Picture this. A member opens their health plan’s app, searches for an in-network cardiologist near them, finds one listed as accepting new patients, and calls the number. The line rings to a disconnected tone. They try the address on the next listing — the office relocated eight months ago. On the third try, the provider is still in the directory but retired two years back.
By the time that member dials your member services line, they are not thinking about provider data infrastructure. They are thinking about switching plans.
A June 2026 report from the HHS Office of Inspector General found significant gaps in how Medicare Advantage plans maintain and present provider information to enrollees. In their review, plans were unable to provide a working phone number for more than one in four network providers. Even more striking, nearly half of the providers appearing on official network lists were absent from the online directories that members actually use to find care — raising serious questions not just about directory accuracy, but about whether the network lists health plans submit to regulators reflect the same reality as what enrollees see.
Provider data verification is the process designed to prevent exactly this from happening. And for most health plans, it is not working well enough.
What exactly is provider data verification?
At its core, provider data verification is the process of independently confirming that the information a health plan holds about its network providers is accurate, current, and complete — and that it reflects what is actually true at the practice level today.
That word — independently — carries most of the weight. Health plans receive provider information from multiple upstream sources: roster spreadsheets submitted by large medical groups, CAQH ProView applications, Availity feeds, manual uploads from credentialing teams. All of that data arrives as-submitted. A plan has no way of knowing, at the point of ingestion, whether the address is still active, whether the phone number connects to that practice, or whether the provider listed as accepting new patients closed their panel six weeks ago.
Provider data verification goes back to the source. It means calling the practice directly, querying primary source databases — state licensing boards, NPDB, OIG, SAM — and cross-referencing what was submitted against what can be independently confirmed. The outputs are field-level: each data attribute gets a verification status, not just the record as a whole.
The specific fields that verification covers include:
Practice address — whether the provider is actually seeing patients at the listed location
Phone number — whether the listed number connects to the right practice and reaches a live operator
Provider NPI verification — confirming that the NPI on file corresponds to the provider practicing at that location
Specialty — whether the provider is actively practicing the specialty listed in the directory
Panel status — whether the provider is currently accepting new patients
Scheduling availability — including telehealth options and appointment window accuracy
None of these fields can be reliably confirmed through database queries alone. Address and phone confirmation, panel status, and scheduling availability all require direct outreach to the provider’s office — which is why phone-based verification remains the most irreplaceable component of any serious provider data validation workflow.
Why provider data degrades faster than most people assume
Here is something that surprises most people outside of provider operations: provider data is not stable. It is constantly in motion.
Studies find that between 20% and 30% of directory data changes annually. A physician practice contracts with an average of more than 20 health plans — meaning every time a provider moves, retires, changes their panel, or updates their contact details, that change needs to propagate across two dozen directories, each with its own submission process and update timeline.
The structural problem, as the AMA and CAQH laid out in a joint white paper, is that no single party controls all of the relevant information. Practices hold contact data and group rosters. Health plans control which physicians are under contract and which locations are covered. Panel status — whether a provider is accepting new patients under a specific plan — is jointly controlled by both parties, and neither has a reliable mechanism to keep it synchronized in real time.
The difference between verification and credentialing — and why it matters
One of the most persistent misconceptions in health plan operations is treating a completed credentialing cycle as evidence of directory accuracy. It is not — and the gap between the two is where most of the real risk lives.
Credentialing is the formal evaluation of a provider’s qualifications before they join a network: licenses, education, training history, malpractice record, board certifications. It happens at defined intervals — at onboarding, and typically every two to three years at re-credentialing. The output is a decision: this provider is qualified to participate in this network and bill for services.
Provider directory validation is something different. It is the ongoing confirmation that the information in your directory accurately reflects what is happening at that practice today — not when the provider was credentialed. A provider who was fully credentialed three years ago may have moved across town, changed their specialty focus, reduced their patient load, or had their license placed under a restriction. None of that surfaces in a credentialing record until the next re-credentialing cycle, which could be two years away.
The practical consequence: a directory can be populated entirely with fully credentialed providers and still be significantly inaccurate, because the real-world data — where they practice, what they do, whether they are available — has moved on. Provider data validation is the process that closes that gap between credentialing cycles.
What the compliance picture actually looks like
CMS has made provider directory accuracy a formal enforcement priority, and the numbers from its own audits are striking.
CMS’s national review found that 48.74% of provider locations in Medicare Advantage online directories contained at least one inaccuracy—wrong phone numbers, incorrect addresses, or outdated patient acceptance status.
The required threshold is 85% accuracy. The average plan, based on CMS’s own data, is nowhere close. Plans that fail CMS’s quarterly secret shopper audits — conducted without advance notice — face warning letters, corrective action plans, and civil monetary penalties. The Star Rating impact is arguably more significant: persistent directory failures feed directly into member experience scores that determine whether a plan qualifies for quality bonus payments worth hundreds of millions of dollars annually.
The regulatory floor is also rising. The REAL Health Providers Act, passed as part of the Consolidated Appropriations Act, 2026, adds mandatory proactive 90-day verification for every provider record starting plan year 2028. Beginning plan year 2029, directory accuracy scores will be publicly visible on Medicare Plan Finder — meaning a health plan’s provider data validation performance will be a consumer-facing metric visible to prospective enrollees at the moment they are choosing coverage. That changes the stakes considerably.
Why manual verification cannot keep up
The conventional response to directory inaccuracy has been to run verification campaigns: dial through a list, update what’s wrong, close the ticket. The problem is not the effort — it is the math.
For a plan with 50,000 provider records, that is $200,000 in direct outreach costs before accounting for management overhead, rework cycles, and the inevitable lag between when a campaign closes and when the corrected data actually makes it back into the directory. And by the time that campaign is done, the data has already started drifting again.
Manual verification also produces inconsistent outputs. One agent hears “we moved last spring” and updates the address. Another hears the same thing and marks the call as inconclusive. There is no standardized field-level outcome, no automatic audit trail, and no pipeline back into the system of record. The work gets done, but the documentation rarely holds up under regulatory scrutiny.
What modern provider data verification looks like
The shift happening at leading health plans is away from periodic verification campaigns and toward continuous, automated provider data validation — treating verification as ongoing infrastructure rather than a quarterly project with a start and end date.
Modern automated outreach platforms handle the phone-based verification layer that manual teams have always struggled to scale: calling practice locations, navigating IVR systems, conducting structured conversations with front desk staff, and returning field-level outcomes — Accurate, Inaccurate, or Inconclusive — for each data attribute per provider. Every interaction produces a timestamped audit trail with a full call transcript. Structured results feed directly into directory management and credentialing pipelines without anyone manually re-keying data.
The operational result is a verification program that runs continuously, keeps pace with the rate at which provider data actually changes, and produces the kind of documented evidence that holds up when an auditor asks how a specific directory entry was verified and when.
Provider directory validation is not a back-office function to be optimized around the margins. It is the foundation on which member access, regulatory compliance, and network credibility all rest — and the health plans building that foundation on automated, continuous infrastructure are the ones best positioned as the compliance requirements ahead continue to tighten.

