HFMA 2025 Conference Recap, Part 1: Tackling Denials and Embracing AI

For revenue professionals, the HFMA 2025 Annual Conference in Denver offered a clear message: the pressure on revenue cycle teams from rising denial rates and staffing challenges is real—and growing. 

As a client partner in the healthcare revenue cycle management space, I attended to learn more about these concerns and the momentum around applying technology and AI to respond to these challenges more effectively. This is the first of my three-part series on key HFMA 2025 topics and my observations.

Want to continue the discussion? Schedule a chat with Seth.

Denials and Coding: Still a Top Concern, but AI and Automation Technologies Show Promising Answers

Alicia Soratt of WellStar Health System shared that coding audit discrepancies are a primary source of denials, citing that 85% of HIM professionals identify them as a leading issue. The operational burden is multifaceted: unpredictable payer rules, clinician engagement, limited staff capacity, and identifying root causes with limited visibility. These aren’t new problems—but the urgency to address them is increasing.

One of the more pragmatic sessions, How Autonomous Coding for Physician Claims Streamlines, Increases Satisfaction and Efficiencies,  focused on one possible solution: autonomous coding. The promise here isn't just cost savings, but better clinical documentation, coder satisfaction, and faster reimbursement. A 5% revenue uplift and shortened AR cycles (from 14 to just 7 days) were cited as achievable targets.

This also leads to a shift in how coders spend their time. By automating the bulk of repetitive work, coders are being redeployed into roles that focus on auditing and mentoring, higher-value activities that benefit both operations and job satisfaction.

Another approach included using generative AI.  In No More Business As Usual: How Generative AI is Real and Already Paying Off in RCM, Cleveland Clinic shared a case study where GenAI helped streamline mid-cycle, inpatient coding—an area bogged down by manual review of dozens of documents per case. They partnered with external vendors to implement their approach rather than building it internally, citing both speed and specialization as factors. In Cleveland Clinic’s case, Coders transitioned into validator roles, with AI assisting them in their work, rather than replacing.

UChicago Medicine presented a complementary view in AI Solutions in Revenue Cycle Management: How to Overcome Payment Barriers to Find New Revenue: while their clean claim rate is high (98.23%), their 7.9% denial rate still represents a costly inefficiency. By integrating AI into their pre-claim editing process, they’ve proactively addressed common denial triggers before claims even leave the EHR.

Appeals Automation: High ROI, Low Effort

McKinsey presented compelling data on using GenAI for denial appeals—one of the more under-resourced areas in RCM. With 50% of denials going unappealed due to time constraints, automation led to a 75% reduction in processing time and a 15% increase in cash recovery. 

As someone who focuses on intelligent automation, this tangible example of AI delivering measurable financial impact was compelling.

My Takeaways: Where to Focus Now

1. Start with Root Cause Denial Analytics: Pinpoint your highest-cost, most frequent denial reasons by payer and code. This precision guides where technology can have the biggest impact.
2. Adopt a “Shift Left” Mindset: Fixing issues earlier in the lifecycle (during pre-claim or documentation) is far more cost-effective than managing post-denial.
3. Focus AI Efforts on High-Value Use Cases: Don’t try to “boil the ocean.” Start with areas like coding complexity and appeal generation, where AI has a clear track record of success.
4. Invest in Upskilling Coders: Based on multiple sessions at the 2025 HFMA conference, coders are evolving into validators, auditors, and mentors. Support this transition with training and engagement.
5. Define KPIs and Track Outcomes: Make sure new technologies are aligned with specific, measurable goals—like reducing AR days, denial rates, or increasing net collections.

These items signal a clear shift: denial management and revenue optimization are no longer just about reacting faster; they're about working smarter. As AI and automation continue to mature, providers who take a strategic approach to technology adoption, workforce transformation, and performance measurement will be best positioned to weather rising complexity and build a more resilient revenue cycle.

Want to continue the discussion? Schedule a chat with Seth.