River
Y CombinatorBacked by Y Combinator

HealthcareFree

Utilization Review Summary and Change Log

Each day's summary built as a documented change since the last, in the reviewer's structure and length, with the clinical judgement left where it belongs.

Start here

River's utilization review summary takes the documentation a clinician has already written and puts it into the form a concurrent reviewer reads: the reviewer's structure, inside the reviewer's word limit, dated to the day it covers. It carries the clinical content forward as the clinician recorded it and adds none of its own. What it contributes is the shape, the chronology, the mapping onto the criteria the payer stated, and the discipline of saying what changed.

The templates that rank for this search all treat the summary as a status report, and that is the mistake that ends stays. A continued stay review does not ask whether the patient was acute on admission. It asks whether today differs from the discharge criteria. So a narrative that restates yesterday's story reads to a reviewer as stable, and stable is the word that closes a case. The document has to be a difference, and the difference has to be legible in one pass.

There is a second reason to send it, which has nothing to do with persuasion and everything to do with procedure. This is for inpatient and behavioral health utilization review staff, case managers, and the clinicians whose hour it currently costs each day. When a stay is denied outright and the fight moves to the written record, that is a medical necessity letter, a different document with a different job entirely.

Silence lowers the bar for denying you

The hospital utilization review condition of participation says something most staff have never read. Before deciding a continued stay is not medically necessary, the committee must consult the responsible practitioner and afford them the opportunity to present their views. If those views are presented, an adverse determination takes at least two committee members. If the practitioner concurs, or fails to present views when given the chance, one member is enough. The asymmetry buried in there is the whole point.

So the summary does something before anyone reads a word of it. Filed inside the window, it doubles the number of physicians who have to sign off on ending the stay. Missed, it halves that number. That is a procedural consequence of the document existing, independent of how well it argues, and it is the reason a late summary costs more than a thin one. Calderbrook missed the window on 39 of 411 submissions last quarter.

The log that accumulates is worth as much as the summaries. Concurrent review is a nonquantitative treatment limitation, and the parity rule requires the processes applied to mental health benefits be comparable to and no more stringent than those applied to medical and surgical benefits in the same classification. Calderbrook's behavioral health stays were reviewed every two days against every four for medical stays under the same plan. One stay proves nothing. Four hundred is a record.

How it works

  1. Send the record

    Whatever the clinician documented for the day, plus the last summary you filed on this stay.

  2. Name the reviewer

    Which payer, which criteria set they cited, and the structure and length they expect.

  3. Get the delta

    What changed since the last submission, mapped against the discharge criteria the payer stated.

  4. File inside the window

    On the record, on time, because the filing changes what the plan has to do.

What you get

  • Each day's summary written as the documented change since the last one submitted
  • The reviewer's structure and word limit treated as hard constraints, with cuts explained
  • Clinical content carried forward exactly as the clinician recorded it, never generated
  • Days with no documented change flagged before the summary goes out, not after
  • Every submission timestamped against the payer's own window, because the filing itself matters
  • A running log of review frequency and criteria cited, per payer and classification

Common questions

Does it write the clinical content?

No, and it refuses to. Clinical assertions are carried forward exactly as the clinician recorded them, and anything the record does not contain is marked as a gap and left empty. A summary with an invented observation in it is worse than a late one, because it puts something in the chart that nobody said.

Why does a late summary matter so much?

Because the rule turns on whether views were presented when the opportunity was afforded. Present them and an adverse determination needs at least two committee members. Fail to present them and one is enough. The document is doing procedural work the moment it is filed, before its content persuades anybody of anything.

What happens on a day when nothing changed?

It gets flagged before the summary goes out rather than after the denial arrives. That flag is an administrative prompt with two administrative answers: something changed and was not documented, or the record supports discharge planning. Which of those is true is the clinician's call, and the tool does not guess at it.

How long should the summary actually be?

Whatever the reviewer said, treated as a ceiling rather than a target. Calderbrook's median landed at 418 words against a stated 600 limit, and 104 submissions ran over before editing. Overrunning does not add information, it moves the operative sentence somewhere the reviewer stops reading. Cuts get listed so nothing vanishes silently.

Is the review log useful for anything beyond the stay?

It is how a pattern question ever gets raised. Medicaid managed care contracts must include mechanisms to ensure consistent application of review criteria, and concurrent review is also a nonquantitative treatment limitation under the parity rule. Two reviewers citing different criteria for the same service is a finding. One stay cannot show it. A quarter can.

The stay was denied anyway. What now?

Then the venue changes and so does the document. A written appeal argues against a decision already made, which is a medical necessity letter built off the criteria the payer cited. If a call is offered instead, peer-to-peer preparation is the shorter version a clinician can hold in nine minutes.

The reviewer keeps citing criteria we have never seen.

Then ask for them, and there is often a disclosure obligation behind the request. Reading the criteria into a structure you can actually write against is a payer policy criteria summary. Without it every summary is a guess at what the reviewer is checking, which is how a well documented stay still fails a review.

Utilization Review Summary and Change Log

Fill in the form and your workspace opens with the work already underway.