HomeProductProvider & Venue Outlier Map
Provider & Venue Outlier Map

Providers and venues, mapped against the portfolio's own severity patterns.BETA

Provider and venue outlier mapping software, in beta, surfaces which providers, venues, and treatment patterns correlate with higher severity or cost across your claim portfolio. Every pattern is benchmarked against your own portfolio, cited back to the underlying claims, and shown only once it clears a minimum sample size.

Portfolio severity signalsProvider & venue
Cedar ValleyWithin range
Pinnacle Ortho & SpineWithin range
County MedicalRouted to SIU review
Pattern signal only, requires minimum sample size · not a fraud determination
BETA STATUS
In active beta, refined hands-on with early carrier and TPA customers.

Severity and cost patterns, mapped by provider and venue.

The map aggregates claim outcomes by provider, venue, and treatment pattern across the portfolio you route through Medrecords AI, then benchmarks each one against the portfolio's own norms, with the sample size shown next to every pattern.

Every pattern shows the sample size behind it
Benchmarked against your own portfolio, not a public database
Severity vs. portfolio norm
Cedar ValleyWithin range
Pinnacle Ortho & SpineWithin range
County MedicalOutlier
How a pattern gets to review
Pattern detected
Across provider, venue, or treatment
Sample size check
Below minimum, nothing shown
Routed to SIU queue
A signal for human review

A signal that routes to a human, never a verdict.

Patterns below the minimum sample size aren't shown at all. Patterns that clear it route to your SIU or review queue as a signal, cited back to the specific claims that produced it, for a person to open and investigate.

Below the minimum sample size, nothing is shown
Every flagged pattern links back to the underlying claims
The guardrail

A pattern signal, never a fraud label.

The map never labels a provider or venue as fraudulent. It surfaces a statistical pattern, cost or severity that sits outside the portfolio's normal range, and nothing more. Correlation shown here is not an accusation.

Every pattern requires a minimum sample size before it's shown, so a single unusual case never triggers a flag. When a pattern clears that bar, it routes to your SIU or review queue as a signal for a human to investigate, never as an automated determination.

From portfolio to signal, in three steps.

Aggregate, benchmark, and route, with a human deciding what happens next.

1. Aggregate the portfolio

Claim outcomes are aggregated by provider, venue, and treatment pattern across the book you route through Medrecords AI.

2. Benchmark against the portfolio norm

Each pattern is measured against the severity and cost range typical for the portfolio, with the sample size shown alongside it.

3. Route outliers to review

Patterns that clear the minimum sample size route to your SIU or review queue as a signal, cited back to the claims behind it.

Built for the teams that manage a claim portfolio.

A portfolio-wide view for the organizations that route large volumes of claims.

FAQ

Provider & Venue Outlier Map, answered.

It surfaces which providers, venues, and treatment patterns correlate with higher severity or cost across a claim portfolio, benchmarked against the portfolio itself and cited back to the claims behind each pattern.

No. It never labels a provider or venue as fraudulent. It shows a statistical pattern relative to the portfolio; what happens next is entirely a human decision.

It routes to your SIU or review queue as a signal for a person to investigate, cited to the specific claims behind the pattern. It's not an automated determination.

Yes. A pattern has to clear a minimum sample size before it's shown at all, so a single unusual claim never triggers a flag.

Yes, in beta. The Provider & Venue Outlier Map is live and testable now on your own portfolio; we're refining it hands-on with early customers, and if your use case is a good fit we'll work with you directly.

Related capabilities.

See the outlier map on your own portfolio.

Join the beta and run it against a real book of claims, or book a demo first.