819,398 Providers. Zero Watching the Data Between Renewals.

The Claim That Came Back Wrong

Picture a billing coordinator at a mid-size multispecialty practice, a Tuesday morning, opening a denial in their queue. The reason code is familiar enough: provider not found in directory. But the provider is credentialed. Has been for three years. The license is current. The CAQH profile was attested four months ago.

So what happened?

She pulls the payer directory record. The address on file is the old location, the one the practice left eight months ago. Nobody caught it. Nobody was watching.

That is not an edge case. That is, from what I keep seeing in the data, the median credentialing failure in 2025.

An HHS-OIG audit found that 97% of Medicare provider records had data mismatches between databases (Prime Credential, March 9, 2026). Not a rounding error. Not a bad quarter. Ninety-seven percent, and the most common problem was something as mundane as an address that no longer matched.

What the 819,398 Number Actually Tells Us

We currently monitor 819,398 providers across 219,144 practices. That number is useful not because it is large, but because of what it reveals at scale.

When you watch that many records continuously, a pattern emerges that point-in-time credentialing simply cannot see. The risk is not concentrated at renewal. It is distributed across every day of the credentialing cycle, quietly accumulating in NPPES, in payer directories, in state license databases, in CAQH profiles sitting just inside their 120-day attestation window.

I think of it like this: your credentialing system tracks what you filed. It does not verify whether it is still true today.

Of the 219,144 practices we monitor, 12,026 currently show at least one provider record with a detectable mismatch between a primary source and a downstream directory. That is roughly one in eighteen practices with a live data integrity problem they may not know about. The mismatch is already there. The denial has not happened yet.

The Forty-Week Window Nobody Owns

Most practices credential on a two to three year recredentialing cycle. Between the initial enrollment and the next formal review, a provider can move offices, add a location, lose a hospital privilege, have a state board action opened, or simply let a CAQH profile slip past its 120-day attestation date. CAQH's own requirements are clear: providers who miss the 120-day re-attestation window risk losing active status, and payers rely on that data to verify enrollment before processing claims (Medix Revenue Group, February 9, 2026).

None of those events send a notification to your credentialing system. They just happen. And somewhere downstream, a payer directory record drifts quietly out of sync.

NCQA recognized this formally. Effective July 1, 2025, updated NCQA standards moved from episodic verification toward continuous monitoring, requiring organizations to review every provider at least every 30 days for license status, OIG exclusions, state board actions, and SAM.gov screening (insights.wchsb.com, January 27, 2026; medcaremso.com, April 1, 2026). That is not a minor workflow adjustment. It is a structural acknowledgment that the between-cycle window is where the compliance risk actually lives.

CMS went in the same direction from the payment side. The CMS-4208-F2 final rule, effective January 1, 2026, requires Medicare Advantage plans to update provider directory data within 30 days of any known change and to attest to accuracy annually (Credentialing DDS, June 11, 2026, citing the Federal Register, September 19, 2025). For multi-site groups operating across Medicare and Medicaid, the stakes got materially higher: starting in 2026, CMS strengthened cross-program termination enforcement, meaning a sanction or termination in one program can cascade across a provider's entire payer enrollment mix.

One data problem. Every payer at once.

One Address Change, Six Weeks of Drift

Here is how it actually unfolds. A clinic in Texas we scanned last quarter had a provider who relocated to a second office location and updated her group's website, her email signature, and her scheduling system. The practice administrator submitted the change to one payer. The NPPES record was not updated for eleven weeks.

In the meantime, three other payers were still pulling directory data from NPPES. Their records showed the old address. One of those payers runs a quarterly directory reconciliation. When the reconciliation flagged the mismatch, it did not just flag the address. It flagged the provider's entire record for review, which triggered a temporary hold on claims processing while the discrepancy was investigated.

The hold lasted 23 days. Services rendered during that window were not delayed. They were denied and, per the payer's billing rules for out-of-network submission, not recoverable.

The NPPES update that would have prevented it took about four minutes to file. The window between the address change and someone noticing the downstream consequence was six weeks.

What Continuous Monitoring Actually Does

Argoseer is not a credentialing system. I want to be direct about that. We do not perform primary source verification under NCQA standards, we do not issue licenses, and we do not replace CAQH, Medallion, Modio, or any CVO your team already uses. What we do is watch the data those systems depend on, every day, between the moments your credentialing workflow is active.

We run delta monitoring across NPPES, state license databases, and payer directory sources. When a record changes, whether it is an address, an affiliation, a license status flag, or an NPI anomaly, we surface it before it becomes a denial. The 12,026 practices in our current dataset with active mismatches did not get an alert from their credentialing system. The data changed and nothing fired.

MGMA estimates each day a provider is delayed in credentialing costs a hospital between $6,000 and $15,000 in lost revenue (pie Health, November 26, 2025). Sirius Solutions Global puts the per-lapse average at $7,500 per day (Qualigenix, June 3, 2026). And the part that does not show up in those figures: the loss is not temporary. Revenue for services rendered during a credentialing gap is permanently unrecoverable, not just delayed (Human Medical Billing, November 17, 2025).

Legacy workflows built on spreadsheets and periodic attestation cycles cannot detect this category of risk. Not because the people running them are careless, but because the systems are not built to watch the space between audits (Becker's Hospital Review, April 8, 2026; Verisys, May 26, 2026).

The Question Nobody Has Answered Yet

The regulatory direction is clear. NCQA mandated 30-day monitoring cadence as of July 2025. CMS built 30-day update requirements into the 2026 MA final rule. The financial math on undetected data drift is not ambiguous.

What is less clear to me is who, inside most practices, owns the between-cycle problem. The credentialing team owns renewal. The billing team owns the denial. But the 40 weeks between those two moments, the window where an address changes and a payer directory silently disagrees, that window belongs to no one.

If you want to explore what continuous monitoring looks like in practice, the product page at argoseer.com/product/monitor walks through how our pipeline works. But the more interesting question, I think, is not which tool you use. It is whether anyone in your organization is actually watching the data the day it changes.