Directory coverage is not evenly spread across the professions it appears to describe. It tracks who bills Medicare.
We reloaded all 45 GB of the 2026-08-20 release and re-ran every measurement. Where-they-work records went from 7.0M to 16.5M, and coverage moved only five points because most landed on clinicians who already had one. Every profession improved, most at the bottom of the table. A hierarchy field that resolved to nothing in May now resolves completely. 233 health plans and 27 insurers appeared. Endpoint attribution went backwards, 16.9% to 14.7%.
One of the two new resource types is complete on arrival. The other carries no descriptive field at all.
We checked all 227,727 health workers the national directory lists as active in Pennsylvania, and asked which ones it says have a workplace. It is not evenly spread. 77.9% of nurse practitioners and physician assistants have one, and 1 of 12,995 pharmacy workers does. The split tracks who bills Medicare, not who provides care. Includes a county map you can zoom. Measured against the 2026-05-08 release; every profession improved at 2026-08-20 and pharmacy went from 1 to 526.
76% of the endpoints the NDH cannot name are already named by the vendors, in public files.
We checked 114,071 web addresses in the national directory of doctors. Only 19,334 say who owns them, about one in six. Nothing is broken, the names were simply never filled in. Includes a free list of the ones that work and why sixteen software companies have never filled the name in once. Measured against the 2026-05-08 release; at 2026-08-20 it is 16,262 of 110,973, and health insurers have since arrived.
16.9% of NDH FHIR endpoints can be attributed to an organization. The rest are URLs with no owner.
27 payer organizations arrived. None of them publishes an endpoint.
34.4% of US hospitals sit in nonmetro counties serving 13.8% of the population.
What arrived when the NDH went from six resource types to eight
Vendor-published endpoint files can name most of what the NDH cannot
Payer directories carry the affiliation the NDH leaves empty
The role gap is a Medicare-billing gap
Endpoint-to-organization linkage
NDH payer endpoint and organization coverage
National rural hospital baseline
Pennsylvania rural hospitals: FHIR endpoint publication and EHR concentration
State Medicaid provider-directory coverage and liveness
CEHRT-published FHIR endpoints missing from the NDH, by state
Endpoint metadata coverage vs the HTE submission spec
Practitioner phone-number reachability
Federally excluded NPIs billing Medicare Part B by HCPCS code post-exclusion
NPPES taxonomy vs Medicare-billed-specialty divergence (behavioral specialty drift)
Federally excluded NPIs whose post-exclusion Medicare Part B billing is telehealth-dominant
PECOS PROVIDER_TYPE vs NPPES NUCC taxonomy disagreement
Behavioral-health PECOS taxonomy misalignment (highest-recoupment cohort)
Multi-enrollment NPIs with conflicting state addresses
Federally excluded providers paid by Medicaid (HHS spending dataset)
Federally excluded VA-resident NPIs billing Medicare Part B (CY 2023)
Federally excluded VA-resident NPIs prescribing Medicare Part D (CY 2023)
NPPES-deactivated VA-state NPIs still billing public claims data
Federally excluded NPIs receiving industry payments (Open Payments)
DMEPOS suppliers on federal exclusion lists
Internal CMS contradiction: POS-certified providers also flagged NPPES-deactivated
SNF, hospice, HHA, hospital owners on federal exclusion lists (Stage B: NPI-keyed + facility-state demographic)
High-volume Medicare billers absent from NDH (directory completeness)
Endpoint URL validity + machine-readable share
Endpoint liveness
NPI and taxonomy correctness
Temporal staleness
Referential integrity
Duplicate detection
Social Security Numbers exposed in the NDH bulk export
Network adequacy gauge
OIG LEIE excluded providers in NDH
SAM.gov excluded providers in NDH
VA payer networks containing federally excluded providers (methodology demo)
High-risk provider cohort