Ask your team that question right now. I ask it everywhere I go. I have sat with some of the sharpest risk adjustment operators in the country, and I have never once heard a confident yes.
Today we are changing that. Invent Health is launching Encounter Submissions, a closed-loop engine that carries every encounter from the source chart to CMS acceptance, for Medicare and ACA, in one system. It has already done this work at scale, inside operations supporting more than five million members. Now it is a product any plan can run. This is the most important release since I founded the company. The reason we built it is bigger than the product.
You already know how the ground shifted this spring. Unlinked chart reviews are going away, and about $7 billion moves with them in 2027, per CMS’s 2026 Rate Announcement Every plan is audited every year now. The 2026 dates of service that decide 2027 payment are moving through your pipeline as you read this. For two decades, this industry treated “submitted” as good enough. CMS just made “accepted” the only word that pays.
The vacuum we built this for
Before I founded Invent Health, I spent twenty years with systems that move data between payers and providers. We connected almost everything, and I learned that connection is not intelligence. You can wire every system to every other system and still not answer the one question that decides your revenue.
The whole industry raced to find one more diagnosis. Billions went into the hunt: suspecting, chart chases, coding at scale. Almost nothing went into proving the find survived, because proof was treated as plumbing.
So, a gap opened inside every plan, between what was coded and what CMS counted, and nobody owned it. Not the coding vendor, whose job ended at delivery. Not the submission team, whose report stopped at “sent.” Not the risk team, who never saw the response files. The most important number in the business sat in a no man’s land the industry built by accident and then stopped noticing. Keeping the three systems in step, encounter submission, risk analytics, and coding, is real work that fell to no one. The seams between them are where the money leaks out.
We started this company to capture the opportunities no one is watching. This is the biggest one we have ever found.
What we see that you cannot
When we reconcile a plan’s chart-coded diagnoses against CMS acceptance end to end, we find money. We find rejected encounters nobody resubmitted, chart reviews pointing at encounters that were voided months earlier, and vendor reports that say linked while the CMS response files say otherwise. None of those failure modes will surprise you. You have lived with all of them.
Here is what should stop you: almost no plan has ever produced the number. The full, end-to-end gap between what your coders documented and what CMS actually counted. Your teams are strong and your vendors are expensive, and the number still does not exist. The industry never built the machine that produces it.
Every budget, every vendor scorecard, every board report about risk performance rests on a figure nobody has verified against the only ledger that pays. That was survivable when the rules forgave it. The rules stopped forgiving it this spring.
A plan gets paid for the risk it can prove, judged on the quality it can show, and audited on the evidence it can produce.
We built the engine that makes all three the same motion.
How Encounter Submissions closes the loop
This is not a better submission tool. The market is full of tools that generate a file and call the job done. A closed loop means nothing leaves the pipeline without a resolution.
Encounter Submissions generates, validates, submits, and reconciles every encounter, from the source chart to CMS acceptance, for Medicare and ACA in a single engine.
When an encounter breaks, the AI drafts the fix and an analyst approves it before anything moves. Our AI surfaces what matters but never invents a diagnosis, so nothing enters the pipeline that real documentation cannot support. And because it runs beside our risk analytics and coding work, the chart, the evidence, the diagnosis, and the CMS acceptance live in one place.
We aim for more than 99% first-pass acceptance across Medicare and ACA from day one. Inside a closed loop, a rejected encounter stops being a line in a monthly file. It becomes a work item with a fix, an owner, a resubmission, and a confirmed acceptance at the end.
The audit trail is not something you rebuild when CMS calls. You kept it the whole way through. That is the difference between defending your revenue and hoping it’s okay.
Where this is heading
I believe the accepted encounter record is becoming the plan’s real ledger. Payment already reads from it. Audit defense reads from it. Quality data rides on it. Every rule CMS has recently written points in the same direction, and I do not expect the direction to change.
