Building a Reliable National Provider Directory—Lessons From Other Sectors

US residents searching for a new health care clinician face a disheartening and onerous experience. In principle, provider directories are an essential tool used by patients to select in-network clinicians and by regulators and policymakers to audit access to care and monitor compliance with network adequacy regulations. However, across insurance markets, geographies, and health plans, provider directories are so highly inaccurate as to render them essentially meaningless. A 2018 report by the Centers for Medicare & Medicaid Services (CMS) found that more than half of clinician listings in Medicare Advantage directories included at least 1 error,¹ matching findings from audits and secret shopper studies.² A 2023 audit of directories from 5 large national health insurers found inconsistencies in 81% of entries³—a level of inaccuracy consumers would be unlikely to tolerate for any other service or industry.

This abysmal state arises from several factors. One challenge is the capricious and complex nature of the underlying data: directory data are akin to a phone book in a town where substantial shares of residents move every month, change their names, and sometimes list their neighbors’ phone numbers instead of their own. Monitoring these shifting data requires substantial time and resources, but because of fragmented data collection, maintenance, and verification processes across stakeholders, there is marked inefficiency and weak incentives to rectify inaccuracies.⁴ Administrative burdens to maintain data accuracy are borne not only by insurers but also by clinicians who must respond to frequent, redundant reporting requirements. A 2019 survey from the Council for Affordable Quality Healthcare estimated that physician practices collectively spend $2.76 billion annually on directory maintenance alone—equivalent to nearly $1000 per month per practice or 1 full-time staff member per week.⁵ Practices with fewer resources may lack the capacity to report frequent staffing changes and may be more prone to erroneous data inputs.

Moreover, regulatory enforcement has been historically weak, although there have been recent efforts to strengthen it. The 2021 Consolidated Appropriations Act specifies that health insurers, both private and public, now hold a default responsibility for maintaining accurate provider directories. All health plans are required to publicly list the clinicians and facilities with which they contract to deliver care to their members. Clinicians and health care facilities are also required to inform health plans of any material changes to their directory information or network status, with mandated timely updates and regular verification “no less than once every 90 days.”⁶ The law also establishes that consumers who see a clinician wrongly listed as in network are held only to in-network cost-sharing requirements. However, despite this new legislation, real-world compliance continues to be poor, with 1 study identifying that 45% of directory errors persisted well beyond 90 days.² Evidence so far suggests that enforcement alone is unlikely to transform directory quality without requisite changes in data input, verification, and coordination.

Finally, a core challenge in directory accuracy stems not from data quality alone but from potentially divergent purposes: directories are both administrative tools for reimbursement and patient-facing tools for navigation. These uses may be misaligned. For instance, group practices list clinicians as in network across multiple locations due to billing arrangements, even if those clinicians rarely practice at those sites. Although such listings may be technically accurate from a contractual standpoint, they misrepresent and overstate availability from the patient perspective. A truly functional national provider directory must be transparent about this tension without sacrificing usability, potentially by distinguishing contractual network presence from active clinical availability.

Universally poor directory data have spawned an industry of data brokers offering purportedly higher-quality clinician and health care facility information. However, existing commercial vendor solutions are proprietary, fragmented, and costly to maintain, with insufficient incentives to aggregate and verify clinician data across insurers, health systems, and regulatory boundaries. A privately owned national directory would also raise concerns about transparency, governance, and data stewardship, particularly if access was limited. The recognition that public infrastructure is more likely to ensure equity, accountability, and public trust has prompted federal efforts to address this problem. In 2022, CMS released a request for information seeking public input on a potential “national directory of health care providers and services” (NDH). By enabling a centralized data hub for directory and digital information, an NDH would build upon existing CMS systems, streamline updates and validation, and maintain directory information in a standardized, interoperable, and shareable format through a common application programming interface. Additional patient-centered features of the proposed NDH include new data inputs, such as clinician demographics, linguistic capabilities, telehealth service offerings, and acceptance of new patients.

Given the deeply entrenched dysfunction in the data ecosystem, an NDH is promising but stands to gain immensely from other sectors’ experiences. For example, Google Maps—a ubiquitous component of daily life for many—applies a multimodal approach that is successful in part because businesses have a strong incentive to maintain their own data and can upload information for free. Google also uses sophisticated machine learning and artificial intelligence algorithms to analyze satellite imagery, user reports, and traffic patterns to automatically identify and update changes while collaborating with government authorities to ensure the accuracy and consistency of the data collected. Simultaneously, crowdsourcing facilitates real-time updates, with users encouraged to report inaccuracies that integrate into data sources at little to no cost. Online services that rely on business directories, like Yelp, Tripadvisor, and Glassdoor, are variations on this model, which could be adopted in part or wholly in an NDH.

Many analogous government-sponsored projects have been successful with partnerships and learning from the private sector. For example, during the COVID-19 pandemic, the federal government rapidly sourced real-time data on pharmacy supply nationally to facilitate vaccine and COVID-19 therapeutic access. HealthCare.gov, recovered from its disastrous launch, now reports a public-facing, extremely complex patchwork of insurance plans with changing benefits and networks, while College Scorecard maintains detailed metadata and statistics on thousands of educational institutions from the US Department of Education’s mandated data collection. A universal provider directory is arguably more complex, but these examples from the private and public sectors show that the provider directory problem is not one of technological or logistical barriers.

Building a universal provider directory would not be prohibitively costly relative to the immediate public benefit and it could be feasible, if adequately funded, through the US Department of Health and Human Services. Several existing data sources, such as the Medicare Provider Enrollment, Chain, and Ownership System; Medicare Data on Provider Practice and Specialty; and the National Plan and Provider Enumeration System, could serve as foundational inputs, although each requires extensive validation and supplementation. Some states have already begun to operationalize centralized provider directories. California’s Symphony system centralizes data submission, validates participating health care clinician and facility information, and facilitates reconciliation of data discrepancies.⁷ Symphony now serves as the single data source for all qualified health plans in the state, reflecting the scale of infrastructure and coordination. Other states, such as Maryland, Rhode Island, Indiana, and Michigan, are exploring or already leveraging existing health information exchange systems to support similar efforts.⁷ These initiatives support the argument that a national directory not only has consumer-facing value but could also serve as critical infrastructure for health information exchange and interoperability. Federal efforts should layer onto, rather than duplicate, these promising state-level programs.

As policymakers seek tangible, high-impact improvements to the US health care system, a well-designed national provider directory represents a pragmatic and achievable step to improve transparency; reduce administrative burdens; and meaningfully enhance care experience, navigation, and access for all US residents.