New York's Adult Day Care Audit: What the Receipts Prove—and What They Don't
A state audit identified major oversight failures and questionable Medicaid payments. Jason Wisdom's follow-up research asks a narrower question: which original records can the public actually verify?
New York's Office of the State Comptroller published audit 2023-S-21 on February 6, 2026. It examined the Department of Health's oversight of managed long-term care plans and social adult day care (SADC) services during January 2019–October 2024.
The auditors identified more than $285 million in questionable encounter payments associated with service dates after providers had been terminated from at least one of six reviewed plan networks. The report also details $1.3 million in improper payments for services lacking supporting documentation during visits to three SADCs. These are the Comptroller's findings—not a court judgment that every payment was fraudulent, nor a demonstrated amount recoverable from specific individuals.
The story beneath the headline
The audit found that documentation standards and oversight had gaps. In one example, sign-in records did not support 7,964 of 92,969 reviewed encounter claims, representing $672,147. In another, encounter data reported 530 members at a site with a documented combined capacity of 323 on a particular date. Both findings demand scrutiny; neither alone establishes an individual's criminal intent.
Verification over narrative: A questionable payment is a signal for reconciliation. It is not automatically a finding of theft. The next proof requires service dates, contract or network status, attendance evidence, payment trails, and applicable authority.
What Jason's ACT3 research found
On November 12, Jason's ACT3 research assembled official New York Department of Health guidance on SADC compliance reviews, remediation procedures and determination letters. It documented that such letters are described in agency guidance, but the bounded search did not recover an individual facility's determination letter or the Comptroller's original audit workpaper set.
That result is classified UNKNOWN—not a claim that the records do not exist. The search was limited, and the research files did not preserve the original agency-document bytes for independent hashing.
ACT4: The 180-day response question
ACT4 examined whether a separate New York Department of Health implementation response to the February 2026 audit could be located. It distinguished a September 30, 2025 DOH letter, reproduced with responses to the draft audit in the final report, from a later post-publication report. The earlier letter cannot be treated as the 180-day response to a report issued months afterward.
The ACT4 ledger calculates a nominal 180-calendar-day mark of August 5, 2026 from the February 6 release date, while expressly reserving statutory counting and filing details. The bounded source search recovered no separately linked 180-day implementation response for report 2023-S-21. That is not evidence of nonfiling, concealment or a legal violation.
Evidence ledger: published work versus open gaps
| Claim or control | Grade | Evidence boundary |
|---|---|---|
| Official OSC audit 2023-S-21 | PASS | Public agency report and summary |
| Research file ACT3 → ACT4 parent SHA-256 | PASS | Exact parent digest reproduced from ACT3 bytes |
| DOH draft-audit comments | PASS | Published appendix, not later 180-day response |
| Individual determination letters | UNKNOWN | Not recovered in bounded search |
| Original audit workpapers / transaction lineage | UNKNOWN | Not authenticated in this corridor |
| Separate 180-day implementation response | UNKNOWN | No separately published version recovered by ACT4 search |
| Nationwide court-access / 50-state Merkle inclusion | HOLD | Outside the scope of this NY research linkage |
What would close the gap?
The next step is original-source evidence: an identifiable 180-day response, original plan-level or site-level determinations where lawfully public, and reproducible service-to-payment records. Each would need its own source URL, acquisition date, immutable bytes, digest and clear authority boundary. Until those records are independently obtained, the status stays UNKNOWN.
Read the source record
Reproducibility checkpoint
These digests identify the archived research files themselves, not raw copies of the official audit records:
ACT3 SHA-256
ACT4 SHA-256
Editorial note: This article is a human-readable publication compiled from Jason Wisdom's pre-existing ACT3/ACT4 research and independently checked official audit summary. Source research files remain unaltered. No agency was contacted and no new court or transaction records were authenticated in this publication run.