The healthcare technology sector has spent the better part of the last decade propagating a deceptively simple narrative: medical billing is a rule-bound, repetitive process ripe for total disruption. Startups and software vendors routinely pitch the healthcare revenue cycle as an administrative assembly line. Because claims follow rigid guidelines, payers adhere to established policies, and back-office tasks involve predictable patterns, the industry is often told that algorithms should be able to run the show from end to end—from clinical documentation to final reimbursement—with minimal human intervention.
However, industry leaders operating at the bleeding edge of healthcare delivery offer a significantly more nuanced perspective. Speaking at the Healthcare Financial Management Association’s (HFMA) annual conference in National Harbor, Maryland, Todd Manion, revenue cycle chair at the prestigious Mayo Clinic, challenged the prevailing Silicon Valley ethos of total replacement. While Manion and his team at Mayo are actively harnessing artificial intelligence and machine learning to streamline administrative bottlenecks, he draws a hard, unambiguous line at full-scale automation.
The core limitation, according to Manion, lies in the unyielding gap between human clinical complexity and the rigid, highly structured data requirements demanded by automated systems. Medicine is an art as much as a science; physicians document nuances, clinical impressions, and subtle pathophysiological findings that resist neat translation into standard billing codes.
This deep-dive report examines the friction between tech-driven automation and clinical reality, explores where AI is genuinely transforming the revenue cycle at institutions like Mayo Clinic, and outlines the strategic philosophy that human expertise remains irreplaceable when translating patient care into financial truth.
Detailed Chronology: The Evolution of RCM Automation and the HFMA Disclosures
The Rise of the "Plug-and-Play" Billing Fantasy
For years, digital health venture capital has flooded into Revenue Cycle Management (RCM) automation. The pitch to hospital CFOs drowning in labor shortages and compressing operating margins was irresistible: deploy proprietary machine learning models to ingest electronic health records (EHRs), auto-generate claims, predict denials before they happen, and drastically reduce administrative headcount.
Early iterations of robotic process automation (RPA) successfully tackled low-hanging fruit—such as batch eligibility verification and basic patient scheduling reminders. But as generative AI and large language models (LLMs) matured, tech vendors pushed the envelope further, promising end-to-end autonomous coding and zero-touch claims processing.
The Reality Check at HFMA 2026
It was against this backdrop of escalating hype that Manion took the stage at the HFMA annual conference. In a candid interview session, he provided a reality check derived from managing the vast, highly complex billing operations of one of the world’s premier integrated health systems.
Manion did not discount the transformative power of modern software. Instead, he systematically deconstructed the myth of the fully autonomous revenue cycle, illustrating how clinical documentation workflows inherently clash with payer adjudication rules. By highlighting real-world friction points—such as the linguistic disconnect between treating a patient for pneumonia versus documenting a "pulmonary infiltrate"—Manion underscored why algorithms cannot simply bridge the gap without human validation.
The disclosures at HFMA marked a pivotal shift in the broader industry discourse: moving away from the utopian dream of a "lights-out" revenue department and toward a pragmatic, hybrid model where AI serves as a force multiplier for human intelligence rather than a wholesale substitute.
Supporting Context & Metrics: The Friction Between Clinical Language and Administrative Codes
To understand why full automation stumbles, one must examine the operational mechanics of how patient care transforms into hospital revenue.
The Semantics Gap: Clinical Nuance vs. Payer Adjudication
In clinical practice, physicians are trained to document observations with maximum precision, prioritizing patient safety, diagnostic clarity, and continuity of care. They use granular terminology that reflects the dynamic, uncertain nature of human disease.
Conversely, the administrative and financial machinery of healthcare operates on a binary, highly codified system. Payers do not reimburse for narrative descriptions; they reimburse for specific, validated codes (such as ICD-10 for diagnoses, CPT for procedures, and HCPCS for supplies).
During his remarks, Manion shared a classic, highly illustrative scenario that plays out in hospital billing offices nationwide every single day:
The Clinical Reality: A patient is admitted with severe respiratory distress. The care team administers every evidence-based medication, respiratory therapy, and monitoring protocol specifically indicated for pneumonia.
The Documentation: The attending physician documents the condition in the progress notes as a "pulmonary infiltrate" rather than explicitly writing "pneumonia."
The Administrative Dead End: Even though the clinical evidence overwhelmingly points to pneumonia, and even though the payer can read the exact same medical record, the coder’s hands are legally and procedurally tied.
Payers operate under strict regulatory and contractual guidelines requiring an explicit, signed diagnosis from a licensed clinician, entered into designated structured fields within the EHR, to substantiate a claim.
"Sometimes I think there’s a misunderstanding that the entirety of the medical record can be used to, yes, treat the patient — but unless that diagnosis is in a specific place, we can’t apply it to the claim without then going back to the provider and querying," Manion explained during the HFMA session.
This administrative reality shatters the illusion of seamless end-to-end automation. An algorithm can read the phrase "pulmonary infiltrate," but without human clinical judgment to interpret intent, evaluate context, and initiate a provider query, an automated system will either generate a denied claim or prematurely halt processing.
