FDA Clinical Investigator Disqualification & Debarment Check

Bulk-screen clinical investigators, sub-investigators and study staff against the two FDA registers you have to clear before putting someone on a trial or naming them in a drug application:

1. Clinical Investigators — Disqualification Proceedings (21 CFR 312.70) — 232 entries
2. FDA Debarment List, drug product applications (21 U.S.C. 335a) — 149 named individuals, 0 firms at present

Give it a list of names. It returns, for each one, FDA's verbatim status and a plain answer to the only question that matters: is this person barred right now?

---

What this does that a name-match does not

Being on the FDA disqualification register does not mean you are disqualified. Measured against the live register on 2026-07-30:

FDA status	Entries	Barred right now?
Disqualified	144	Yes
Totally Restricted	14	Yes
Restricted	23	Only under FDA's conditions
Restrictions Removed	27	No — cleared
Not Disqualified	8	No — cleared
Reinstated	7	No — cleared
Adequate Assurances (and variants)	9	No — cleared

74 of 232 entries — just under a third — are listed but not disqualified. The register exists to record proceedings, including the ones FDA closed in the investigator's favour or later lifted. Any tool that answers "found on the FDA disqualification list" is wrong for all 74, in the direction that costs a real investigator a contract.

This Actor returns currently_barred as barred, restricted, not_barred or unclear, always alongside the exact FDA status string it came from, so you can see the reasoning rather than trust it.

What it does not cover — read this before you rely on a clear result

It checks these two FDA registers only. It does not check HHS-OIG LEIE, GSA SAM exclusions, state Medicaid exclusion lists, DEA actions, or state medical boards. FDA debarment for device applications is a separate list and is not included either.

A NO_MATCH means "not on these two FDA registers" — not "cleared generally." Every output item repeats this in a coverage_note field so a row can never be read as broader than it is.

Input

{
  "names": [
    "Jane Q. Doe",
    "Roe, John A.",
    { "name": "Alex Smith", "recordId": "emp-4417" }
  ],
  "maxNames": 100
}

Either name order works — First M. Last or Last, First M. Degrees and generational suffixes (MD, DO, PhD, DVM, Jr, III) are ignored, and accents and hyphens are normalised, so Anzures-Camarena, María and Maria Anzures Camarena reach the same entry. Pass objects with a recordId to join the output straight back onto your own roster; it is echoed on every row, including errors.

Output

One row per name.

{
  "input_name": "Jane Q. Doe",
  "recordId": null,
  "status": "ok",
  "result": "LISTED_BUT_NOT_BARRED",
  "match_count": 1,
  "matches": [
    {
      "register": "clinical-investigator-disqualification",
      "name": "Doe, Jane Q., MD",
      "center": "CDER",
      "status": "Restrictions Removed",
      "date_of_status": "06/23/1986",
      "date_nidpoe_issued": "08/06/1980",
      "date_nooh_issued": "11/12/1981",
      "match_confidence": "exact",
      "currently_barred": "not_barred",
      "what_this_means": "Restrictions were lifted; eligible."
    }
  ],
  "same_surname_not_matched": [],
  "same_surname_not_matched_count": 0
}

result is one of BARRED, RESTRICTED, REVIEW, LISTED_BUT_NOT_BARRED, NO_MATCH, and status is ok or error.

Namesakes. A surname on its own is never treated as a match. Entries sharing the surname but not the given name are returned separately in same_surname_not_matched so a screener can eyeball them instead of never seeing them. match_confidence is exact when the given name matches in full and initial when only an initial was available on one side.

Why you can trust a clear result

Both sources are HTML pages, so the real risk is not an outage — it is the page quietly changing shape and the Actor confidently reporting an empty register. Four guards run before anything is charged:

1. The disqualification page states its own record count (No. of Records Found: 232). The parsed row count must equal it exactly.
2. Both tables must present their expected column headers verbatim.
3. Both registers must clear a floor row count.
4. A name whose surname cannot be read is returned as status: "error" with the guidance treat it as UNCHECKED, not as clear — and is not billed.

If any guard trips, the run fails loudly and costs you nothing, rather than returning a confident all-clear.

Pricing

• $0.05 per run — charged once, after both registers download and pass their integrity checks. A failed or unreachable source is not charged.
• $0.25 per name screened — charged only for names actually answered.

Background-screening vendors sell a healthcare-sanctions check at $12.50 per name, and the published coverage lists in that market routinely include the FDA Debarment List while omitting the disqualification register entirely. 200 investigators a month costs $50 here.

Source and licence

U.S. Food and Drug Administration public registers, retrieved live on every run:

• accessdata.fda.gov/scripts/SDA/sdNavigation.cfm?sd=clinicalinvestigatorsdisqualificationproceedings
• fda.gov/inspections-compliance-enforcement-and-criminal-investigations/compliance-actions-and-activities/fda-debarment-list-drug-product-applications

U.S. federal government works, public domain under 17 U.S.C. §105. No API key, no account, no CAPTCHA; robots.txt permits both paths.

This Actor reports what the registers say. It is not legal advice, and it is not a substitute for the debarment certification required by 21 U.S.C. 335a(k)(1).