CodaMetrix and five major health systems have formed a council to address inconsistencies in how coding quality is measured across healthcare organizations. The initiative, led by CodaMetrix’s director of coding applied research, Jamie Noorlander, and Monica Watson, corporate coding director at Allegheny Health Network, aims to create a unified standard for evaluating coding accuracy, whether performed by AI or human coders. The council, described as a governance body, brings together subject-matter experts from Allegheny Health Network, Mayo Clinic, Henry Ford Health System, Oregon Health & Science University, and CU Medicine, reflecting a collaborative effort to standardize definitions and metrics across the industry. Noorlander, who previously worked in coding at academic medical centers, emphasized that the initiative stems from early observations with CodaMetrix’s customers, who noted significant variations in how quality is defined, measured, and audited across health systems. These discrepancies, she explained, often stem from differences in audit scope, cadence, and the focus on human versus AI-driven coding processes, creating a gap in industry-wide benchmarks.
The council includes coding leaders from CU Medicine, Mayo Clinic, Henry Ford Health System, Allegheny Health Network, and Oregon Health & Science University. Watson, who has spent 25 years in medical coding, emphasized that discrepancies in definitions and metrics have long been a challenge. “If we don’t know what we’re calculating or what the expected outcomes are, we can’t ensure technology achieves them,” she said. Noorlander noted that early CodaMetrix customers shared concerns about varying audit scopes and quality benchmarks, which created a gap in industry-wide standards. Watson, who oversees coding operations for Allegheny Health Network’s inpatient, outpatient, emergency room, and professional organization, highlighted the importance of standardizing definitions to ensure that coding accurately reflects patient care. CodaMetrix, which partners with over 30 leading health systems across 27 states, operates within a network representing more than $191 billion in net patient revenue, 30 million patients, and 120,000 physicians, showing the scale of its impact.
Disparities in measurement could lead to disputes between payers and providers, Watson warned. “Having AI tools gives us a chance to level the playing field,” she said. The council seeks to define industry-wide metrics that both AI vendors and payers can adopt, reducing friction between coding and payer systems. Noorlander added that expanding the framework to payers could further align expectations and improve collaboration. Watson noted that current disputes between payers and providers often arise from inconsistent definitions of coding quality, a challenge that AI adoption could help mitigate. She explained that AI tools, by providing consistent benchmarks, could enable health systems and payers to work from a shared understanding of quality, supporting more transparent and efficient interactions. This alignment, she argued, is critical as both payers and health systems increasingly integrate AI into their operations.
Related: Dietzman Succeeds Nakashima as eHealth Exchange Leader
The council’s initial focus is on radiology, with plans to refine the quality framework through feedback from medical coders and professionals. Noorlander explained that the council is starting with radiology, the most mature service line within CodaMetrix, to pilot the quality framework. A white paper study is underway, inviting volunteer coders to participate and provide input on the framework’s effectiveness. This iterative process, she emphasized, is essential to ensuring the framework remains adaptable and responsive to evolving industry needs. Watson also noted that the council’s efforts are not limited to AI but aim to standardize measurement across all coding platforms, including human-driven processes, reflecting a full approach to quality assurance.
But are you starting to see some efficiency gains so that people are able to do more with AI assistance? Watson affirmed that AI has not only increased productivity but also encouraged coders to move beyond routine tasks. She explained that coders often develop muscle memory for repetitive tasks, leading to complacency, but AI tools prompt them to engage in more critical thinking and problem-solving. Noorlander echoed this, noting that successful AI adoption hinges on a shift in mindset among customers, who must view AI not as a tool for error detection but as a partner in optimizing workflows. She emphasized that health systems with strong medical management teams and a culture of innovation tend to achieve the most success, as they collaborate closely with AI vendors to refine models and improve outcomes. The council’s quarterly meetings, she added, are designed to continuously vet and refine the quality framework, ensuring it evolves in tandem with industry advancements and stakeholder feedback.