The True Cost of Administrative Friction
The disconnect between clinical documentation and billing codes is not merely a bureaucratic nuisance; it is a multi-billion-dollar drag on the U.S. healthcare system. According to recent industry analyses:
Denial Rates are Climbing: Initial claims denial rates across U.S. health systems have steadily hovered near historical highs, frequently exceeding 10% to 15% for complex inpatient encounters.
The Burden of Rework: Hospitals spend billions annually on administrative overhead dedicated solely to reworking, appealing, and resubmitting denied claims—much of which stems from documentation queries and code mismatches.
Staff Burnout: Revenue cycle professionals spend countless hours navigating payer portals, waiting on hold for claim status updates, and playing administrative ping-pong between clinical teams and insurance adjusters.
Official Statements and Strategic Insights: Where AI Belongs in the Revenue Cycle
While Manion pumped the brakes on full-scale automation, he was careful not to dismiss artificial intelligence altogether. Under his leadership, Mayo Clinic is deploying targeted, pragmatic AI solutions that generate measurable return on investment—not by replacing human judgment, but by eliminating mind-numbing administrative toil.
Automating the Repetitive, Elevating the Human
At Mayo Clinic, AI has found its true calling in high-volume, low-complexity workflows where algorithms excel. These include:
Claim Status Tracking: Automatically querying payer clearinghouses and portals to determine where a claim sits in the adjudication pipeline.
Remittance Flagging: Identifying outstanding remits and categorizing payment discrepancies.
Aged Accounts Receivables (AR) Follow-Up: Monitoring claims that have lingered past contractual payment windows and generating automated preliminary alerts.
In the past, these routine tasks required skilled revenue cycle personnel to sit on hold with insurance companies for agonizing periods. Today, automation absorbs that friction.
"I don’t need people waiting on hold to figure out where a claim’s status is with the payer," Manion remarked. "There are simplistic tasks that are repetitive that we’ve used AI to simplify so that we can elevate our people toward more complex patient issues."
By stripping away repetitive administrative chores, Mayo is effectively upskilling its workforce. Staff members are freed from mechanical data-retrieval duties and empowered to apply critical thinking, negotiation skills, and clinical acumen to complex denials, intricate payer contracts, and escalated patient advocacy issues.
Redefining the Ultimate Goal of RCM
Perhaps the most profound takeaway from Manion’s industry insights is his fundamental redefinition of what revenue cycle management is actually supposed to achieve.
For many organizations, the RCM department operates like a collection agency—relentlessly chasing claims, fighting off denials, and plugging revenue leaks in a reactive game of whack-a-mole. Manion rejects this adversarial, transactional framework.
To him, the true North Star of revenue cycle management is deceptively simple yet profoundly challenging to execute: accurately reflecting the care that was actually delivered.
"If we can do that and do it accurately, everything else falls into place," Manion declared.
This philosophy shifts the entire paradigm. When RCM is viewed as an accurate mirror of clinical reality rather than a financial weapon designed to maximize yield, the relationship between clinical care, documentation, and billing harmonizes. It validates why human oversight is indispensable; an algorithm can optimize for financial extraction, but only a human clinician and professional coder working in tandem can ensure that the financial record bears an unassailable, truthful witness to the healing work performed at the bedside.
Future Outlook: The Hybrid Horizon of Healthcare Finance
As health systems nationwide grapple with persistent labor shortages, wage inflation, and tightening operating margins, the pressure to adopt automation will only intensify. Technology vendors will continue to iterate, promising ever-smarter LLMs capable of interpreting complex clinical narratives.
However, the insights shared by leaders like Todd Manion at Mayo Clinic establish a vital roadmap for sustainable innovation over the next decade:
The Death of "Full-Auto" Claims: Healthcare organizations that chase the elusive dream of 100% "lights-out" revenue cycle automation will likely encounter severe compliance risks, widespread claim rejections, and damaged payer relations. Clinical complexity resists wholesale algorithmic compression.
Rise of the Augmented Professional: The future belongs to hybrid operating models. AI will act as an indispensable co-pilot—handling status checks, clearinghouse pinging, and predictive denial analytics—while human professionals retain control over clinical validation, provider queries, and high-stakes appeals.
Documentation-First Strategies: Forward-thinking health systems will increasingly invest upstream, focusing on real-time Clinical Documentation Improvement (CDI) at the point of care. By helping physicians capture clinical nuance accurately before the chart is closed, hospitals can minimize the downstream administrative friction that requires heavy intervention later.
Cultural Realignment: RCM leaders will pivot away from adversarial collections metrics and embrace Manion’s philosophy: measuring success by the fidelity with which financial records reflect true patient care.
In the end, healthcare finance is fundamentally tethered to the human condition. While algorithms can process the rules of the game, the art of medicine will always require human wisdom to translate care into coin. Institutions that recognize this boundary will harness the true power of AI without falling victim to its overpromises.
