Defensible risk adjustment requires one evidence chain

Health plans experience risk adjustment as a continuous cycle. Prospective identification happens before the encounter, concurrent review happens during the visit, retrospective validation happens after claim submission, and member management runs across the plan year. The technology architecture built to support that cycle either compounds evidence at every stage or starts over at each handoff. Payer strategy teams tend to price the second option as a workflow inconvenience. The first months of 2026 made clear that it is a compliance liability with a dollar figure attached.

Fragmentation now carries a measurable cost

Payer strategy leaders already sense the shape of this problem, and McKinsey's own administrative-simplification research puts a figure behind it. The firm estimates administrative work consumes close to a quarter of the nearly four trillion dollars the country spends on healthcare annually, and that $265 billion of administrative spending could be eliminated. Of that total, $35 billion comes from closing handoffs between separate organizations, and another $105 billion comes from broader structural change across the industry. The same pattern describes what happens when prospective and retrospective risk adjustment run through separate technology partners. Reconciliation cost is a documented, budgeted line item, and it grows with every additional system a program depends on.

Compliance risk stacks directly on top of that cost. On February 3, 2026, OIG released its Medicare Advantage compliance guidance, the first update since 1999. The guidance named add-only chart review programs, retrospective processes that surface new diagnoses without a parallel discipline to remove unsupported ones, as a defined compliance failure with direct enforcement weight. Whichever technology partner produces those diagnoses has to apply that same discipline in both directions, adding and removing, since a program built only to add is the pattern OIG's guidance treats as a compliance failure regardless of how many vendors are involved. A health plan running multiple technology partners has to verify that discipline separately for each one, an oversight burden a single evidence chain removes.

CMS closed the same vulnerability from the payment side. The CY2027 Rate Announcement, finalized April 6, 2026, excludes diagnoses from unlinked chart review records from risk score calculation altogether. A diagnosis that cannot be tied to a documented encounter loses eligibility for payment, independent of which system produced it, and independent of how confident that system's model was when it surfaced the code.

I discussed this same shift, from RADV as a periodic event to RADV as a year-round operating condition, in a Becker's Healthcare interview earlier this year, and the reconciliation problem described here is the operational detail that conversation left unexplored. Stronger clinical data infrastructure and year-round documentation readiness only pay off when the systems producing that documentation agree with each other in the first place.

Multiple technology partners create multiple systems to reconcile

A payer strategy team that runs prospective identification through one technology partner and retrospective validation through another is running two programs built on the same coding and clinical support standards, but processed through separate systems; separate encounter processing, and separate claims processing on the health plan side. In 2026, reconciling those systems only when a chart is questioned is when it costs the most. Accuracy metrics on either system miss this discrepancy entirely. The cost accumulates in the compliance team's calendar, in the hours spent rebuilding a record that should have arrived whole.

Where multiple systems add work and risk

Maintaining two distinct systems has implications across evidence standards, audit response, encounter linkage, and partner management.

‍

Prospective and retrospective risk adjustment must work as one program

Prospective point-of-care capture improves documentation timeliness and reduces reliance on after-the-fact chart mining, but it leaves two gaps behind for whichever process runs after it. Members without a recent visit may have conditions that were never documented at all, and members who were seen may have a condition the visit documented but the claim never billed or submitted to CMS. Reveleer has made the case for a blended approach before. Prospective and retrospective methods solve different halves of the same problem, and the health plans treating them as substitutes for one another are the ones still explaining discrepancies to an auditor.

Retrospective review catches what prospective capture missed, but a retrospective program lacking upstream discipline is the exact pattern OIG's guidance now names directly, an add-heavy process that, absent a parallel deletion standard, reads to an auditor as a program built primarily to inflate risk scores. A payer strategy team stitching both halves together across separate technology partners inherits the harder problem underneath the workflow stream, multiple varying evidence standards and a fragmented record in place of a single source of truth to hand a coder, a compliance reviewer, or a federal auditor when a code is questioned.

One evidence chain creates one defensible record

Reveleer supports health plans across risk adjustment, quality improvement, and member management as a single technology partner, structured around four connected workflow families, Prospective, Concurrent, Retrospective, and Member Management, which also carries the submissions work that keeps encounter data flowing to CMS. A code captured prospectively keeps the same evidence standard through concurrent review and into retrospective validation, because it moves through one evidence chain, held end to end instead of handed off between systems built by different teams for different specifications.

For a payer strategy leader, a shared evidence chain eliminates the task of checking prospective suggestions against retrospective findings from two systems running on different logic, then explaining the mismatch to whoever asks. One audit trail replaces many, on whatever timeline an OIG inquiry sets. RAF accuracy reflects one program's discipline as opposed to a blended average from multiple systems that never had to agree with each other.

2027 renewal decisions should prioritize evidence continuity

Health plans renewing or rebidding risk adjustment technology partner relationships heading into 2027 should ask a direct question of every candidate. Does a code captured at one stage of the cycle move its evidence forward to the next stage automatically, or does that fall to the health plan's own team to reconcile? The McKinsey research and OIG's guidance both answer the question of what the second scenario eventually costs, in dollars and in exposure. The first is the one worth building toward now, before an auditor asks which scenario describes the program in front of them.

Risk adjustment vendor consolidation FAQs

What changes when a health plan consolidates prospective and retrospective risk adjustment under one technology partner?

The evidence standard becomes singular instead of split. A diagnosis captured prospectively keeps the same clinical evidence and encounter linkage through concurrent review and into retrospective validation, instead of being re-evaluated against a second technology partner's independent logic. The health plan's compliance team produces one audit trail instead of reconciling two, and RAF accuracy reflects one program's discipline instead of the average of two systems that were never built to agree with each other.

Does moving to a single technology partner require replacing existing chart retrieval or coding infrastructure overnight?

Consolidation is typically staged by workflow family, prospective first, retrospective and member management following, on a timeline the health plan controls. The objective is a single evidence chain across the stages a health plan chooses to consolidate, achieved through a controlled migration that respects infrastructure still performing well.

How does the CY2027 exclusion of unlinked chart review codes affect technology partners versus health plans directly?

The payment impact falls on the health plan, since CMS calculates risk scores at the contract level, but the operational burden of proving encounter linkage falls on whichever system produced the diagnosis. A health plan running two technology partners absorbs that burden twice, once per system, which is precisely the reconciliation cost a single evidence chain removes.

I raised the same theme on a Becker's Payer Roundtable panel on pioneering innovations for health plan executives. Technology earns its keep in payer operations by removing coordination work, and it loses that value the moment it becomes another system for a payer strategy team to manage.

‍

‍