How to Check OIG Exclusions

Checking one person takes a minute. Checking a hundred people every month, keeping a record of it, and handling the false positives that common names produce is the part that turns into work. This guide covers both, and is honest about which step the public data cannot finish for you.

Three ways to check

1. OIG's own search

exclusions.oig.hhs.gov is the authoritative interface, maintained by the agency that imposes the exclusions. It is free, it needs no account, and it is the place to confirm anything you are about to act on. It searches one name at a time and offers a five-at-a-time option, so it does not scale to a roster.

2. The downloadable file

OIG publishes the complete LEIE as a CSV every month. If you have somebody comfortable with a spreadsheet, this is a legitimate way to screen a roster: download it, match your list against it, keep the file as evidence of what you checked. The catches are that matching is harder than it looks — names differ between systems constantly — and that a spreadsheet is a poor audit record a year later.

3. A screening service

Which is what NPISignal is. The value is not that it can search — anybody can search — it is that it screens the whole roster on a schedule, explains each potential match, keeps every decision, and can reproduce what a run looked like months afterwards. Its free exclusion search checks one party at a time with no account.

Checking one person, properly

1. Search the surname first, on its own. A search that includes a first name will miss a record filed under a middle name, a maiden name or an initial.
2. Compare every field you have against every result: first name, middle name or initial, date of birth, state, and specialty. Two or three agreeing fields is a lead; a name alone is not.
3. If the person has an NPI, check whether any result carries the same one. An NPI agreement is an identifier match rather than a resemblance, and it is the strongest signal the public data offers.
4. If nothing matches, record what you searched, when, and against which published version of the file. 'We checked' is not a record; 'we checked the August 10 file on August 22 and found no potential match' is.
5. If something matches, do not stop at the search screen. Go to step four below.

What to do with a possible match

This is the step most guides skip, and it is the one that actually matters. Common names produce matches constantly — the real LEIE contains hundreds of people sharing each of the most common American surnames. A match is a lead, not a finding.

1. Gather the identity fields you hold and compare them to the record side by side. A date of birth that disagrees is strong evidence of two different people; a date of birth that agrees exactly is the strongest non-identifier signal available.
2. Check whether the exclusion record carries an NPI, and whether it is your person's NPI.
3. Look at the specialty and the state on the record. A pharmacy technician in Ohio is not your physician in Arizona, however identical the name.
4. If the identity is still ambiguous after all of that, use OIG's own verification process. The public file cannot resolve it — see below.
5. Record the decision and the reasoning, whichever way it goes. A dismissal with no reason attached is indistinguishable from not having looked.

Checking a roster

The mechanics are the same; the difficulty is volume and continuity. Three things make the difference between a roster screening that survives scrutiny and one that does not.

• Screen everybody in scope, not just billing clinicians. The payment prohibition reaches anyone whose work contributes to an item or service paid for with federal healthcare funds.
• Collect a date of birth where you can. It is the single field that turns a common-name match from an unresolvable question into a decision. Without one, a screening of a roster full of common surnames produces leads nobody can close.
• Keep dismissals attached to the facts they were based on. A match you cleared last year should not reappear every month — but it should reappear if the identity information changes, because that is new information.

What a clean result actually means

No potential match was identified in the data the screener held, as of a stated date. That is the whole claim. It is not a determination of eligibility, it does not establish licensure, and it says nothing about an exclusion imposed after the file you searched was published. Any tool that phrases a clean result as 'compliant' or 'cleared' is overstating what it did.

Exclusions Database — HHS Office of Inspector GeneralThe authoritative search interface. Confirm anything you are about to act on here.