Agarwal noted. “We need proactive monitoring, validation, and enrichment of the data ourselves.”

Accuracy has to be embedded into workflows — not applied as a downstream correction.

Fragmented Incentives and the Ghost Network Problem

One of the more difficult dynamics in provider data management is that the organizations contributing data don't always share the same priorities around accuracy. Provider groups have operational reasons to over-list providers. Payers absorb the compliance and member experience consequences.

The result is ghost networks: directory listings that don't reflect actual network availability, frustrating members and drawing regulatory scrutiny.

The panel's view was that intelligent validation infrastructure and unified data governance can help realign those incentives. When data integrity becomes a shared operational standard rather than each party's internal problem, the network-level benefits compound.

As Agarwal observed, “Accuracy, efficiency, and member experience can move in the same direction when incentives are aligned.”

What Future-Ready Provider Data Infrastructure Looks Like

The panel converged on a clear picture of what scalable, interoperable provider data management requires:

• National visibility into provider practice locations and network overlaps

• National visibility into provider practice locations and network overlaps

• Infrastructure capable of supporting directories, referrals, and claims simultaneously

• Intelligent survivorship logic that resolves conflicting records and establishes a single source of truth

AI Depends on the Integrity of the Data Layer

The AI dimension matters here too. Downstream workflows — from claims routing to member-facing directories to network analytics — increasingly depend on AI-assisted processing. But those workflows inherit whatever data quality exists at the foundation.

Agarwal was direct: “AI is only as effective as the data feeding it. Stability at the data layer is essential.”