Your CDI Program Isn’t the Problem
If clinical denial rates are still climbing despite your Clinical Documentation Improvement (CDI) investment and coder education program, the issue isn’t upstream—it’s that payers have changed how denials are generated on the back end. Commercial payers have deployed machine learning systems to conduct retrospective payment integrity reviews at a scale no human review team could sustain, targeting DRG assignments, medical necessity determinations, and Complication or Comorbidity, or Major Complication or Comorbidity, status on already-adjudicated claims. What looks like a documentation problem from the provider side is, in many cases, an algorithmic output from the payer side.
The Back-End Shift Most Providers Missed
The pattern making this especially difficult to see coming is one Aspirion’s Chief Strategy and Client Officer Jim Bohnsack has tracked closely: payers are announcing reductions in prior authorization requirements—which registers as progress—while simultaneously expanding post-payment medical record review requests.¹ Less friction at the front door. A far more complex dispute concentrated downstream, after discharge, after the account closes, after the clinical team has moved on. That shift—from concurrent review to retrospective audit—transfers the entire evidentiary burden to the provider, against criteria never disclosed before care was delivered.
Outright denials, though, are only part of what’s being lost. The American Hospital Association reported $130 billion in underpayments from Medicare and Medicaid in 2023 alone.² Those errors don’t arrive as denials—they close as zero-balance accounts and disappear into the general ledger, never triggering a worklist or a follow-up.
Why Standard Workflows Break Down
A defensible appeal for a clinical validation denial or DRG downgrade requires navigating thousands of pages of unstructured documentation, mapping findings against InterQual, MCG, or proprietary payer policy, and building a legally coherent argument a physician reviewer won’t easily dismiss. That’s not one job description—it’s four. Historically, roughly 60% of denied claims went uncontested,3 not because providers lacked a winnable argument, but because assembling the expertise cost more than the claim was worth. Payers built their audit strategy around that threshold.
The same breakdown applies to underpayments. Identifying a short payment means extracting contract terms, comparing them against remittance data, and determining recoverability—across every account, every payer, at full claim volume. Teams stretched across denial queues don’t run systematic zero-balance review on top of everything else. Those accounts stay closed, and the revenue stays buried.
How Artificial Intelligence (AI) Restructures the Economics of Revenue Recovery
Large language models change the calculus on both fronts. AI platforms purpose-built for revenue cycle work can ingest denial correspondence and clinical documentation, validate evidence against clinical guidelines and payer-specific policies, and produce structured appeal drafts that clinical and legal staff refine and own. 4In initial blind testing, Aspirion found AI-assisted appeals outperformed those written by experienced human professionals across every evaluated dimension—with a 30% gain in medical record review productivity and a 40-day reduction in placement-to-payment for some health systems. 5 Across its deployments since, Aspirion also reports first appeals generated 2.2x faster, overturn rates improving by more than 10%, and cash in the door 20 days sooner—though outcomes will vary depending on payer mix, denial complexity, and implementation.
On the underpayment side, AI contract modeling extracts payment rules and compares them against billed claims data at scale, surfacing variance that manual review may never reach. Nationwide data shows unknown and underworked denials represent 54% of total recovery opportunity, 6 sitting almost entirely outside the traditional denials management workflow—and recoverable at a scale that manual processes simply cannot reach.
The Real Problem Is Hiding in Plain Sight
This is not a CDI gap or a staffing problem. It is an information-processing competition playing out on three fronts—payers generating clinical denials algorithmically, underpayments closing as zero-balance accounts, and clinical appeals going unfiled because building a defensible case has historically cost more than the return. Industry analyses suggest contract complexity underpayments alone account for up to 10% of claims, 7 with providers losing 1–7% of net revenue annually—before accounting for clinical denial revenue that never reaches the appeals queue. For health systems on sub-5% margins, 8 that cumulative exposure has been normalized—and the only model that addresses all three at scale pairs AI-driven analysis with clinical and legal expertise across the full spectrum of revenue at risk.
Sources
¹ HealthLeaders. “The Winning Edge for Defeating Denials.” healthleadersmedia.com/revenue-cycle/winning-edge-strategies-defeating-denials-source
² American Hospital Association. “Hospitals and Health Systems Squeezed by Persistent Economic Challenges.” aha.org/press-releases/2025-04-30-new-aha-report-hospitals-and-health-systems-squeezed-persistent-economic-challenges
3 HealthLeaders. “Denials at Scale: Using AI to Work Claims Once Thought Unwinnable.” healthleadersmedia.com/technology/denials-scale-using-ai-work-claims-once-thought-unwinnable
4 Aspirion. “Denials Management Services & DocIQ.” aspirion.com/services/denials-management/
5 Aspirion. “Lessons Learned on the AI Journey.” aspirion.com/lessons-learned-on-the-ai-journey-aspirions-innovative-approach-to-complex-clinical-denials/
6 Aspirion. “Why 1% of Revenue Is Worth Fighting For.” aspirion.com/why-1-of-revenue-is-worth-fighting-for/
7 HFMA. “Why AI Is Such a Promising Tool for Eliminating a Hospital’s Revenue Leakage.” hfma.org/ai/why-ai-is-such-a-promising-tool-for-eliminating-a-hospitals-revenue-leakage/
8 KFF. “Hospital Margins Rebounded in 2023.” kff.org/health-costs/hospital-margins-rebounded-in-2023-but-rural-hospitals-and-those-with-high-medicaid-shares-were-struggling-more-than-others/
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
