In the past week, the federal government released two Requests for Information (RFIs): one from CMS seeking input on the Health Technology Ecosystem and its impact on Medicare beneficiaries (which includes mentions of provider directory), and another from the Department of Veterans Affairs (VA) on its Community Care Provider Directory. Meanwhile, CMS is actively piloting its National Directory in Oklahoma for exchange plans.
These efforts represent an opportunity for coordination between the two agencies on provider directories. As part of Defacto’s response to VA’s RFI, we encouraged VA to explore collaboration with CMS and potentially align on approaches related to data sourcing, accuracy assessment, and shared directory infrastructure. A substantial portion of the VA’s Community Care Network (CCN) overlaps with providers participating in CMS programs, including Medicare FFS, Medicare Advantage, Medicaid, and Marketplace plans. In fact, most physicians participating in Medicare are also part of the VA Community Care Network (source). Given the overlap, it follows that there are opportunities to rationalize administrative workflows and infrastructure.
Before either agency invests heavily in transformative IT infrastructure, they should consider a coordinated approach: one that reduces burden, lowers costs, and improves the chances of meaningful adoption.
How VA Can Benefit from CMS’s Directory Efforts
CMS has issued rules requiring data transparency around payer networks, and it has also been monitoring the accuracy of Medicare Advantage provider directories using ‘secret shopper’ methods for nearly a decade. VA has the opportunity to leverage CMS’s efforts around directory data improvement to supplement its own efforts. Here are some examples of how VA can immediately benefit from CMS’s directory efforts:
1. Leverage Payer Provider Directory APIs to Improve Data Quality – CMS requires all Medicare and Medicaid plans to publish standards-based Provider Directory APIs. While the data in any one given API contains errors, it is possible to predict record accuracy by measuring consensus across the hundreds of plans that participate in CMS programs. VA can adopt a “wisdom of the crowd” approach, comparing directory entries across payer APIs, to identify reliable data and flag discrepancies in its own CCN directory.
2. Cost Recovery Opportunities – By law, VA can bill third-party insurance when it knows that a Veteran has it. In the case of Community Care, if a Veteran goes to an in-network (relative to the Veteran’s private insurance) provider, then VA can bill third-party insurance when the care is for non-service related conditions. By using data sourced from payers’ FHIR APIs, VA can identify and refer Veterans to CCN providers with third-party insurance participation that align with Veterans’ coverage. There are also modern, coverage discovery databases that VA can use to discover third-party coverage of Veterans. The combination of ‘Insurances Accepted’ data for providers, and coverage discovery for Veterans, expands the opportunity for cost recovery over time, especially as an increasing volume of care occurs within the Community Care Network.
3. Support Seamless Clinical Data Exchange – As CCN providers join TEFCA-participating QHINs or expose ASTP/ONC-required Patient Access APIs, VA can leverage these to support clinical data exchange. As endpoint directories that include provider digital endpoints mature and become more available, VA will be able to use these to capture relevant clinical information about Veterans even in non-VA settings. While not a stated CCN goal in the RFI, enabling this exchange would enhance care coordination for Veterans.
4. Network Adequacy Alignment – VA and CMS could benefit from sharing appointment availability (and other) data to support network assessments. CMS is currently requiring payers to collect this information manually from providers. VA itself is receiving pressure to improve oversight of its TPAs supporting the Community Care Network. As both agencies are collecting information on location and appointment availability to support network assessments, they can share this data to hold operators of TPAs and plan networks accountable.
How VA Can Complement CMS’s Efforts
If VA moves forward with the work described in this RFI, it will be pushing the envelope in terms of innovation around provider directory accuracy, directory data interoperability, and dynamic appointment scheduling. This directory-focused effort echoes VA’s innovations around Blue Button in the 2010s, which paved the way for broader patient access to data in the following decade (among Medicare FFS and private carriers participating in Medicare, Medicaid, and Exchange programs funded by CMS). If VA demonstrates success in provider directory, it can similarly pave a path for CMS and the broader industry for better directories.
1. Promote Use of Standards – As VA seeks directory data from TPAs, it should promote conformance with the Da Vinci Plan-Net FHIR Implementation Guide. That is, the TPAs responsible for the networks should expose directory data via FHIR standards to VA. When working with CCN provider systems, VA should encourage use of Argonaut IGs for provider directories and appointment scheduling. VA and its contractors would then be able to tap into FHIR APIs to interact with directories and appointments. Promoting these standards will accelerate industry alignment, reduce long-term maintenance costs, and mitigate vendor lock-in.
2. Reopen and Maintain Its Own Provider Directory API – VA previously published a Provider Directory API, but it has been offline due to policy constraints. Bringing this API back online would signal VA’s support for interoperability and open data, while allowing others to build on and validate VA’s directory data. VA should lead by example and make its own data transparent, so as to encourage other health systems to do so.
3. Publish Appointment Scheduling Endpoints – VA could publish appointment scheduling endpoints for its own facilities and providers. This could enable innovation in Veteran access, especially in dynamic scheduling within VA facilities. This would serve as another successful reference implementation that private, non-VA health systems and providers could use as a model for surfacing their own scheduling endpoints.
4. Define “Accuracy” and How to Measure It – VA has stated a goal of achieving 90% data accuracy, which is a currently unachieved level within provider directories. VA also has not specified how accuracy should be measured. Defacto has seen multiple approaches: conformance with attested data, harmonization across internal systems, secret shopper audited data, and consensus scoring. VA should work with CMS to establish a shared, repeatable methodology for measuring accuracy that could become an industry standard.
5. Secondary Uses of Appointment Scheduling APIs – If providers expose standards-based scheduling APIs, this infrastructure could become a powerful tool beyond VA use cases. Appointment availability can serve as a proxy for provider directory accuracy: an indicator that a provider is seeing patients at a location and accepting new patients. This data could help payers clean up directories and support CMS in assessing access metrics without phone calls, laying the groundwork for availability-adjusted network adequacy standards.
The Case for Collaboration
Neither CMS nor VA needs to build for the other, but they should explore where their needs overlap, and identify common sense collaboration opportunities. Their goals are aligned, their networks overlap, and both stand to benefit from shared infrastructure and data. Even if they maintain separate systems, they can identify areas for joint work, such as standardizing accuracy definitions or exchanging data for accuracy verification. A coordinated federal approach would strengthen the entire health data ecosystem, lower administrative burden, and improve provider directory accuracy not just for Veterans or CMS beneficiaries, but for everyone.