Skip to content

Local market data

Where our local numbers come from

Every number on our city pages, such as SEO for dentists in Houston, comes from public U.S. government data. Here is what we use, how we count, and the rules we follow before we call one place higher or lower than another.

Sources

All seven are works of the U.S. government in the public domain. We publish totals only, never the name or NPI of an individual provider.

How we count providers

  • We count active National Provider Identifier (NPI) registrations of individual clinicians by their primary specialty and primary practice address, from the full monthly NPPES file published by CMS.
  • A registration is not a license. NPIs are rarely deactivated when someone retires or moves, so a count includes some clinicians who no longer practice there. Read our counts as registrations, not a headcount.
  • A provider counts for a city when their practice address has that city as its mailing city, or sits in a ZIP code that lies mostly inside city limits under a neighborhood name (so Brooklyn counts for New York, and Van Nuys for Los Angeles), but not a suburb that is a city of its own.
  • Metro areas follow the federal Core Based Statistical Area definitions. A provider is placed in a county by their mailing city when that city's county is one the ZIP code actually covers, and otherwise by the county holding most of the ZIP's land.
  • There is no registry or specialty code for med spas, so we never publish a med spa count. Med spa pages show dermatologists and plastic surgeons instead, labeled as such.

How we compare places

  • Ratios such as “one dentist per 1,440 residents” use metro, state and U.S. populations, never city limits alone: offices cluster in central cities, which would overstate how many providers serve each city resident. Ratios are rounded to the nearest 10.
  • We say “higher” or “lower” only when the gap is real for that source: Census survey estimates whose 90% margins of error don't overlap; CDC and cancer rates whose 95% confidence interval excludes the comparison; counts and ratios that differ by at least 10%. Otherwise we say “about the same”.
  • CDC PLACES city figures are model-based, so we compare them only with model-based state averages, never with the CDC's directly surveyed state numbers.
  • Washington, DC is both a city and a state-level area, so its pages compare with the U.S. instead.

What we leave out

  • Wages whose estimate is too imprecise (BLS relative standard error above 10% for the mean wage or 30% for employment). Where a metro estimate fails, we show a dash.
  • Census estimates with a coefficient of variation above 30%.
  • Figures with a known upstream defect: the 2022 Mississippi dental-visit measure, for example, is far outside its historical range, so we don't show it.
  • Cancer rates the NCI suppresses for small numbers, and per-resident ratios built on fewer than 10 providers.

A value we can't stand behind is left out, never replaced with an estimate of our own.

How often it's updated

The provider counts on our pages are from the NPPES file as of Oct 11, 2026. We refresh them from the monthly file; the other sources publish once a year, and we update when a new release comes out. If you spot a number that looks wrong, email support@sirvo.ai and we'll check it.