When I first walked into the revenue cycle department of a midsize outpatient clinic in Raleigh, I expected to see stacks of patient charts and billing ledgers. What I found instead was a quiet crisis: denial slips stacked in bins, staff wandering between desks chasing payers on the phone and a sense of resigned frustration.
I knew from my research at the University of Pennsylvania and a collaborative partnership with Duke Health, that this wasn’t just busywork, but a growing systemic failure of clinics to recover revenue because they treated claim denials as paperwork instead of as actionable data.
For years I’ve studied how artificial intelligence can help clinics reclaim what is essentially wasted cash and even wrote a bestselling book “Insured to Death” on the topic.
The average claim denial rate in the United States hovers around 15% or more, and for many payers and plans it ranges still higher. That means roughly one out of every five claims submitted is being rejected – despite clinical documentation, prior authorization, and patient eligibility checks.
And even more troubling, many denied claims are ultimately overturned, meaning the claim should never have been denied in the first place.
In our work focused on predictive models for claim denial risk I came face-to-face with the paradox: clinics have data, but they don’t always use it. I saw how workflows that intentionally built feedback loops around denials improved outcomes. At my startup Counterforce Health, we apply those lessons, and we build tools that scale them for smaller clinics that don’t have big revenue cycle teams.
My biggest observation from this work has been that many clinic leaders view denials as an accounting nuisance, a cost of doing business. I think the more interesting angle is to view them as intelligence. Each denial carries metadata – payer, denial reason, procedure code, date, appeal outcome. Yet only a minority of clinics mine that information at scale. According to recent research, 69% of users of AI solutions report reduced claim denials and improved resubmissions.
Clinic leaders should stop treating denial bins like the garbage can and start treating them like gold mines. Because each denied claim that never gets appealed is revenue left on the table. It’s the difference between surviving and thriving in an era when margins are already thin.
Here are the concrete shifts healthcare providers and clinics need to make:
Change the workflow: instead of manual appeals happening when someone notices a denial, build automated triggers at submission. For example, AI can flag high-risk claims before submission (eligibility mismatches, coding anomalies) and mark denials for appeal or corrective action in real time.Build feedback loops: Collect the reasons for denials and tie them into payer-specific playbooks. Patterns will emerge like “this payer rejects this CPT code for this diagnosis” or “this provider type consistently misses this modifier.” Use analytics to feed corrective training.Scale the machine: Many clinics cannot afford a full appeals department. The solution isn’t to hire five more billers but rather to build one AI system that triages claims, generates appeal documentation, attaches supporting records, and routes high-value cases for human review. Research shows these automated steps reduce wasteful spend.Make it “board room material”: When denial rates are confined to back office metrics, they don’t get strategic attention. So, elevate them. C-suite should see “denials = lost revenue” on the same level as “no-shows = lost margin” or “supply waste = lost opportunity.” Fixing denials should be viewed as a growth lever, not a cost center.Lean on partnerships and aggregators: For smaller clinics, access to collective payer analytics, shared benchmarks, and best practices may be the only way to compete. My work at Counterforce Health aims to democratize that access so that non-hospital-system clinics don’t bear the full burden of building their own RCM analytics from scratch.
I think the real revolution in healthcare billing is already underway, as artificial intelligence is no longer a “future” promise but already alive in revenue cycle operations. But many clinics are still asleep, treating appeals as reactive messes instead of proactive opportunities. They chase payers on the phone instead of chasing intelligence in their data.
The question that stuck with me during our research was: what if you could reduce your clinic’s denial rate by even five percent? In many clinics that small shift equates to tens or hundreds of thousands of dollars annually. Why then treat this as optional?
We go through life by telling ourselves stories like ”denials happen,” “that payer always does it,” “it’s just the billing team’s problem.” But a story told long enough becomes a script for failure. It’s time for clinic leaders to re-write the script.
Claim denials are not the cost of doing business, but the cost of ignoring business. Automation amplifies revenues and reduces time spent on the most boring work in healthcare. The clinics that engage AI-powered denial intelligence will not just recover revenue but actually build resilience. And in a system where margins shrink and complexity grows, resilience is everything.
Neal K. Shah is a healthcare researcher specializing in artificial intelligence. He is the Principal Investigator on the Johns Hopkins YayaGuide AI Innovation project and co-Principal Investigator on the University of Pennsylvania’s artificial intelligence for health insurance denials CounterforceAI project. Neal also serves on North Carolina’s Steering Committee on Aging. He is CEO of CareYaya, Chairman of Counterforce Health and the author of Insured to Death: How Health Insurance Screws Over Americans – And How We Take It Back.