Compliance Tool · OIG Enforcement

Corporate Integrity Agreements Dashboard

Active HHS OIG Corporate Integrity Agreements entered 2023–2026, with violation summaries, the compliance obligations read from each agreement OIG publishes, and a plain read on what it means for your own program.

Source: OIG CIA Browser ↗  ·  Compiled as of August 30, 2026; reverified September 20, 2026 against OIG's own page for every agreement: all 68 effective dates, term lengths and Active statuses agree, and 66 of the 68 agreements were read in full: 65 as text PDFs, and one scanned image rasterised at 300 dpi and put through OCR on 2026-09-21, which confirmed its Claims Review obligation and five year term from the agreement's own words, and whose pages were then rendered and read so that the sections its obligations sit in could be cited; the OCR reading order, not the scan, is what had made those numbers unreadable. OIG publishes no agreement document for the remaining two, and none was found elsewhere, so their provisions were not read and no claim here rests on them. Their OIG category and report type were checked against OIG's own detail pages, which is a different thing from reading the agreement. Read the Active status as OIG's listing, not as a term calculation. Every record here is one OIG lists as active, and OIG's browse listing lags nominal term expiry, so agreements keep their Active status past the date their stated term runs out. Five of the 68 were past that date on September 20, 2026, the most recent being Connex Family Services LLC, whose three-year term from September 15, 2023 elapsed on September 15, 2026 and which OIG still listed as Active when checked that day. That count rises on its own as terms run out between refreshes, so treat a record past its own effective date plus term as one OIG has not yet moved, and confirm current status on the linked OIG page before relying on it. Settlement figures follow the DOJ press release, not OIG's Settlement Amount field. Those two disagree for eight records, so the field is not a check on this table: for three of them OIG states one component of a larger resolution, for two it truncates cents DOJ reports, and for three it does not match its own linked release. 63 of the 68 figures were confirmed against the DOJ release; the other five are composites or entity shares whose basis is stated in each record's note. Rite Aid's $409,368,524 is OIG's own Settlement Amount to the dollar; what it is not is the sum of the two components DOJ announces, because DOJ rounds both. Review obligations were read from the agreements themselves. The review type shown is the review the agreement names in its Independent Review Organization section: 49 of the 52 records that name one already agreed, three named a review their agreement does not impose and were corrected, and three that said Other were filled in. Six records keep Other because their agreements impose several reviews that no single category describes. Each agreement's obligations were read out of its own sections. On 2026-09-21 all 66 readable agreements were walked section by section, and every obligation shown on a card carries the section it was read from, with the review obligation quoted in the agreement's own words. That produced 872 cited obligations, between 6 and 20 per agreement. Every section label on an obligation line has been checked back against its agreement. The 66 review citations were checked on 2026-09-21, by walking each agreement's heading tree and reading the printed page where the text layer could not settle it, and eight of the 66 named the wrong section. 872 of the 872 obligation labels have now been checked the same way, and 36 of those 872 were wrong and were corrected on 2026-09-22. 