The days you served but never got paid for.
Reeve reads adult day authorizations, attendance, claims, and remittances today.
Its claims-side detectors read remittance and claim data, so they work here now: denied balances still inside the filing window, adjusted and written-off balances, service-code mismatches, units the payer cut back, and paid lines compared against the state's published fee schedule in the 22 states where Reeve has read one. What it will not do is guess at a dollar on your delivered-versus-billed gap. It hands you the rows and the units and says so on the report, rather than returning a quiet zero. You run the program. Your staff sign the roster, the vans roll, the day happens. Then somewhere between the authorization, the attendance sheet, and the claim, units go missing and nobody catches it. Reeve is a read-only margin engine that reads your own billing export and shows you exactly where, with the record attached. Reeve reads. It does not write, does not file a claim, does not touch your money.
We price the claims side. Your delivered-versus-billed gap comes back as records and units, priced against your own rate sheet.
Denied balances still inside the filing window, adjusted balances, written-off balances, service-code mismatches, units the payer cut back, and paid lines compared against the state's published fee schedule in the 22 states where Reeve has read one, with the citation on the finding. These read remittance and claim data, so they do not care how the service was delivered. On an adult day book they run today and carry defensible dollars.
Reeve's unit and duration logic is calibrated for home care units, and a per-diem or tiered line does not convert cleanly into them. The per diem and half day codes on a day services book fall outside it entirely: T2020, S5101, and S5102 pay by the day or the session, and Reeve holds no verified way to turn a day into an authorized unit. Those rows come back as attended days, missing claims, and open authorizations, with no dollar attached to any of them. The report says so on its face instead of quietly returning a zero.
Rate variance, clawback exposure, lapsed authorizations, and rows whose rate had to be imputed all carry a real dollar figure on the report, marked as not counted toward recoverable. Reeve puts a number on them. It does not tell you that number is money you can go collect.
One detector and no others: denied balances still resubmittable inside the filing window. Nothing else reaches the headline figure on this book. Everything else is reported beside that figure, never inside it.
Denied-and-resubmittable balances are dollarized only where Reeve has read the timely-filing window from a primary source, which is 45 of the fifty states and the District of Columbia. In the 6 it has not read, AK, OR, TN, WV, WI and DC, the balance still comes back with the record attached and no dollar on it. Elsewhere those rows are surfaced at their real amount and routed to indeterminate, with no dollar figure attached. Reeve's unit and duration pricing logic is calibrated against Medicaid home care, so on an adult day book it prices the claims-side columns and reports the rest as records and units against your own rate sheet. Reeve is read-only: it does not file, does not appeal, does not write to an EMR, does not move money.
Every day of service you provide should move cleanly through four columns. In day habilitation and adult day programs, the money leaks in the gaps between them.
The state Medicaid program or the managed care plan approves a member for a set number of units or days under an HCBS waiver. Day habilitation is usually authorized in per diem days, billed as T2020, or in 15 minute units, T2021. Adult day services run on S5100 per 15 minutes unless the state flips it above a threshold, S5101 per half day, and S5102 per diem. S5100 bills in 15 minute units, and in at least one state it flips to a daily unit above four hours in a day. Reeve reads that flip from the state's own rule and says so on the line. The authorization is your ceiling. Every unit under it that you actually delivered is money you earned.
Your attendance roster and, where your state requires it, your EVV or check-in record show who actually attended and for how long. This is the true count of service delivered.
What your EMR or clearinghouse actually sent to the payer, on the right code, with the right modifier, at the right waiver rate.
What the state or the MCO actually remitted, after denials, partial pays, and adjustments.
Reeve lines up all four columns on your own export and shows you every row where they disagree. The gaps are where earned margin sits.
- 01Attended but never billed. A member is on the roster, the day was delivered, but it never converted into a T2020 or S5102 claim. A handoff dropped it. This is the most common and the most painful, because the service already cost you staff and space.
- 02Authorized units left on the table. The plan approved 20 days a month, the member attended 20, but only 17 billed. Three per diem days quietly evaporate every month, member after member.
- 03Rate or modifier mismatch against the waiver fee schedule. The claim went out on a rate or modifier that does not match the current waiver schedule for that program and setting. It pays, but it pays wrong, or it denies for a reason nobody reads.
- 04Unit rounding and partial-day errors. A half day, S5101, billed where a full per diem was delivered, or 15 minute units, T2021 or S5100 where the state holds it at 15 minutes, rounded down against the actual attended time.
- 05Timely-filing expirations. Medicaid filing windows vary widely by state and by managed care plan, commonly ranging from about 90 days to 365 days from the date of service. A claim that sits in a worklist past that window is gone for good. The clock is the quietest thief in this business.
- 06Denials that were never reworked. A denial for a missing authorization reference or a documentation gap that could have been corrected and resubmitted inside the window, but got filed and forgotten.
