The healthcare industry has become obsessed with “coverage” in the race to digitize everything. On the surface, it sounds like a straightforward metric. You’ll see claims of 90% coverage or “nationwide health system access.”
But there is a widening gap between data connection and data completeness.
If we are to move toward a future of high-functioning AI, reliable risk stratification, and true patient autonomy, we must stop measuring success only by the number of logos on a slide and start measuring it by the granularity of the data retrieved.
To understand what complete coverage really looks like, here are the four factors that matter most:
1. The Myth of the “One-and-Done” Connection
How data is stored is a fundamental flaw in current industry metrics. Many believe that connecting to a major EHR vendor or a single health system “checks the box” for all associated data.
The reality is much more fragmented:
- Onboarding Lag: Not every EHR vendor automatically enables connectivity across their entire client base. Individual provider organizations often require separate onboarding, a process that can take months.
- The Multi-System Practitioner: Doctors do not exist in a vacuum. A single practitioner often documents care across multiple locations using entirely different EHR systems.
Consider the “Dr. Plauche” scenario: If a physician documents care at four different organizations using four different systems (e.g., Evident, Athena, Medhost, and Epic), a connectivity partner that only counts “Health Systems” or “Epic access” will miss 75% of that doctor’s patient records. To achieve a longitudinal record, you cannot just connect to the “mother ship.” You must map data down to the individual NPI (National Provider Identifier) level and the specific clinics where care actually occurs.
2. The Multi-Pathway Necessity
Relying on a single network or a single API standard is not suited for a long-term strategy. True coverage requires a “redundant pathway” approach to ensure no data is left behind. A connectivity strategy must simultaneously leverage:
- Patient Access APIs: Mandated by law under ONC g(10) and CMS 9115.
- National Networks: Voluntary Participation in TEFCA, CommonWell, Carequality, CMS Aligned Networks, etc.
- Regional Expertise: Integration with State and Regional HIEs.
- Payer Data: Incorporating patient access APIs with payers.
Another nuance is that not all pathways return the same data. Different exchange methods operate at different levels of data richness. As an example, TEFCA transacts on document exchange at the USCDIv1 level, which lacks the unstructured clinical notes and care plan context that are essential for comprehensive patient understanding. In contrast, (g)(10) APIs are mandated to support USCDIv3, which includes these critical data elements.
By utilizing multiple pathways, you maximize record completeness and remove the “single point of failure” that plagues many legacy connectivity solutions.
3. Why Data Depth is the “Oxygen” for Health AI
AI and advanced analytics are now expected to identify patterns, detect drug interactions, and stratify patient risk.
AI is only as intelligent as the data it consumes. Incomplete records don’t just lead to “messy” charts; they lead to:
- Inaccurate Insights: A missed lab result or a specialist consultation not captured by a primary health system can fundamentally change a clinical recommendation.
- Erosion of Trust: Users lose faith in digital health solutions when they see glaring gaps in their own medical history.
- Compromised Intelligence: Without the “full story”, including labs, pharmacies, and social determinants, the AI cannot accurately generate recommendations, predict outcomes, or surface meaningful insights.
4. Redefining the Metric: From Percentages to Precision
The industry must move away from percentages without clear denominators. Claiming “90% coverage” is meaningless if the 10% missed contains 2/3 of a patient’s longitudinal history.
We must stop conflating “endpoint counts” with actual coverage. An endpoint is a technical connection. There’s a fundamental disconnect because patients don’t think in terms of organizations or endpoints. They think in terms of their doctor’s name or the clinic’s name where they receive care.
A patient knows “Dr. Smith at Main Street Family Practice,” they don’t know that their records might technically be behind an endpoint registered to “Regional Healthcare Consortium.” The hard work is in creating the linkages from that endpoint down to the individual NPI practitioner level and the care delivery locations, not just establishing the technical connection. Without this granular mapping, endpoint counts become meaningless to the consumer if they cannot actually find their doctor. The true measure of coverage is whether a patient can find their provider and successfully obtain their data.
The new standard for transparency should include:
- NPI-Level Mapping: Verified practitioners rather than just “affiliated systems.”
- Granular Enumeration: Listing specific clinics, ambulatory surgery centers, and retail clinics individually.
- Explicit Payer Lists: Direct transparency into which insurance feeds are live.
Moving Forward
At b.well, we have spent over a decade building a comprehensive ontology that maps the healthcare ecosystem with this exact level of precision. We provide the largest set of live connected health data because we understand that in healthcare, “mostly covered” is not enough. Our collaboration with CMS on provider directories and our transparent lookup tools for providers and health plans, which enable access to health records, are designed to set a new bar for the industry. We believe that when coverage is measured accurately—by the record, by the NPI, and by the patient—the entire industry wins.
