A claim comes back paid, the account closes, and the team moves on to the next one. That’s fine, as long as the payment was right.
The trouble is that it isn’t always easy to say what a claim should have paid under its contract. Without that number, an underpayment looks like every other paid claim. Artificial Intelligence (AI) is changing that, making it practical to know what every claim should have paid.
Fewer Denials, Same Cash Problem
On paper, denials look better. A Kodiak Solutions analysis of claims from more than 2,300 hospitals and 375,000 physicians found that the overall initial denial rate fell to 10.63% in the first half of 2026, down from 12% a year earlier.¹ Revenue pressure hasn’t eased so much as shifted later in the cycle, after claims have already been paid. Over the same period, payer takebacks rose to 1.57% of accounts receivable, up from 1.38%, a 13.8% increase. Meanwhile, providers recovered less when they pushed back, with post-payment recoveries falling 12.2%.²
A denial is visible and can be worked. A takeback arrives after the revenue is booked. Underpayments are harder still to see: the claim is paid, just not in full, and nothing flags the gap unless someone knows what the claim should have paid. Fewer claims are being denied, but cash is still at risk.
Why Claims Get Priced Wrong
Payer agreements layer rate exhibits, amendments, and interdependent pricing terms. In many organizations, an analyst reads each agreement and hand codes every term as if-then logic in a SQL database or contract management system: if the claim falls in this DRG, pay this rate; if charges cross this threshold, switch to a percentage of charges.
It’s slow, error-prone, and requires rework any time the contract changes. If a term is missed or a rate is entered incorrectly, the system calculates the wrong expected payment, and an underpaid claim passes as paid correctly.
Where Underpayments Hide
Complicated provisions leave more room for error. Lesser-of terms, stop-loss thresholds, and carve-outs each change how a claim is priced, and pass-throughs pay certain costs, such as implants, on top of the base rate.
Consider a hypothetical inpatient claim. The contract calls for a $27,000 DRG-based payment plus a $7,500 pass-through for an implant, so the expected payment is $34,500. The payer pays $27,000.
If the pass-through was never built into the health system’s contract model, the claim shows as paid in full. Nobody denied the $7,500. Nobody collected it either.
Pricing Every Claim
AI replaces antiquated manual contract modeling with automated modeling at scale. A large language model reads the agreement and pulls out its rates and rules, and a rules engine turns them into pricing logic.
Claims are priced in seconds instead of days. And instead of reviewing only a small sample of claims, as manual audits typically do, every claim in a batch is checked. Each claim can be scored on how likely it is to be a true underpayment, so staff start with the strongest opportunities. Every expected reimbursement is explained rule by rule, with the specific contractual language that supports an appeal, so analysts spend their time building recovery arguments instead of paging through PDFs.
AI works best paired with expert review. A model can misread a rate table or record the same rule twice, so experienced people check every extracted rule against the contract before it prices claims. AI takes on the volume, and people make the judgment calls.
Seeing the Pattern
Because AI prices every claim, underpayments can be sorted by payer, service line, and facility. One-off errors can turn out to be one repeating problem.
Some of those problems start outside the contract. In a 2026 HFMA and Trek Health survey of 161 provider leaders, 68% cited payer policy changes not reflected in contracts as a revenue leakage risk, yet only 26% said they proactively track and model those changes across major payers before reimbursement is affected.3 Tracking where shortfalls cluster gives health systems evidence for appeals and leverage at the next negotiation.
Regulators are paying attention, too. Through Worksheet S-12, IPPS hospitals must report median payer-specific charges they negotiated with Medicare Advantage organizations by MS-DRG,4 and CMS intends to use that data to set MS-DRG weights beginning in fiscal year 2029.5 That raises the stakes on knowing exactly what each contract pays.
The Bottom Line
Lower denial rates are welcome, but they don’t tell you whether you’re collecting what your contracts promise. You can’t recover what you can’t price. AI, backed by experienced reviewers, makes it possible to know what each claim should have paid and to act on the difference.
Sources
- TechTarget Rev Cycle Management (news article). “Insurers Deny Less Just to Claw Back More Later: Analysis.” https://www.techtarget.com/revcyclemanagement/news/366651012/Insurers-deny-less-just-to-claw-back-more-later-analysis
- Kodiak Solutions (press release). “Falling Rates of Insurance Claim Denials Fail to Drive Better Financial Performance for Hospitals, Health Systems and Medical Practices, According to a Kodiak Solutions Analysis.” https://www.businesswire.com/news/home/20260915724994/en/Falling-Rates-of-Insurance-Claim-Denials-Fail-to-Drive-Better-Financial-Performance-for-Hospitals-Health-Systems-and-Medical-Practices-According-to-a-Kodiak-Solutions-Analysis
- Becker’s Hospital Review (reporting HFMA and Trek Health survey). “The payer policy problem draining hospital revenue: 6 things to know.” https://www.beckershospitalreview.com/finance/the-payer-policy-problem-draining-hospital-revenue-6-things-to-know/
- Centers for Medicare & Medicaid Services. “Worksheet S-12.” https://www.cms.gov/medicare/audits-compliance/part-cost-report-audit/worksheet-s-12
- Forvis Mazars. “New Cost Report Requirement: Reporting MA Negotiated Rates.” https://www.forvismazars.us/forsights/2026/01/new-cost-report-requirement-reporting-ma-negotiated-rates
Becker’s 3rd Annual Fall Chief Pharmacy Officer Summit is the top tier event that the people leading the future of pharmacy convene. This 3-day event is built specifically for Chief Pharmacy Officers and senior pharmacy leaders. The sessions examine the workforce, financial and clinical priorities shaping pharmacy today. Becker’s is where candid discussions flourish and new ideas thrive. Becker's 3rd Annual Fall Chief Pharmacy Officer Summit
