AI Won’t Save You from Bad Claim Strategy

Last week, EnableComp Chief Strategy Officer Mike Esworthy laid out a clear and compelling framework for tackling one of the most persistently underestimated problems in healthcare finance: the drag of complex claims. These are the 10 to 15 percent of encounters that exist outside the logic of routine reimbursement. They resist automation, evade standard EHR logic, and disproportionately strain revenue cycle teams already operating with fewer resources and tighter margins.

Esworthy’s argument is both a technical and a strategic one. His call to action reframes complex claims from operational nuisances into a central lever of performance. To address them is not to fix a broken process, but to seize a neglected opportunity.

That opportunity is growing more urgent. In its latest analysis, Kaufman Hall reports that median hospital margins remain below pre-pandemic levels for a fifth consecutive quarter, with revenue cycle teams struggling to do more with fewer FTEs. Labor constraints, inflationary pressure, and shifting payer behavior are not temporary conditions. They are the new baseline. Yet even under these conditions, too many hospitals continue to pour resources into standard claim paths while treating edge cases as afterthoughts.

This misalignment is expensive. The American Hospital Association estimates that administrative inefficiencies, especially those tied to eligibility errors, misrouted claims, and payer miscommunication, cost U.S. hospitals nearly $39 billion annually. Claims involving Veterans Affairs, workers’ compensation, and TRICARE are among the most error-prone. They demand highly specific documentation and vary wildly in processing rules. Denials are not just more frequent. They are more difficult to reverse and more likely to result in delayed or lost reimbursement.

The natural temptation is to throw technology at the problem. But as McKinsey observed in its 2024 generative AI deployment guide, general-purpose tools underperform in high-complexity environments. It is not the sophistication of the AI that matters. It is its relevance. Systems that understand payer nuance, denial code specificity, and historical claim patterns will outperform generic solutions every time.

This is where Esworthy’s framework gains traction. He argues not for more automation in general, but for automation grounded in domain-specific intelligence, or what some call context-augmented generation. This allows complex claims to be flagged, routed, and resolved with precision instead of trial and error. But his broader point is about structure. To manage complexity, hospitals need to stop thinking in terms of bolt-on fixes. They need to architect around complexity as a core part of their business model.

That means designing intake processes to detect non-standard payer scenarios early. It means outsourcing the 10 percent of claims that devour 40 percent of RCM bandwidth. And it means measuring performance not by generic KPIs, but by the yield, velocity, and resolution rate of the hardest claims in the portfolio.

In short, complex claims are not just a workflow issue. They are a strategic frontier. And for health systems with the discipline to approach them as such, they may represent the most untapped source of margin recovery in modern hospital operations.