Payment Integrity is a Systems Problem, not a Data Glitch

A man recovering from heart surgery at home is shocked when he finds out his insurance has denied coverage for a critical part of his hospital stay. When he looks into what happened, he finds a billing code that indicates his operation was elective. But it wasn’t — he was admitted through the ER with severe chest pain. He was told surgery was the only option. 

So, he calls the hospital. They confirm the procedure was indeed medically necessary. Yet, because of a single incorrect entry — one that no clinician flagged while he was in the hospital — the system now treats his life-saving care as “elective.” 

The error only cascades from there. Payment is denied, appeals are triggered, and now, the man recovering from surgery is caught between clinical truth and administrative fiction. From a payer’s perspective, the claim looks sound. The codes are all technically correct, and the documentation checks out. 

The issue only surfaced when the patient scrutinized his bill. 

Now, multiply this one incident across millions of claims. The result is not just administrative overhead or confusion — it’s systemic inefficiency. Time is wasted correcting data, money is spent resolving preventable disputes, and trust is diminished between payers, providers, and patients. Despite this, many continue to respond by layering on more checks or outsourcing to vendors who promise to find mistakes after they’ve happened.

This approach only addresses symptoms, not root causes. It highlights a deeper structural issue: our healthcare system’s dependence on fragmented, reactive processes rather than proactive system design.

Identifying the problem 

The primary antagonist behind claims inaccuracy isn’t bad behavior or ill-intentioned misclassification. The culprit is fragmented systems that don’t communicate well. As our healthcare system has evolved to provide higher quality care to more — and increasingly clinically complicated — people, the disconnect between what’s billed, what’s documented, and what’s actually true has deepened.  

That’s why administrative waste in healthcare still exceeds $1 trillion annually, despite decades of digitization and vendor optimization.

Historically, the upstream resolution of payment errors has been seen as unsolvable. The system’s complexity has notoriously made automation difficult. There are more than 700 Diagnosis-Related Group (DRG) categories, each with their own layered severity and pricing logic. Medicare alone operated over 30 payment programs last year, and billing rules vary widely between hospitals and health plans. Add inconsistent clinical documentation and ambiguous policy language, and the result is the same: manual reconciliation of information that should have aligned from the beginning.

For years, the primary argument has been that these problems are data-centric; that better data, more audits, or even more codes will solve payment integrity challenges. But no volume of raw data can fix a fundamentally flawed process. Data without aligned, intelligent workflows just creates more noise. 

Health plans have never had the tools to leverage noisy data intelligently at scale. But technology has changed. Today’s AI systems can understand language, follow policy logic, and evaluate complex clinical and contractual data in real time. Just like you’d want a second opinion from a doctor before a major procedure, patients and payers alike deserve a system that double-checks critical decisions before they create problems downstream. 

The remaining barrier is cultural, not technical. Too many organizations still assume that payment integrity must be reactive. Disputes are treated as inevitable. Errors are something to fix later rather than prevent now. But that assumption is outdated.

Adopting a proactive approach to payment

With lightspeed advances in AI, healthcare providers and health plans can now have the tools to ensure payment accuracy from the start. Intelligent systems can be trained to understand the full picture of a member’s care and billing journey, from what their policy says to what their record documents to what a contract dictates. Humans will always remain an essential part of the process, with AI enabling fast approvals and teeing up potential inconsistencies with the relevant context for human experts to proactively address. 

It’s not about replacing people. It’s about giving clinicians and claims teams the equivalent of a real-time second opinion — one that doesn’t just spot errors, but can prevent them from ever impacting the patient experience.

As we move from an era of “intelligence scarcity” to “intelligence abundance”, we have an opportunity to rethink how we can harness AI second opinions for the greater good.  

For the heart surgery patient, an AI-driven system would have flagged the erroneous elective procedure code immediately, comparing it against clinical documentation, admission type, and policy rules. The inconsistency would have been caught and corrected before the claim was ever submitted, preventing a costly denial, a prolonged appeals process, and a deeply stressful experience for someone trying to get healthy.

These AI systems work by integrating data streams — clinical, financial, policy — and applying advanced logic continuously, not retrospectively. By aligning these inputs up front, they enable the ecosystem to “get it right the first time,” avoiding costly cycles of denials and rebilling. 

This shift doesn’t require a total overhaul of the system. It requires applying existing rules clearly and consistently leveraging AI as a tool for intelligence amplification and improved accuracy. 

Building these technologies is a heavy lift, but the most daunting requirement will be cultural. Organizational alignment is required to allow information to flow across departments and systems. Silos — whether data, departmental, or process-driven — must be eliminated. Clinical and financial logic should operate together, not in isolation.

That’s how we move from a reactive payment system to a proactive one. By ensuring the accuracy of what goes into the system, we remove the need to clean up what comes out of it — and eliminate the fear, confusion, and waste associated with incorrect claims.

Photo: lbodvar, Getty Images

This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.