859 were found in their agreement's heading tree, and the heading's printed title was read against the obligation. Seven the walk cannot settle, because the text layer splits a heading, a label names two provisions or footnotes, or the provision sits in the definitions, were checked by finding the provision's own heading and words in the agreement's text. Community Options' six were read off its rendered pages, because it is a scan. Rite Aid's eleven were checked first, and six of them pointed elsewhere in the same agreement. Rite Aid cited its compliance officer, compliance committee, board oversight and management certification lines at III.1 to III.4 where the agreement has III.A.1 to III.A.4; its IRO engagement at III.C.2, which is Board Training, where the agreement has III.D.1.a; and its IRO independence certification at III.C.3, which is Training Records, where the agreement has III.D.3. A label of the form III.1 cannot be a section of these agreements at all, because a roman numeral takes a letter before it takes a number, and looking for that shape found fourteen more across five other records, each corrected against its own agreement. Reading the other 841 found sixteen more wrong. Eleven cited the wrong provision: an IRO engagement at Training Records or at a records-retention clause, an independence certification at a section with no such item, focus arrangements procedures at Covered Persons training, an Arrangements Review at the Anti-Kickback section that precedes it, two EyePoint policies at the compliance-program item instead of their own, and a Veloxis line that cited one of the two programs it names. Five named an Independent Review Organization in agreements that engage none: four quality-of-care agreements engage an Independent Monitor, and McKinsey's engages an Independent Compliance Expert. All sixteen are corrected. That leaves 0 of the 872 labels unchecked. A checked label says where the obligation is; it is still a summary of the obligation, not the agreement's words. The controls this dashboard suggests are separate and are not the agreement's terms. 43 of the 268 lines previously shown attributed a duty to an agreement: 31 were replaced by the cited obligation itself, and 12 were rewritten as this dashboard's own suggestion. Seven stated something their agreement does not say rather than merely overstating its scope: a prospective coding review no agreement in this set requires, a speaker-program monitoring threshold given as 2 where the agreement says 35, owner-level certifications in an agreement whose only certification is by the compliance contact and the chief executive, management certifications in an agreement that has none, a grant-expenditure review that is actually a review of Federal Financial Reports, an arrangements review called annual that runs in the first and fourth reporting periods, and owner accountability described as built into obligations that reach owners only by making them Covered Persons. An earlier word-frequency pass over the same 268 lines is what decided which to read first; word overlap settles nothing about whether a document imposes a duty, and none of the results above rest on it. 19 of these 68 are Integrity Agreements and one is a Recipient Compliance Agreement, per OIG's own Report Type field, and those impose fewer of the usual elements. Counting the cards that list each provision, out of the 66 with a list: exclusion screening 66, a review by an IRO, a monitor or a compliance expert 65, annual reports 65, training 64, notification of investigations 63, reportable events 60, record retention 54, written standards 49, a compliance officer 47, management certifications 47, a compliance committee 46, a disclosure program 40, board oversight 37, an overpayment procedure 20, a notice posted for patients 18, and a lesser compliance contact in place of an officer 9. No element but exclusion screening is on every card, which is why no card shows a generic list any more. These count the cards, not the agreements: a card lists the obligations that were located and read, so an agreement can carry a provision its card does not list  ·  68 agreements  ·  2023–2026  ·  active CIAs effective in this window