The plans that close the loop first will run differently. They will know their gap between coded and counted the way a CFO knows cash. They will stop paying for chase work that produces diagnoses their pipeline cannot land. And when the auditor arrives, they will not assemble a defense, because their operations already has it covered. I built this company for those plans, and for the leaders who run health plans that refuse to walk away from their members. Proving every dollar is how they afford to stay.
Here is what you need to do
Start with three questions. Do your encounters live in a separate system from your risk and coding work? Who owns your submissions, and if it is a vendor, would you know what they missed? And the question I opened with: can you prove every coded diagnosis reached CMS and was accepted?
If any answer gives you pause, the gap is already costing you, and now you can measure it. We will walk your pipeline with you, from source chart to CMS acceptance, and put your number in front of you. You will see what is clean, what is leaking, and what to fix first. The real question is whether your current systems are ready for the 2027 rules. Invent Health is built for these changes, so you meet them without rebuilding your pipeline under a deadline.
Frequently Asked Questions
What is Invent Health Encounter Submissions?
Encounter Submissions is Invent Health's closed-loop encounter engine, now available to all health plans. It generates, validates, submits, and reconciles every encounter from the source chart to CMS acceptance, for Medicare and ACA in a single engine, and it runs alongside Invent Health's risk analytics and coding work.
How can a plan prove its diagnoses reached CMS?
Proof requires end-to-end reconciliation: matching every chart-coded diagnosis to a submitted encounter, then confirming that encounter in the CMS response data as accepted. Reports that stop at “submitted” are not proof. Only the CMS acceptance record shows which diagnoses actually count toward risk scores and payment.
What is closed-loop encounter submission?
Closed-loop encounter submission means every encounter is carried through a full cycle: generated, validated, submitted, and reconciled against CMS acceptance. Nothing leaves the pipeline without a resolution. A rejected encounter is fixed and resubmitted, and every chart-coded diagnosis is confirmed to have landed on an accepted encounter.
What is an open loop in risk adjustment?
An open loop is a risk adjustment operation that tracks what was submitted but never confirms what CMS accepted and counted. Charts are coded, and files are sent, but rejections, voids, and broken chart review linkage go unresolved. The revenue impact stays invisible because nobody reconciles the two ends.
What is the claim linking problem in risk adjustment?
The claim linking problem is the difficulty of tying chart review diagnoses to an accepted base encounter at CMS. A chart review that points to a rejected or voided encounter is unlinked in practice, even if a vendor's system shows it as linked. Only linkage that CMS recognizes supports the risk score.
Why do encounters get rejected by CMS?
Encounters are rejected when they fail CMS audits, such as missing or mismatched member data, invalid codes, or conflicts with previously accepted records. Rejections are normal in every pipeline. The revenue damage comes when nobody catches the rejection, fixes the encounter, and resubmits it before the deadline passes.
What changed with unlinked chart reviews for 2027?
CMS finalized the exclusion of unlinked chart reviews in its April 2026 Rate Announcement, then set the details in a July 9 2026 HPMS memo. Diagnoses from unlinked chart reviews for 2026 dates of service will not count toward risk scores. The exclusion does not start at the beginning of 2027. It begins at the 2027 midyear model run, around July 2027, and CMS goes back and adjusts the payments it already made for the first half of the year. There is an exception for members who switch to a Medicare Advantage contract under a different parent organization.
How do plans reconcile encounter submissions with CMS?
Reconciliation means comparing what CMS accepted against what the plan submitted and what its coders documented, member by member and diagnosis by diagnosis. Many plans stop at tracking what was sent. Full reconciliation reads the CMS response data and confirms every chart-coded diagnosis landed on an accepted encounter.
What happens to diagnoses on rejected encounters?
Diagnoses on rejected encounters do not reach the risk score, so the plan is not paid for them. Many plans miss these losses because their reporting tracks submissions, not acceptances. Finding and fixing rejected encounters is one of the fastest ways to recover revenue the plan already earned.
