Provider universe
We start from NPPES (National Provider Identifier registry) and keep only individuals (entity type 1) whose primary taxonomy belongs to the Dental Providers grouping & Dentist classification, with no deactivation_date. As of the last refresh, that's ~373K active U.S. dentists.
Sub-specialty resolution
NPPES exposes a 16-classification dental sub-specialty taxonomy (orthodontist, pediatric dentist, oral & maxillofacial surgeon, endodontist, periodontist, etc.). We use the primary taxonomy specialization first.
For providers whose primary taxonomy is the generic 122300000X (Dentist) — which is the case for ~35% of the universe — we attempt a fallback resolve from the credential string with regex (orthodontist, OMS, pedo, etc.) and an LLM transform with deterministic gating. If neither resolves cleanly, the provider is labeled General Dentist.
HPSA dental flag
HRSA publishes Health Professional Shortage Area designations for dental health (alongside primary care and mental health). We cross-reference each provider's practice state and county against the active dental HPSA list. If at least one designation matches, we flag is_hpsa_dental = true and surface the HPSA score (0–26).
HRSA refreshes monthly; our HPSA snapshot lags by at most 30 days.
Content score (the ranking input)
For every active individual dentist, we compute a deterministic content_score as the sum of the components below. Higher score = richer profile content for SEO purposes; this is not a quality-of-care score and should not be read as one.
- MIPS final score — 0–100, raw, when a CMS Quality Payment Program score is on file.
- Medicare beneficiary count, scaled by 1.6 and capped at 100.
- Medicare service count, scaled by 30 and capped at 50.
- Open Payments total, log-scaled (× 7) and capped at 50.
- Medicare-enrolled: +20.
- Has CMS data: +20.
- Hospital affiliation (count > 0): +15.
- Multi-location practice: +10.
- HPSA dental service: +15. We deliberately reward shortage-area practice — that's an editorial choice we believe the site's YMYL position justifies.
- Profile completeness: credential (+5), graduation year (+3), phone (+3), address (+3).
Public top-10 selection
For every (state, dental_subspecialty_slug) partition, we promote the top 10 providers — ordered by content_score DESC NULLS LAST, npi — to is_public = true. These are the only providers exposed through the public site. OIG-excluded providers are filtered out before ranking.
Re-rank is implemented as a forward-only SQL migration db/migrations/000N_*.sql. We don't hand-edit ranks on production; if a re-rank ships, the migration is reviewable in git.
Profile completeness bucket
We measure how filled-in the federal NPPES record is across 12 fields: full name, credential, address line, phone, primary taxonomy code, graduation year, dental school, gender, NPI enumeration date, last update date, entity type, and state. We count non-null/non-empty values, divide by 12, then bucket the share:
- High — 80% or more of the 12 fields filled.
- Medium — 60% to 79% filled.
- Sparse — under 60% filled.
This measures NPPES record fullness, not clinical quality. A sparse record can still belong to an excellent clinician — it just means less verifiable information is available through federal channels.
MIPS final score (CMS Quality Payment Program)
The Merit-based Incentive Payment System scores Medicare-billing clinicians on a 0–100 scale across four categories: quality, cost, improvement activities, and promoting interoperability. CMS publishes the final score per cycle.
On provider profiles we surface the raw MIPS final score plus its national percentile against all dentists who report MIPS. State and specialty pages show the median, p25, and p75 across reporting dentists with their sample size.
Coverage caveat: Only ~10–20% of US dentists report MIPS — primarily those with enough Medicare billing volume to meet the threshold. Absence of a MIPS score is not a negative signal.
Medicare patient volume
From the CMS Medicare Provider Utilization & Payment Public Use File. We bucket on beneficiary count using PUF percentile cutoffs:
- Low volume — fewer than 50 beneficiaries.
- Moderate volume — 50 to 249.
- High volume — 250 or more.
Most dental work is private-pay or Medicaid; Medicare covers limited dental services. Volume here reflects only Medicare-billable work, not total practice volume.
Industry payments (CMS Open Payments)
The federal Sunshine Act requires drug and device manufacturers to disclose transfers of value (consulting fees, food, travel, research grants) to physicians and dentists. CMS publishes Open Payments data annually.
We display total disclosed payments and the count of distinct payments per provider, with a link to the official Open Payments record. We do not bucket payments as "low" or "high" — disclosure is a transparency mechanism mandated by federal law, not a quality flag. Implant- and surgery-heavy specialties trend higher because of device manufacturer relationships.
HPSA dental score (0–26)
HRSA scores each Dental Health Professional Shortage Area on a 0–26 scale. Higher scores indicate greater unmet population need and higher federal funding priority. We surface the raw score plus a priority bucket:
- Lower priority — 0–12.
- Moderate need — 13–19.
- High need — 20–26.
The score reflects the population's dental access need, not the provider's quality. Practicing in a high-scoring HPSA is an access-equity signal we believe deserves visibility on a dental-directory site.
Indexed dentists per 100K residents
Computed as (indexed_count / state_population) × 100,000 using U.S. Census Bureau Vintage 2024 state population estimates as of 2024-07-01. The number reflects the public-tier subset surfaced on this site, not all active dentists in the state. The national all-active-dentists baseline runs around 116 per 100K (~389K active dentists / ~334M residents).
Refresh cadence
- NPPES — quarterly
- HRSA HPSA — monthly
- CMS Open Payments — annual (CMS publishes once a year)
- OIG Exclusions — monthly
- CMS Quality Payment Program (MIPS) — annual
See Sources for upstream URLs and license terms.
What you'll see and what you won't
We surface only signals we can defensibly tie to a public record. That's why we don't display patient star ratings, prices, wait times, or insurance acceptance: there's no federal source of truth for any of those. For Medicaid acceptance specifically, see our caveat copy.