HRSA NHSC & NURSE Corps Clinician Workforce Site Monitor

Pull the official list of HRSA-approved clinical sites eligible for federal loan-repayment and scholarship programs — NHSC (National Health Service Corps), NURSE Corps, and STAR LRP — as clean, structured rows. Every site comes with its type, approval status, HPSA shortage scores, phone, email, full address, county FIPS, congressional district, and lat/lon, plus an optional state / program / site-type rollup.

Data source: HRSA Bureau of Health Workforce public ArcGIS server (gisportal.hrsa.gov → BHW/BHW_Sites). Keyless, public, refreshed by HRSA.

Who it's for

• Clinician & locum staffing agencies — NHSC/NURSE Corps sites are loan-repayment-eligible employers: warm recruiting leads for physicians, NPs, PAs, dentists, and behavioral-health clinicians who want their loans paid.
• Health-workforce planners & researchers — model provider supply, shortage (HPSA) exposure, and rural access by state, county, or congressional district.
• FQHC / RHC vendors & GPOs — a targeted list of safety-net facilities with contact info for outreach.
• Grant writers & policy teams — quantify program footprint and shortage severity across geographies.

Example input

{
  "state": "California",
  "program": "NHSC",
  "siteType": "FQHC",
  "ruralOnly": false,
  "minHpsaScore": 15,
  "includeSummary": true,
  "maxResults": 5000
}

Leave everything blank to pull all NHSC sites nationwide. Set program to NurseCorps or STAR for those program lists. state accepts a full name ("California") or a 2-letter code ("CA", "PR", "GU").

Output fields

One row per site (record_type: "site"): site_id, objectid, site_name, program, site_type, site_status, cross-program flags (nhsc_site_type, nhsc_status, nursecorps_site_type, nursecorps_status, star_site_type, star_status), address, city, county, county_fips, state, state_abbrev, zip, phone, fax, email, hpsa_primary_care, hpsa_dental, hpsa_mental_health, rural_status, region, cong_district_code, cong_district_name, census_tract, latitude, longitude, result_truncated, source_status.

result_truncated is true when maxResults cut the roster short while HRSA still had matching sites, false when the roster is complete for your filters, and null if it was not evaluated. A partial roster is never presented as a complete one.

source_status is "ok" when the row came from an HRSA fetch that completed every page it needed and passed the live state-verification gate, and null when that was not established. It is never "ok" by default — a run that could not finish fetching fails instead of emitting rows.

With includeSummary: true, one extra record_type: "summary" row is appended with total_sites, sites_with_email, sites_with_phone, rural_sites, and counts by_state, by_site_type, and by_site_status, plus result_truncated, pages_fetched, state_predicate_mode, source_status, fetch_elapsed_ms, fetch_budget_ms, and drift_gate_status.

fetch_elapsed_ms / fetch_budget_ms report how much of the run's wall-clock retrieval budget HRSA actually used — useful for spotting an HRSA slow phase. drift_gate_status is "pass" when a full state name was rewritten to a state code and every returned row was verified to really be that state, or "not_applicable" when a 2-letter code (or no state) was given and there was nothing to verify. It is never "pass" for a check that did not run.

Reliability

HRSA's GeoPortal is intermittently slow: the first page of a large state roster typically answers in under 4 seconds, but a continuation page has been measured anywhere between 29 and 138 seconds. This Actor gives the whole retrieval a single wall-clock budget and clamps every individual request to whatever is genuinely left of it, so a slow HRSA phase can never make the run overrun.

If any page cannot be retrieved inside that budget, the run fails and emits nothing rather than publishing a partial roster that reads as complete — so you are never billed for a short answer that looks whole. The failure message distinguishes an HRSA outage (retry shortly, or narrow the query so fewer pages are needed) from a genuine source change that this Actor must be updated for.

simulateOutage is an optional diagnostic input, defaulting to "none" (a true no-op). Set it to "all" or "page2" to prove the failure behaviour on demand: the run fails, emits no rows, and bills nothing.

Use as an MCP tool

This Actor is callable by AI agents (Claude, Cursor, etc.) via mcp.apify.com. The field-level output descriptions make it a clean, chainable tool — e.g. "list active NHSC FQHC sites in Texas with a primary-care HPSA score above 18 and return their phone and email."

FAQ

Where does the data come from? HRSA's public Bureau of Health Workforce ArcGIS service (BHW/BHW_Sites). No API key required.

What's the difference between NHSC, NURSE Corps, and STAR? They are three HRSA loan-repayment/scholarship programs; each has its own list of approved sites. Pick one via the program input. Each returned row also carries the other programs' type/status flags so you can cross-reference.

What is an HPSA score? A Health Professional Shortage Area score (0-26) rates how severe the provider shortage is at that location — higher means more underserved and higher loan-repayment priority. Filter with minHpsaScore.

How do I get a downloadable clinician-recruiting lead list? Filter by state and siteType, run, then export the dataset as CSV/Excel/JSON.

Does it cover territories? Yes — Puerto Rico, Guam, and the US Virgin Islands appear when present in HRSA's data.

Notes & caveats

• Coverage = whatever HRSA currently publishes in BHW/BHW_Sites; site rosters change as programs re-approve sites.
• This layer does not expose a per-site "date added" field, so the Actor ships a program/state/type/status rollup rather than a date-based new-since delta. For change tracking, schedule the Actor and diff dataset snapshots on site_id.
• Email is only present when HRSA publishes it (not every site).