Reeve's claims-side detectors run on an adult day book today and produce real dollars from your own export. The per diem and half day side is different: a day is not a unit, and Reeve has no verified conversion between them, so it returns those rows as counts and records with no dollar attached. Any dollar figure on this page is illustrative and synthetic.
14 per diem day habilitation days across 6 members appear on the attendance roster with no matching T2020 claim in the billing column. Reeve lists every one, with the member coded reference, the date, and the roster reference. It attaches no dollar figure, because pricing a per diem day would mean converting a day into a unit on an assumption Reeve cannot defend. The rate is on your own waiver schedule.
Across one closed month, adult day authorizations total 340 per diem days, attendance shows 331 delivered, billing shows 309. Reeve reports that 22 day gap as days sitting inside a live authorization, and shows the authorization it sits inside. It reports days, not dollars.
28 day habilitation claims billed on a 15 minute unit code with a modifier that does not match the current waiver fee schedule for that setting. They are at risk of a downstream denial or a partial pay. This is the one finding on this page carrying a figure, because T2021 bills in 15 minute units and Reeve prices units, unlike the per diem and half day codes named above it. Illustrative exposure at the corrected rate is about $2,100, and it is synthetic. Reeve flags the mismatch and the correct schedule reference; your biller decides and files.
9 delivered but unbilled adult day dates fall within 15 days of an illustrative 90 day filing deadline for the plan. Reeve sorts these to the top, because the clock is what makes them urgent rather than the size of the figure. On per diem dates the finding is the count and the deadline, and no value is quoted against it.
6 T2020 claims denied for a missing authorization reference, still inside the resubmission window, never reworked. Where the remittance carries the balance, Reeve prices it from the remittance itself rather than from a rate it inferred. Where the remittance does not, the row comes back as a claim to rework and not as a number.
Why an annual audit or the EMR alone misses this.
Your EMR is very good at the job it was built for: scheduling the day, tracking attendance, and dropping a claim. What it does not do is stand outside itself and ask whether every authorized and delivered day actually turned into a paid dollar. That is a four-column reconciliation, and most EMRs only ever show you one or two columns at a time.
The authorization lives in one screen, attendance in another, the claim in a third, and the remittance in a fourth. Nobody sits with all four side by side, member by member, day by day. The gaps hide in the space between screens.
An annual audit misses it for a simpler reason. By the time an audit runs, the filing windows have already closed, commonly about 90 to 365 days from the date of service. A finding you discover in an October audit about a January date of service is a finding you can no longer act on. The money is real, and it is already gone.
Reeve does the one thing neither the EMR nor the annual audit does. It puts authorized, delivered, billed, and paid in four columns next to each other, on your own data, while there is still time to file.
What you may have to give back, found while it is still fixable.
Money you can collect and money you may have to return are two different numbers, and Reeve never adds them together. Exposure is the second one, and Reeve surfaces it before the payer finds it.
- 01
Care delivered past what the authorization approved, with the authorization and the delivered record side by side.
- 02
Service and attendance records that failed the payer's visit verification system, flagged with the reason they failed.
- 03
Visits and days missing a signature or a required timestamp.
- 04
Conflicts inside one person's own delivery day: two visits by the same person whose recorded times overlap, and hours that cannot both be true.
Exposure never enters the recoverable total. It sits on its own line, with its own figure, attached to the record it came from. What you do about it is yours to decide.
Plain answers, on the record.
What Reeve looks at, what happens to your file, and how it reads an adult day book today.
Collect is money you never captured. Cover is money a payer can still take back.
Reeve reports the two separately and never adds them together, because only one of them is yours to go and get. The Margin Review reads both on your own export and costs nothing.
One pass over your own export, in your browser. The findings are yours to keep, with no obligation.
The recovery lane. Care you delivered and never billed, units short of what was authorized, lines paid under the published rate.
Everything in Collect, plus the exposure lane. Retired codes, authorizations at the end of their period, care delivered past what was approved.
Month to month, no annual contract. Read-only in every tier. Run the free review.
Run it on your book. Priced where we can defend it.
Export one file from the system you already use. Reeve reads it in your browser, replaces every name with a coded reference before any math, and hands back a ranked worklist. The claims-side findings carry dollars. The delivered-versus-billed rows are shown and left unpriced, and the report tells you which is which on its face.
- 01
Read-only. Reeve looks at your billing and does not write to it, does not file a claim, does not move money.
- 02
Names coded on your machine. Names and identifiers are replaced with a coded reference unique to your agency locally, in your own browser, before anything is analyzed.
- 03
Reeve does not receive your raw file. You keep your data.
- 04
Honest by default. Every dollar you see comes from your own export, measured on closed periods.
Any dollar figures in examples are illustrative and synthetic, method rather than results. Reeve never invents a rate; dollars come from the rate on your own export or the rate your own remittances actually paid most often.