Healthcare / Health Plans · Fraud, Waste & Abuse / Special Investigationshealth plan
Investigation Case Building & Referral
Readiness: Now — Deployable with established commercial tools today · Near-term — Proven but early — expect one to three years to mainstream · Emerging — Demonstrated, not yet production-mainstream
Readiness reflects an editorial assessment against a published rubric as of August 2026 — an observation about current tool maturity and adoption, not a prediction about specific products or timelines.
Trajectories describe the observable direction of human effort — not a prediction about specific roles, headcount, or individual careers.
What You Do Today
A lead becomes a case when someone has to prove it. You pull the subject's full claim history and every adjudication decision behind it, request records for the dates you intend to rely on, and read what the documentation says against what was billed, date by date. You look for the note that does not mention the service, the time-based code with no time recorded, the identical note appearing across unrelated members, the signature that is not there. You interview: the member who can say whether the visit happened, the former employee who can say how the billing was actually done, sometimes the provider. You calculate exposure across the claims that fit the pattern you have established, and where extrapolation is contemplated the sample design has to be defensible before it is drawn. Then you decide the disposition, and each one goes somewhere different — an education letter, a recovery demand, placement on pre-payment review, a contract action, a referral to the state fraud bureau, to a Medicaid Fraud Control Unit, or to federal investigators. The file has to survive being read by someone who was not there and has no reason to be persuaded: a hearing officer, a prosecutor who will decline it if it is thin, opposing counsel. Meanwhile the same provider may be disputing claims, coming up for recredentialing and negotiating a contract, and none of those teams knows what you know.
AI Technologies
Roles Involved
How It Works
Assembly gathers what an investigator would otherwise collect by hand — every claim the subject submitted and how each one adjudicated, authorizations, prior contacts and prior investigations, the provider file and its credentialing history, correspondence, and records the plan already holds from earlier reviews — into one file that records where each item came from, who pulled it and when. Documentation comparison sets the record beside the claim for each date of service and reports the specific element that is missing or contradicted rather than a pass or a fail, because which element fails is what separates a documentation problem from a coding problem from something else, and because near-identical notes across unrelated members are visible to text comparison in a way they are not to a reviewer reading one chart at a time. Chronology construction orders claims, records, contacts, ownership changes and prior actions onto a single timeline, which is how a pattern becomes a narrative someone else can follow. Quantification applies the pattern the investigation established to the universe of claims that fit it, supports the design and draw of a sample where extrapolation is contemplated, and keeps the arithmetic reproducible, because the number will be challenged and whoever produced it will have to walk through it.
What Changes
An investigator opens a file that is already assembled rather than spending the opening of every case assembling it. Records are read against claims across all the dates rather than a sample of them. Where each piece of evidence came from is recorded as the file is built instead of reconstructed when somebody asks. Exposure is computed the same way from case to case, which makes cases comparable and dispositions more consistent. Duplicate documentation across unrelated members becomes visible. And the referral narrative starts from an ordered record rather than from a blank page.
What Stays the Same
Intent is the whole case and no model can see it. The line between a billing error, abuse and fraud is what the person knew and meant, and it decides everything downstream — whether this ends in an education letter or with a prosecutor — so it is established by people who can be cross-examined on how they established it. Whether a record supports what was billed is a clinical determination made by a qualified clinician who can be named and whose reasoning has to hold up when the treating provider reads it back. Interviews are the case as often as the data is: what a member says about whether the visit happened, what a former employee says about how the billing was done, what a provider says when asked directly. Evidence has to be answerable for — how it was obtained, under what authority, within the limits on what the plan may pull and disclose — and a file that cannot say where something came from does not survive the first serious challenge. Referral is a judgment about somebody else's threshold rather than the plan's: a thin case that gets declined costs more than the case, because it spends the credibility that makes the next call answerable. Once a plan identifies an overpayment it carries a legal obligation to act within a statutory deadline, so the moment something becomes 'identified' is a determination made with counsel rather than a status a system sets, and running the detection creates the knowledge that starts the clock. Putting a provider on pre-payment review or out of the network reaches the members who see that provider, and access is a real cost that belongs in the decision. A model built on the unit's closed cases learns the cases it closed, including the ones it should have lost. And closing a case as unsubstantiated, plainly and on the record, is a discipline, because the pressure to justify the time already spent is what turns a weak case into a bad referral.
Evidence & Sources
- •False Claims Act enforcement by the U.S. Department of Justice
- •HHS Office of Inspector General Self-Disclosure Protocol
- •HHS Office of Inspector General List of Excluded Individuals/Entities (LEIE)
- •State Medicaid Fraud Control Units certified by the HHS Office of Inspector General
- •CMS requirements for Medicare Advantage and Part D sponsors to report and return identified overpayments
- •HIPAA Privacy Rule permitted disclosures for health oversight activities and law enforcement purposes (45 CFR 164.512)
- •NAIC Insurance Fraud Prevention Model Act and state insurance fraud bureau reporting requirements
- •AHIMA documentation and coding integrity guidance
Sources listed are directional references, not formal citations. Verify against primary sources before using in business cases or presentations.
Last reviewed: August 2026
What To Do Next
This section won't tell you what your numbers should be. It will show you how to find them yourself. Every instruction below produces a real, verifiable result in your organization. No benchmarks, no projections — just the steps to build your own evidence.
Establish Your Baseline
Know where you are before you move
Before adopting AI tools for investigation case building & referral, document your current state in utilization management.
Without a baseline, you can't tell whether AI actually improved investigation case building & referral or just changed who does it.
Define Your Measures
What to track and how to calculate it
patient outcomes
How to calculate
Measure patient outcomes for investigation case building & referral before and after AI adoption. Pull from your special investigations case system.
Why it matters
This is the most direct indicator of whether AI is adding value to utilization management.
clinical documentation quality
How to calculate
Track clinical documentation quality using the same methodology you use today. Don't change how you measure just because you changed how you work.
Why it matters
Speed without quality is just faster mistakes. Measure both together.
Start These Conversations
Who to talk to and what to ask
CMO or VP Clinical Operations
“What's our plan for AI in utilization management? Are we piloting, planning, or waiting?”
This tells you whether to experiment quietly or push for formal investment in investigation case building & referral.
your special investigations case system administrator or vendor
“What AI capabilities exist in our current EHR system that we're not using? Most platforms are adding AI features faster than teams adopt them.”
The cheapest AI adoption is the features already included in your existing license.
a practitioner in utilization management at another organization
“Have you deployed AI for investigation case building & referral? What worked, what didn't, and what would you do differently?”
Peer experience is more useful than vendor demos. Find someone who has actually done this.
Check Your Prerequisites
Confirm readiness before you invest
Check items as you confirm them.