Interoperability’s Next Job Is Adjudication

Pay for performance is the most agreed-upon idea in American healthcare and one of the least achieved. Payers endorse it. Drugmakers keep proposing warranties tied to it. And yet after a decade of concerted national effort, 28.7 percent of healthcare payments carried downside risk in 2024, barely up from 28.5 percent the year before. The industry’s favorite explanations involve courage and culture. The binding constraint is duller. Nobody can prove an outcome happened at a price that leaves the contract worth signing.

This piece is about that constraint, and about the odd fact that the tools to remove it were just built for unrelated reasons.

The Referee Problem

Every performance-based arrangement lives or dies on a single question: did the outcome occur? A readmission avoided, a tumor that did not progress, an A1c under target. Today that question gets answered with claims data that lags by months and chart pulls performed by hand, with each side working from its own extract of reality. Reconciliation for a typical risk deal commonly lands a year or more after the performance period, and disputes are routine because there is no shared evidence, only competing spreadsheets.

The cost of that verification machinery sets the floor for what can be contracted. When proving performance is a major project, only broad, blunt contracts clear the bar, so the market defaults to upside-only arrangements that demand less proof. Pharma outcomes warranties exist mostly on paper for the same reason: the refund is easy to promise and expensive to establish. What the market has lacked is a referee both parties can afford and accept.

The Rails Got Built Anyway

While pay-for-performance idled, the country built national data infrastructure for other reasons. TEFCA, the federal trusted exchange framework, went live in December 2023 and has now moved more than 1 billion records across more than 70,000 participating organizations. Its approved exchange purposes were never limited to treatment. Payment and healthcare operations sit on the list, and response requirements began expanding into operations use cases like quality measurement this year. On a parallel track, CMS finalized rules requiring Medicare Advantage, Medicaid, CHIP, and exchange plans to stand up FHIR interfaces for patient, provider, and payer-to-payer data by January 1, 2027.

All of it was justified in the language of care coordination and patient access. The unexploited fact is what the same rails can carry. A performance contract is, at bottom, an evidence problem, and the evidence now moves: lawfully, because payment and operations are permitted purposes, and technically, because the networks and interfaces exist. Using interoperability to adjudicate contracts is the most valuable application nobody built it for.

Adjudication at Contract Speed

The standard objection is speed, the worry that clinical data cannot flow fast enough to settle money in real time. The objection aims at a bar no contract actually sets. Settlement runs on a calendar, and today’s calendar is a reconciliation that lands a year or more after the performance period. An adjudication process that assembles the record and reads the outcome within weeks would feel instantaneous by comparison, and a quarterly cadence would transform the economics of every risk deal in the market. Nothing about this requires milliseconds. It requires evidence that arrives before the dispute does.

In practice it looks like this. The contract names the measure and the evidence that satisfies it. As care happens, records accumulate across sites and move over the rails. An adjudication layer assembles them, extracts the outcome, and attaches a citation to the source document behind every fact, so both parties open the same evidence rather than exchanging summaries of their own. Settlement then follows whatever calendar the contract sets.

Picture an outcomes warranty on a single high-cost therapy. The contract names the measure, response at a defined interval. Records from every treating site accumulate across the rails as care proceeds. The adjudication layer reads the outcome and cites the source document behind it. The refund settles in weeks, not after a claims runout. This is a hypothetical, but nothing in it waits on technology that does not exist.

What Becomes Contractable

Collapse the cost of proof and the contract space expands in every direction. Outcomes warranties become viable for a single therapy or a single patient, because establishing the refund no longer costs more than the refund. Provider risk arrangements settle quarterly, and disputes narrow to questions of interpretation rather than questions of fact. Episode-based deals get adjudicated from the chart instead of waiting out claims runout. Purchasers who never trusted a performance guarantee get guarantees they can check. And the interoperability budget line, booked everywhere as a compliance cost, starts functioning as deal infrastructure.

The Layer on Top

The rails move documents. Adjudication needs facts. Records arrive as discharge summaries, lab feeds, scanned notes, and imaging reports scattered across formats and sites. Someone has to resolve them into clinical facts precise enough for a contract to bind to. That is the layer we built at xCures_®. The Clinical Clarity Engine turns records gathered from care sites nationwide into the specific facts a contract names. Each fact carries a citation to the exact source document behind it. The extractors are validated against clinical review and the results are published: 95 percent or better positive predictive value and sensitivity across our clinical checklists. A referee’s methods must survive inspection from both sides of the table. That standard was set in oncology, the most demanding records in medicine. Put this layer on top of the rails and the neutral referee stops being hypothetical. Neutrality lives in the citation, not the vendor. Either party can open the source document and check.

The Second Act

I argued recently in Health Affairs that the way Medicare Advantage prices risk rewards paperwork over patients. Paying for performance has suffered a version of the same disease, promises denominated in outcomes and settled on artifacts. Interoperability’s first act was clinical, helping the next clinician see the chart, and it was worth building for that alone. The second act is economic. The same rails that carry a record to a bedside can carry proof to a settlement, and proof is the only thing pay-for-performance was ever missing.

 See what xCures clinical clarity looks like in your workflow.