DirectTrust recently submitted comments in response to the CMS National Directory RFI.   As CMS outlines, their “request for information solicits public comments on establishing a National Directory of Healthcare Providers & Services (NDH) that could serve as a “centralized data hub” for healthcare provider, facility, and entity directory information nationwide.”

As an organization dedicated to instilling trust in health data exchange, we’ve built our own Directory to support trusted and timely Direct Secure Messaging. Earlier this year, we publicly announced our Directory Improvement Initiative to increase data quality of Direct addresses listed in our Directory. Our Directory Improvement Initiative is ongoing, and our experience has informed a significant part of our commentary.   We know the importance of directories, and as we’ve outlined previously, we support a National Directory effort.

We worked with our membership to craft the following comments in response to the CMS National Directory RFI.  Due to the length of our comments, the full comment letter may provide the best reading experience and is linked at the end of this post. A summary of our general thoughts around the CMS National Directory RFI, including background on our unique position to respond to request for comment, is included below.

Summary of CMS National Directory RFI Comments

The DirectTrust community welcomes the development of a National Directory of Healthcare Providers and Services (NDH).  We believe that a standards-based, national approach that takes advantage of existing networks/assets and requires provider organizations to periodically attest to the accuracy of data they provide can, if appropriately architected, be a significant burden reduction and advance multiple benefits of interoperability for patients, providers, payers, and even healthcare adjacent networks of service providers in social care settings. 

Our years of experience operating a multi-stakeholder, multi-contributor, multi-tiered ecosystem for provider directory aggregation has taught us a great deal about both the technical challenges of such an endeavor and what processes/policies need to be contemplated as a part of a successful model.  We look forward to sharing our experiences both as a part of this RFI and going forward.

Additionally, DirectTrust is in the middle of a substantial project to improve the quality of our directory data through field validation and verification against primary sources, as well as a pivot toward an early instantiation of the National Directory Standards work expected to be the foundation for the NDH which is under development at HL7.  We hope to be at the table as decisions are made about how a NDH will be deployed so we can share our ongoing experiences with both the general challenge of deploying a healthcare directory and with the emerging national standard.  We also hope to promote a role for the DirectTrust Directory as a contributor to the NDH, as well as to generalize the notion of an “Attestation Service Provider Organization” that could serve health systems and others with substantial curation challenges for their directory data. Such a model would allow for the market to advance workflows beyond the capabilities of the NDH itself to improve usability and burden reduction. 

Today’s provider directory environment is another costly artifact of our decentralized free-market healthcare system which deserves our collective attention.  The lack of complete and accurate provider directories creates burdens for patients attempting to locate providers and particularly for the task of choosing among providers across different networks.  Providers also struggle without reliable directory information available in their workflow. We have heard organizations attempting to comply with the ADT Notifications component of the CMS “Interoperability and Patient Access” rule (CMS-9115-F) have fallen short when attempting to locate digital endpoints for downstream providers or provider organizations, even when the patient stipulates a practitioner be notified as is required to be supported in the rule. This may be because of an inability to find an existing address due to ambiguous or inaccurate directory attributes, or because an address has not been published in the DirectTrust Directory at all.  

Providers looking to refer patients outside of their network have a similar challenge.  While many of these referral relationships are known and established, electronic endpoint information such as Direct Secure Messaging addresses for these providers is frequently not available in local workflow systems, even EHRs and the Health Information Service Providers (HISPs) who serve them.  These addresses are almost never available in public-facing marketing-oriented directories published by health systems, which may be appropriate unless the address is built and monitored for the purpose of communicating with patients.  We propose organizations deploy addresses at the organization level, constructed as  “workflow” or departmental addresses which can be discovered in a directory even if a specific provider is sought.  We also advocate that publishing a “purpose” attribute on a directory entry would allow any address to publicize its appropriate uses.  Including workflow/organization addresses, as well as purpose associated with the address, could finally support digital communication mechanisms as alternatives to fax communications and alleviate any unsolicited communication to provider addresses. 

Payer aggregation is perhaps the most costly challenge – addressing the high cost of collecting and verifying  directory information for this constituency could eliminate expensive and duplicative efforts at every health plan and health system that contributes directory data to multiple payers.   Each of the health plan directories are “multi-contributor” in their own right, each with copies of information collected from the practices and health systems in their networks that overlap substantially with the contributions of others.  Several significant multi-payer efforts to rationalize this process have had only modest success because much of the data that needs to be aggregated is seen as proprietary and so is not a part of the consolidation effort. In addition, the attempts to disambiguate, validate, and verify the limited scope of just the demographics of organizations and of individual providers in these models are challenged by the high cost of periodically reaching out to providers to validate that provided information is still accurate.  Provider organizations continue to see substantial reconfiguration caused by acquisitions and other organizational changes.  Requiring attestation periodically to a centralized source that all payers could subscribe to would allow for a single source of truth and eliminate duplicate efforts for both providers and payers. 

Each of the currently available directory systems, whether they are deployed by government or the private sector, are different enough from each other that aggregating all of them will be a difficult task. Different nomenclatures and most importantly, the lack of well-known shared identifiers for individuals and organizations adds complexity to the aggregation problem.  For example, if a provider organization doesn’t share their NPI across directories, duplication or other adverse  aggregation may occur. Because of this challenge, we believe careful consideration of the scope of the “minimum viable product” (MVP) is important.  We propose the MVP be a critical mass of functionality, but with scope limitations around the availability of shared well-known identifiers like NPIs or employer identifier numbers for organizations that won’t be identified by NPIs.   

We also have experience with the question of how reliable “verification” against external sources of truth can be. In our experience, a high percentage of verification attempts will fail, requiring attestors to correct either their contributions or source systems such as NPPES.  No reliable conclusions about whether the source system or the contributor is incorrect can be made on an automated basis making cleanup a strictly manual process..  

As CMS moves to the step of developing an end product, DirectTrust has a desire to remain involved as an organization with an engaged community of active users of one of the largest directories of digital endpoints in the country.  We believe enhancing our aggregated Directory with appropriate interactivity with the national directory will be advantageous for the healthcare ecosystem. Likewise, we believe that the connections our HISPs provide to EHRs, particularly as we provide FHIR-based integration alternatives to the EHR community, represents a significant leverage point for CMS and the market.  The DirectTrust Directory may be the only directory that can provide leverage for the crosswalk between individuals and organizations and their chosen communication mechanism for a given use case.  DirectTrust can provide a substantial jumpstart to the ecosystem the NDH contemplates because our Directory is already integrated within the majority of EHR products in the country, and every record in our Directory includes a digital endpoint (Direct Secure Messaging address) and its individual or organizational relationship. Furthermore, we are well positioned to begin including additional digital endpoints such as a FHIR endpoint. We look forward to contributing to a national effort for a NDH.

Additional and Full Comments

Members of the DirectTrust community contributed to our full commentary and responding to the various questions the RFI specifically asked.  The responses to each of those sections is linked in our full comment letter below.

View the full comments DirectTrust submitted to CMS on the National Directory RFI here.