Inaccurate data in provider directories is costing health plans. Why a national provider diectory and data science are the fix

For health plans, inaccurate data in the provider directory has become a major legal liability and financial risk. On the horizon are policy changes such as the REAL Health Providers Act and the CMS National Provider Directory (NPD) initiative that will reshape how provider data is collected, maintained, and enforced. For health plan executives, knowing how and when these new policies will be implemented and how a better data infrastructure can impact provider directory accuracy is crucial.

Why provider directories are so inaccurate

The core issue to provider directory inaccuracy is that no single entity owns the data. A health system maintains its own internal roster, and payers display whatever information doctors have given them. Online directories can go months or even years without updates. When a physician changes locations, retires or stops accepting a particular insurance, that change rarely propagates consistently across the systems that need it.

Approximately 30% of provider data churns on an annual basis. Physicians move practices. They change which insurances they accept. Reimbursement rates shift. When these changes are not automatically updated across systems, provider directories quickly become outdated.

Additionally, group practices often list every physician as practicing at every location. That's rarely accurate, and it creates real confusion for patients trying to find care at a specific site.

On top of this, before open enrollment, health plans are motivated to list as many in-network providers as possible. The larger the network looks, the more attractive the plan. It is only after enrollment closes that the cleanup begins. Directory accuracy actually tends to decline at the moment patients are making coverage decisions.

The result is what the industry calls a ghost network, a directory full of providers who aren't actually accessible to the patients looking for them. A Health Affairs study found that 53% of patients searching for care in health plan directories found inaccuracies. CMS audits of Medicare Advantage plans’ online directories found error rates approaching 50%.

The legal and financial stakes

Beyond the inconvenience to patients, health plans bear the legal and financial liability when the directories are wrong.

If a directory lists a provider as in-network and a patient receives care under that assumption, the plan can be held responsible for coverage costs even when the provider relationship doesn’t hold up to scrutiny. Recent enforcement shows directory errors can create direct financial exposure for health plans, including per-member CMS penalties and large reimbursement outlays tied to out-of-network claims.

Among the more serious risks to the health plan are regulatory fines and legal exposure. In fact, several major plans have faced significant legal consequences as a result of ghost networks. Health Net, Cigna, Emblem Health, Health First, and Blue Shield of California are among the organizations that have faced lawsuits or regulatory action tied to inaccurate provider directories.

Plans can also face claims leakage, a hit to STAR scores, member churn and an unnecessary administrative burden dealing with appeals processing and customer service issues. All of these legal and financial liabilities collectively become a major issue for the C-suite.

And for members, these inaccuracies can leave them with surprise bills, delayed care and frustrating searches for in-network providers.

The policy response

Two complementary policy initiatives are now moving to address this systemically. The REAL Health Providers Act, signed into law on Feb. 3, 2026, establishes accountability standards for the accuracy of provider information. It introduces new federal requirements to improve provider directory accuracy in Medicare Advantage plans, addressing “ghost networks” and protecting patients from unexpected out-of-network costs. It shifts the dynamic from a voluntary, loosely enforced status quo to one with defined obligations for both providers and plans to keep data current.

The CMS National Provider Directory (NPD) initiative is a centralized infrastructure designed to serve as the authoritative source of provider data across the ecosystem. The NPD doesn’t isn’t just another database. It creates a mechanism to force providers and health plans to keep information up to date in one place, rather than across dozens of disconnected systems.

Together, these two initiatives represent a shift of accountability. Health plans, providers and the broader infrastructure are each being asked to take ownership of data quality.

Beyond policy changes

Fixing provider directories is not just a matter of policy mandates. It requires access to a diverse set of provider data signals and experience in data science.

Inaccurate provider data isn’t random; a small set of fields drives most of it. Accuracy climbs as intake scoring and maintenance take hold, then dips each annual enrollment as networks inflate faster than validation can keep up.

Healthcare organizations need to stop treating provider data as a file you fix on a schedule and start treating it as a living asset. In practice that means three things, working as one connected pipeline: automating how data comes in, how it is kept current, and how it is turned into intelligence.

Location-level accuracy. Interestingly, the provider not being at the listed location is the single largest source of inaccuracy. Approximately 30% of providers move each year. Getting to accurate, location-specific provider data requires combining multiple signals: Electronic health record data from systems like Epic, claims data, direct outreach to provider offices and predictive modeling to fill gaps and flag inconsistencies.

Real-time currency. A directory that's accurate today may be stale in six months. The 30% annual churn rate means that data pipelines need to continuously ingest new signals, such as insurance acceptance changes, practice relocations and patient volume patterns, and updating them accordingly.

Network and access signals. Whether a provider nominally accepts an insurance plan is different from whether they have meaningful availability for patients on that plan. Distinguishing between the two requires richer data, including data from electronic health records, that can show actual patient populations at specific locations.

Health plans that invest in that ecosystem and layer their own data science on top of it will be better positioned to meet compliance requirements and, more importantly, to deliver the accurate networks their members need. A governance layer keeps that record consistent and auditable across the organization, and purpose-built solutions turn it into views each team can act on. That chain is what lets a health plan derive intelligence from provider data instead of simply storing it. The destination is not a better directory tool. It is a living data asset: one canonical, continuously scored record per provider-location that every downstream surface reads from, with intake, maintenance, and intelligence talking to each other rather than running as disconnected projects