Ghosts in the Network

How Inaccurate Medicaid Managed Care Provider Directories Undermine Real Access to Maternal Care in the United States

By Elena Pak, Credentialing Department, WCH

A System that Exists on Paper More than in Practice

The latest findings from the U.S. Department of Health and Human Services Office of Inspector General (HHS Office of Inspector General) point to a structural weakness in Medicaid Managed Care that is rarely visible at the level of policy debate but is immediately tangible at the level of patient experience: the instability and inaccuracy of provider directories.

At first glance, provider directories are a technical artifact of insurance administration. In practice, they function as the operational interface between patients and the healthcare system. For Medicaid enrollees, particularly pregnant patients who must navigate time-sensitive prenatal pathways, these directories are not informational—they are determinative.

What OIG documents is not a series of isolated administrative errors. It is a systematic divergence between three layers of reality: regulatory filings submitted by managed care organizations, internal network databases maintained by insurers, and the publicly accessible directories used by patients attempting to access care. These layers, which should in principle be aligned, increasingly are not.

Maternal Health as the Stress Test

Any discussion of Medicaid network adequacy inevitably intersects with maternal health outcomes. The United States continues to report maternal mortality rates that remain high by high-income country standards. According to CDC surveillance data, 649 women died from pregnancy-related causes in 2024, corresponding to a rate of 17.9 deaths per 100,000 live births. The disparity between racial groups remains substantial, with Black women experiencing mortality rates several times higher than White women.

This context is essential because Medicaid finances more than 40% of all births in the country. In many states, Medicaid Managed Care is not a peripheral payer—it is the dominant financing structure for maternity care.

In such a system, access is not merely a function of insurance eligibility. It is a function of whether a listed provider is actually reachable, actually participating, and actually able to see patients within clinically meaningful timeframes. This is where OIG’s findings become consequential.

The Gap Between “Listed” and “Available”

Across multiple large Medicaid Managed Care organizations—including Centene, UnitedHealthcare, and Elevance Health—the OIG identified persistent mismatches between provider networks submitted for regulatory adequacy review and the networks reflected in patient-facing directories.

The central issue is not simply outdated information. It is structural inflation of network reality.

Providers included in regulatory submissions were frequently found to be no longer participating in Medicaid Managed Care, no longer accepting new patients under those plans, or no longer practicing at the listed location. In some cases, the discrepancy rate approached levels that call into question the reliability of the network submission process itself.

More striking is the divergence between compliance documentation and patient experience. In multiple instances, large proportions of providers listed in regulatory filings did not appear in public directories at all. Conversely, directories often contained providers with incorrect contact information or outdated participation status. The result is a system in which regulatory “network adequacy” and real-world accessibility increasingly operate as separate constructs.

The Operational Meaning of a “Ghost Provider”

The term “ghost provider” has emerged to describe clinicians who exist within insurance datasets but are not functionally available to patients. The concept is deceptively simple but systemically revealing.

A ghost provider may be a clinician who left a network without timely update, a retired physician still listed as active, a provider incorrectly marked as accepting Medicaid Managed Care patients, or a practice that has changed location without synchronized updates across payer systems.

What matters is not the category of error but its effect: the creation of a false positive signal of access. For a Medicaid enrollee, particularly someone seeking obstetric care, this translates into repeated failed contact attempts, delayed appointments, and ultimately restart cycles in provider search. The system produces friction at precisely the point where continuity of care is most clinically sensitive.

From Administrative Inefficiency to Clinical Consequence

The downstream effects of directory inaccuracy are often underestimated because they are diffuse rather than discrete. There is no single catastrophic failure event. Instead, there is attrition through delay.

In maternal care, timing is clinically significant. Early prenatal engagement is associated with improved detection and management of conditions such as preeclampsia and gestational diabetes, as well as better neonatal outcomes. When patients lose days or weeks navigating inaccurate provider lists, the system is not merely inefficient—it is probabilistically degrading outcomes.

This is particularly relevant in Medicaid populations, where structural barriers to access already exist. Directory inaccuracies do not create inequality, but they amplify it by adding an additional layer of uncertainty precisely in populations already at higher clinical risk.

Why the System Produces this Outcome

OIG’s analysis implicitly points to an incentive architecture problem rather than a technical failure.

Managed care organizations are required to demonstrate network adequacy to state regulators. These submissions, however, are episodic and largely document-based. They certify the existence of a network at a point in time rather than its functional accessibility over time.

Maintaining continuously accurate directories, by contrast, requires persistent reconciliation of provider status, participation, location, and availability across multiple systems. This is operationally costly and often not directly tied to financial penalties.

At the provider level, additional fragmentation emerges. Clinicians may participate selectively across Medicaid Managed Care plans, fail to update participation changes across multiple insurers, or operate within administrative environments that do not prioritize real-time data synchronization.

The absence of a unified national provider directory compounds the problem. Each state and payer maintains its own dataset, with no authoritative reconciliation layer. As a result, inconsistencies are not exceptions—they are structurally produced.

A Regulatory System Built on Representation Rather than Verification

Perhaps the most important implication of the OIG findings is conceptual. The current framework of network adequacy is primarily representational: it evaluates whether a network appears sufficient based on submitted data.

What it does not consistently verify is whether that network is operationally accessible. This distinction matters. A system can be compliant in documentation while failing in function. In Medicaid Managed Care, this divergence appears to be widening.

OIG’s recommendations—including “secret shopper” validation, use of claims data to verify active participation, stronger accountability mechanisms, and exploration of a unified provider directory—are not simply technical fixes. They represent a shift from static compliance to dynamic verification.

The Centers for Medicare & Medicaid Services (Centers for Medicare & Medicaid Services) has indicated support for these directions, but implementation remains an open question.

The Deeper Design Failure

At a structural level, the issue of ghost providers reflects a broader characteristic of fragmented health system governance in the United States: reliance on self-reported, periodically updated compliance artifacts in place of continuously verified operational data.

In such a system, incentives naturally favor completeness of listing over accuracy of maintenance. It is easier to demonstrate that a network exists than to ensure that each element of that network remains current, reachable, and actively participating. The consequence is a persistent gap between regulatory reality and patient reality. It is not accidental. It is produced by design.

Access as a Data Integrity Problem

The OIG findings should not be interpreted narrowly as a database quality issue. They point to a deeper misalignment between how Medicaid Managed Care is regulated and how patients actually access care.

“Ghost providers” are not anomalies in an otherwise functioning system. They are predictable outputs of a system that treats provider networks as static compliance objects rather than dynamic service infrastructures.

Until network adequacy is defined in terms of verified, continuously updated access—rather than submitted documentation—the divergence between reported and real availability will persist.

For patients navigating pregnancy under Medicaid Managed Care, this divergence is not theoretical. It is the difference between a listed appointment and an unreachable clinic, between nominal coverage and actual care, and ultimately between a system that appears functional and one that reliably is.

Selected Sources