LEIE Exclusion Evidence API (OIG + NPI + SAM)
Pricing
from $3.00 / 1,000 results
LEIE Exclusion Evidence API (OIG + NPI + SAM)
Source-linked OIG LEIE screening with valid-NPI exact matches, state-bound name review candidates, optional NPPES/SAM checks, and evidence hashes. No eligibility decisions.
Pricing
from $3.00 / 1,000 results
Rating
0.0
(0)
Developer
Bryan
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
0
Monthly active users
3 days ago
Last modified
Categories
Share
Healthcare Exclusion Screening Evidence — LEIE + NPI + optional SAM.gov
Screen provider and entity records against the official HHS-OIG List of Excluded Individuals/Entities (LEIE), correlate valid returned NPIs with the official CMS NPPES registry, and optionally check active SAM.gov exclusions using the caller's own API key.
This Actor is designed to produce reviewable evidence, not a black-box eligibility verdict. Every screening row records which sources were checked, which were skipped or failed, when the source was fetched, the LEIE source-file hash, deterministic identity signals, limitations, and a SHA-256 evidence fingerprint.
HHS-OIG says healthcare entities should routinely check the LEIE to reduce civil-monetary-penalty risk. The Actor supports that workflow; it is not legal advice and does not replace identity verification, licensure checks, or a compliance professional's review.
Screening decisions
screening_decision | Meaning |
|---|---|
confirmed_identifier_match | A checked exclusion source matched the caller's exact NPI. Confirm identity and review the source record before acting. |
potential_match_review_required | A normalized name matched, or SAM.gov returned a name candidate, without an exact NPI confirmation. Do not treat this as identity proof. |
no_match_on_sources_checked | No match was found on every source that the result says was successfully checked at that time. This is not a universal “clear.” |
incomplete_source_check | A requested source failed, was unavailable, or was skipped because a configured query limit was reached. |
The old matched, match_type, and matches fields remain for compatibility. New integrations
should make decisions from screening_decision, sources, and source_errors.
Screen a batch
{"mode": "screen","screenList": [{ "name": "ACME HOME HEALTH", "npi": "1234567893", "state": "NC", "ref": "vendor-118" },{ "name": "JANE DOE", "state": "TX", "ref": "provider-4471" }],"enrichNpi": true,"maxNpiLookups": 10,"includeSam": false}
Exact NPI checks are strongest. Name-only LEIE matches use normalized exact-name comparison and are always labeled for review; the Actor does not pretend an opaque fuzzy score is an identity decision. Supplied NPIs must be ten digits and pass the standard NPI check-digit test. Invalid NPIs, empty identity objects, and malformed state codes are rejected instead of being laundered into a misleading no-match.
Add SAM.gov active exclusions
SAM.gov's official Exclusions API v4 requires an API key and has role-dependent rate limits. Supply
your own key in the secret samApiKey input and opt in explicitly:
{"mode": "screen","screenList": [{ "name": "ACME HOME HEALTH", "npi": "1234567893" }],"includeSam": true,"samApiKey": "YOUR_KEY","maxSamQueries": 10}
The key is sent only to the official api.sam.gov endpoint and is never returned in dataset rows or
status messages. Rows beyond maxSamQueries are marked incomplete instead of silently omitting SAM.
Filter the LEIE source
{"mode": "filter","state": "TX","specialty": "PHARMACY","maxRecords": 100}
Filter rows include source_url, source_fetched_at, source_sha256, and a stable
record_fingerprint so scheduled runs can be compared without guessing whether a row changed.
Monitoring pattern
Save a bounded screening input as an Apify Task, schedule it at the cadence your compliance program
requires, and deliver each run through an Apify integration or webhook. Compare evidence_hash for a
query reference across runs. Do not persist customer screening lists in a developer-owned shared store;
the caller controls their Task input, run retention, and downstream system.
Portable review cases and signed webhooks
Set caseWorkflow to true in screen mode to write a CASE-BUNDLE alongside the ordinary
dataset. Each case binds a hashed caller reference to the screening evidence hash, an evidence
class, a disposition state, and a hash-chained transition history. The case file deliberately does
not retain the submitted name, NPI, or raw caller reference. Keep the bundle in your own system and
return the applicable priorCases on the next run; CivicDataForge does not use a shared developer-
owned customer-population store.
Human actions use optimistic locking. A caseActions item must supply the current case_id,
expected_case_receipt, expected_version, to_status, actor_ref, reason, and ISO at
timestamp. Stale, forged, invalid, or unaudited transitions fail closed. When the evidence hash
changes, a resolved disposition is reopened automatically with a source-monitor transition.
Optional webhookUrl and secret webhookSecret inputs send the exact bundle to a public HTTPS
endpoint. Every retry preserves one event ID and one HMAC-SHA256 signature. Private/internal targets,
redirects, embedded credentials, non-HTTPS URLs, and nonstandard ports are rejected. WEBHOOK-DELIVERY
records accepted/rejected attempts without storing the signing secret or raw transport errors.
This is a review workflow, not a legal eligibility engine. resolved_clear means a named reviewer
resolved the case for the customer-defined purpose and reason recorded in its audit chain; it never
changes the narrower meaning of no_match_on_sources_checked.
Official sources and limits
- HHS-OIG LEIE download:
https://oig.hhs.gov/exclusions/downloadables/UPDATED.csv - CMS NPPES API 2.1:
https://npiregistry.cms.hhs.gov/api-page - GSA SAM.gov Exclusions API v4:
https://open.gsa.gov/api/exclusions-api/
NPPES returns registry identity data but does not establish licensure or good standing. SAM.gov v4 returns active exclusions through the documented API. Source availability, publication cadence, and API rate limits remain controlled by the respective agencies.