HomeProductSettlement Variance Detector
SETTLEMENT VARIANCE DETECTION SOFTWARE

Settlements outside the cohort range, with the medical drivers shown.

BETA

Settlement variance detector software from Medrecords AI flags claims that settled above the comparable cohort without medical drivers that explain the gap — a pattern signal for review, never an accusation. Live in beta, refined hands-on with early carrier and TPA customers.

A pattern flag, never an accusation: a person reviews every flag before any conclusion is drawn about overpayment, error, or misconduct.

Claim #48213 · lumbar strain, settled flagged for review
Settled value $92,000
Comparable cohort range $58,000–$81,000
Surgical intervention on file none found
Imaging findings unremarkable
Treatment course standard
A pattern flag, not a finding — your reviewer determines the reason.
Status BETA · in active testing
Compares settlements against the same comparable-case cohorts used across the platform's benchmarking tools — refined hands-on with early carrier and TPA customers.

Every settlement, compared to its cohort.

A settled claim's value is compared against comparable resolved claims — matched on injury, venue, and era — the same benchmarking mechanics used across the platform, applied at settlement instead of before it.

Matched on injury type, venue, and era, cohort shown
Every comparison cites the cohort it's measured against
Settlement comparison
Cohort match basis shown
Cohort size shown
Comparison type range, not a score
The same comparable-case mechanics as Case Outcome Benchmarks.
Driver check
Surgical intervention checked
Imaging findings checked
Unexplained gap flagged for review
A flag is a starting point for review, never a conclusion.

The medical drivers are checked before anything is flagged.

A flag only appears when the medical severity drivers on file — surgical intervention, imaging findings, treatment intensity — don't explain a settlement above the cohort. If the drivers explain the higher value, nothing is flagged at all.

Checks surgical, imaging, and treatment-intensity drivers first
No flag when the medical drivers explain the settlement
Division of labor by design
The system compares settled value to cohort, checks drivers
The flag a pattern outside the range, cited
Your reviewer determines the reason
Overpayment, error, or misconduct never concluded by the system
The boundary

A pattern flag. Never an accusation.

Settlement Variance Detector compares a settled claim's value to its comparable cohort and checks whether the medical drivers on file explain the difference. When they don't, it surfaces a pattern flag for review — cited to the drivers and the cohort it's compared against.

It never accuses anyone of overpayment, error, or misconduct, and it never reopens or revalues a closed settlement on its own. A flag is a starting point for a person to review, nothing more.

See Case Outcome Benchmarks

From a settled claim to a cited pattern flag.

Three steps, ending with a person's review, not a conclusion.

STEP 1
Settled claims compared to cohort

Matched on injury type, venue, and era against comparable resolved claims.

STEP 2
Medical drivers checked

Surgical intervention, imaging findings, and treatment intensity checked against the gap.

STEP 3
Unexplained variance flagged for review

Cited to the drivers and cohort — never a conclusion about why the gap exists.

Who reviews settlement variance with it.

Built for the teams that audit settlements after the fact — carefully.

FAQ

Settlement Variance Detector, answered.

Yes, in beta. It's live and testable now on your own settled claims; we're refining cohort matching hands-on with early carrier and TPA customers, and if your use case is a good fit we'll work with you directly.

No, and it never will. It surfaces a pattern flag when a settlement sits above the comparable cohort without medical drivers that explain the gap. Whether that means a documentation gap, unusual case facts, or a perfectly reasonable settlement decision is for your reviewer to determine.

Surgical intervention, imaging findings, treatment intensity or duration, and documented complications — each checked and cited before any flag is raised.

No, never automatically. A flag is for internal review only; any action on a closed settlement is a decision made by your team.

Claims are matched by injury type, venue, and era, using the same comparable-case mechanics as Case Outcome Benchmarks, with cohort size and composition always shown.

Related capabilities.

Settlement variance uses the same comparable-cohort mechanics as the platform's other benchmarking tools.

See settlement variance flagged against your own book.

Join the beta and run settlement variance detection on your own settled claims — a cited pattern flag, never an accusation. Or book a quick walkthrough first.