Provider data operations is nobody's job.

Here is a scenario that plays out in health plans of every size. A provider who is contracted and credentialed joins your network. Their information gets entered into the credentialing system, then entered again (slightly differently) into the contracting system, then loaded (with some errors) into the claims system, and then — eventually — appears in the member-facing directory with an address that hasn't been verified and a phone number that goes to a fax machine.

Nobody intended for this to happen. Each team did their part. The credentialing team credentialed the provider. The contracting team filed the contract. The claims team loaded the provider into their system. The directory team published the directory. The problem is that these four data entry events happened in four different systems, by four different people, without a governance structure to ensure they were consistent.

This is the provider data problem. And it affects virtually every health plan, at every scale.

Why the gap exists

Each system that touches provider data has different purposes and different owners. The credentialing system is owned by the credentialing team and is optimized for tracking file completeness and committee review status. The contracting system is owned by the network management team and tracks contract terms, effective dates, and rate schedules. The claims system is owned by operations and is optimized for adjudication — it needs to know whether a claim from a given NPI should be paid at what rate. The directory is owned by... actually, this is where the gap starts.

Directory ownership is genuinely ambiguous at most organizations. IT manages the directory system. Network management is responsible for its accuracy. Marketing cares about how it looks. Member services gets the complaints when it is wrong. Nobody has the explicit accountability to keep all four systems reconciled with each other.

The directory accuracy compliance problem

The No Surprises Act and CMS regulations have made directory accuracy a compliance issue, not just an operational preference. Plans can face civil monetary penalties for directory inaccuracies that result in members being unable to access in-network care. State insurance departments have been increasingly aggressive on directory audits. The standard is essentially: if a provider appears in your directory, a member who calls that provider should be able to make an appointment and have that appointment covered at in-network rates.

Meeting that standard requires more than loading a provider into a directory. It requires that the provider is contracted, credentialed, accepting patients, reachable at the listed phone number, and located at the listed address. Each of those things can change without the directory team knowing about it — unless there is a maintenance process that catches the changes.

The taxonomy code problem

Provider taxonomy codes are the standard classification system used to identify provider type and specialty. There are nearly 900 active taxonomy codes in the NUCC (National Uniform Claim Committee) taxonomy code set. CMS uses specific taxonomy codes to classify providers for adequacy purposes. A provider who has the wrong taxonomy code in your system — say, a 208600000X (Surgery) instead of a 207X00000X (Orthopedic Surgery) — may be credited to the wrong specialty category in your adequacy analysis.

Taxonomy code errors are extremely common and almost always invisible until they cause a problem. The fix is a systematic taxonomy validation against the provider's NPI record in NPPES (the National Plan and Provider Enumeration System) — which is the CMS source of record for taxonomy codes. Any discrepancy between what is in your system and what is in NPPES is a potential adequacy and billing problem.

The claims system doesn't know what the credentialing system knows. The directory doesn't know what the contracting system knows. Provider data operations is the function that knows all of it — and most organizations don't have it.

Building a provider data steward function

The solution is to create an explicit provider data steward function — a person or team whose job is to own the consistency of provider data across all systems. Their responsibilities:

• Maintain the authoritative provider roster and push updates to all downstream systems.
• Run monthly reconciliation reports between the credentialing system, contracting system, claims system, and directory.
• Process provider updates (address changes, phone changes, NPI deactivations, specialty changes) across all systems.
• Validate taxonomy codes against NPPES quarterly.
• Run the directory accuracy verification process — confirming that providers in the directory are still practicing at the listed location and accepting patients.

This is not a full-time job for a large team. It is a consistent part-time function that needs to be explicitly assigned and resourced. What it cannot be is everyone's partial responsibility — which is what it usually is, and why provider data quality at most organizations is worse than anyone admits until a regulator audits the directory.