67 agreements shown
68CIAs (2023–2026)
~$2.74BTotal settlements
$650MLargest settlement (combined)
~66%False Claims Act
5 yr / 3 yrTypical agreement terms

Filters & sort

Cross-CIA takeaways for your organization

Patterns across the 2023–2026 agreements: what to audit, tighten, and watch for in your own compliance program.

Whistleblowers are your early-warning system
Most of these cases were initiated by employee, competitor, or patient whistleblowers under the False Claims Act's qui tam provisions. A disclosure program people actually trust, with genuine non-retaliation protections, is what surfaces these problems internally instead of through a qui tam complaint. The FCA's relator share gives employees a real incentive to go to the government when they can't reach leadership.
Relevant: Kinex, Advanced Urology, Modern Nuclear, Semler, Fresno Community, Matrix, and others
Medical-necessity documentation is a billing matter
Multiple CIAs involve billing for tests or procedures that were medically unnecessary or not performed. Every ordered test or procedure needs a contemporaneous clinical rationale in the record. High-frequency tests (urine drug testing, respiratory panels, wound-care procedures) are OIG audit priorities; audit your own ordering patterns against clinical benchmarks before regulators do.
Relevant: Bethany, Dr. Malone, VHS Holdings, Advanced Urology, Kinex, Patient Choice, Semler
Any financial arrangement with a referring physician requires FMV
Several CIAs involve improper financial arrangements with physicians, from above-FMV supervision fees (Modern Nuclear) to physician bonuses for referrals (Fresno) to kickbacks disguised as marketing fees (Patient Choice). All compensation to physicians who can refer (employment, co-management, supervision, hospitality, technology subsidies) must be at fair market value, commercially reasonable, and documented before execution.
Relevant: Modern Nuclear, Fresno Community Hospital, Patient Choice Labs, NUWAY
Risk-adjustment coding must match the clinical record
Matrix Medical Network ($36.5M) submitted diagnosis codes to Medicare Advantage where the assessment forms didn't contain enough clinical information to support them. For any MA plan, ACO, or value-based contract where diagnosis coding affects payment, each submitted code must be supported by the contemporaneous record. Retrospective coding enhancements without documentation are a known OIG and DOJ target.
Relevant: Community Care Health Network (Matrix)
Quality-of-care failures are also FCA violations
Several nursing-home and disability-services CIAs show that accepting Medicare/Medicaid payment while providing objectively substandard care (understaffing, neglect, failure to meet minimum standards) is itself a False Claims Act violation under the "worthless services" doctrine: if the care is so deficient it has no value, the claims are false. Staffing adequacy is a billing-compliance matter, not just a quality metric.
Relevant: Center for Disability Services, Moroun/Villa, American Health Foundation, Riverpark/Avamere
Controlled-substance oversight is a federal compliance obligation
Walgreens ($300M) and Dr. Kabakibou ($3.5M) show that pharmacies and physicians face concurrent FCA and Controlled Substances Act liability. Pharmacies must maintain and act on suspicious-prescription monitoring. Physicians must be present when signing controlled-substance prescriptions and document medical necessity for every opioid order. Pre-signed prescription pads are both a CSA violation and a source of false Medicare claims.
Relevant: Walgreen Co, Kamal Kabakibou

This dashboard compiles publicly reported HHS OIG Corporate Integrity Agreements and DOJ press releases, summarized for general compliance-learning purposes. The summaries and lessons are interpretive and may simplify complex settlements. This is not legal advice. Brandon Goulter is not an attorney, and using this dashboard creates no professional advisory relationship. Settlement figures, effective dates, and obligations should be confirmed against the linked OIG pages and CIA documents, and current requirements verified with a licensed attorney before acting. Active means OIG still lists the agreement as active, which is not the same as the term having time left on it. OIG keeps an agreement listed as active until it closes it out, and the final annual report and closeout come after the nominal end date, so an effective date plus a duration can land in the past while the agreement is still listed. Four agreements here are in that position as of the compile date, and each was confirmed against its own OIG page rather than inferred from the dates. Settlement amounts are civil recoveries except where a case note says otherwise; the largest-settlement figure ($650M, McKinsey) is a combined criminal and civil resolution, of which $323M was the civil False Claims Act portion. Agreement status was verified against the OIG CIA directory as of September 20, 2026; status can change as OIG closes agreements following final-report review. Effective dates, term lengths, listed status, settlement figures and the named review obligation were checked against OIG's page for all 68 and against the agreement itself for the 66 that are readable, most recently on 2026-09-21. Each card separates two things that were previously mixed. What this agreement requires was read out of that agreement's own text, and every line carries the section it was read from. 66 of the 68 cards have had every reference checked against the agreement's own headings and say so; 872 of the 872 references have been checked in all. The other two cards list no obligations, because OIG publishes no agreement for them. The review obligation is different: for all 66 readable agreements it is quoted in the agreement's own words and its citation has been checked against the headings above the sentence it quotes. It is not a complete list of the agreement's terms. Controls this dashboard suggests is this dashboard's reading of the conduct, written here, taken from no agreement and required by none. Case summaries are likewise this dashboard's reading. For the two matters where OIG publishes no agreement document, the card says so and claims nothing about their provisions. Confirm any obligation against the linked agreement before relying on it.

Privacy preferences

Where enabled for this site, Cloudflare Web Analytics can send page-use and performance measurements to Cloudflare. This panel provides information and does not turn analytics on or off.

This panel stores no tracking choice and does not change your tool saves, screening, resets, reports or exports. Other sites have separate settings.

Read Privacy & Cookies