AI-Driven Coding is Driving Costs Up.  How Can Payers Adapt?

Across nearly every sector in the economy, artificial intelligence is rapidly reshaping organizational business models, and the healthcare industry is no exception. Healthcare’s embrace of AI has boosted its efficiency and accuracy, particularly with clinical and administrative workflows. One area of impact is medical coding: whereas before AI, it was primarily human-driven and slower by necessity, it now helps providers document care with greater specificity and consistency than ever before.   

From a clinical and administrative perspective, AI’s automation offers clear benefits. There are fewer manual errors, fewer appeals and adjustments on claims and documentation that more accurately reflects the variable complexity of patient health. While the clinical potential of generative AI is still unfolding, its coding tools have fundamentally shifted the economics of reimbursement, resulting in complex claims paid out with greater speed than payers have ever experienced and creating new dynamics for how payers manage reimbursement and cost oversight. 

Why AI-driven coding accuracy is raising risk scores and costs

To understand why payers are experiencing rising costs as a result of artificial intelligence, we first must clarify the role AI plays in medical coding. Essentially, the primary impact of generative AI is not what providers are billing for, but the accuracy and efficiency with which they can do it. Historically, medical coding was limited by human capacity, requiring multiple rounds of review, denials and resubmissions that influenced the speed of reimbursement.  

The impact on cost emerges when AI’s efficiency is applied across thousands of patient-provider encounters. AI tools can auto-generate evaluation and management codes, identify missed charges and augment diagnosis specificity, potentially impacting risk scores. Even modest changes in risk assessment can materially affect payer reimbursement and overall costs. As AI-enabled coding increasingly adjudicates and reimburses claims on the first pass, often at higher levels, payers face a growing challenge: traditional cost-containment mechanisms are unable to keep pace.  

How payers can adapt workflows and oversight

With the surge in billing from AI-enabled coding, payers may feel increasing pressure to rely more heavily on denials, payment delays and prior authorizations. While these reactionary responses may appear to slow spending in the short term, they can negatively impact the member experience. Reactive measures also risk creating friction in provider-payer relationships without addressing the underlying factors that drive higher reimbursement.  

Standard review processes struggle to keep up with the demand created by AI, increasing the potential for reactive responses. Payers can better meet demand by adapting the current workflow structure and oversight strategy for the AI era.    

Payers can begin by simply defaulting to the basics; it is more vital than ever to review the entire claims adjudication stack and make sure the appropriate prepayment edits are turned on. This approach means payers working closely with their claims system team and vendors to update workflows. Data and analytics also offer a practical approach to uncovering billing behavior.      

Employ analytics to determine what is driving billing behavior

By employing data and analytics, payers can assess patterns of higher-acuity coding levels rather than focusing on claim-by-claim review. The elevated cost burden that payers are experiencing should be analyzed at the source. Is it the result of AI improving documentation, or is it concentrated among specific providers? A detailed review will identify where these patterns occur, providing insight that helps payers address cost drivers proactively.

Additionally, while AI-enabled billing may be technically accurate, it does not necessarily faithfully reflect the clinical encounter. As providers adopt AI-enabled tools for charge capture and documentation, some billing may result from automation rather than provider intent. This is where a strong fraud, waste and abuse (FWA) system is critical. When payers invest in FWA detection tools, consistent updates and alerts will analyze and flag potential issues, along with implementing workflows to capture additional FWA.

Focus on value-based care  

Once workflow, oversight and analytics are incorporated into payer practice, the resulting data will highlight patterns among specific automated billing codes, laying a foundation for more informed provider conversations, grounded in facts rather than assumptions. Further, these discussions present an opportunity to shift toward value-based care incentives if the data support it.    

In many cases, data will reveal a need for clarification and education over punitive measures.  AI is still relatively new in healthcare and constantly evolving; if payers leverage insights from new data and billing patterns from providers, they can approach negotiations with providers with the goal of collaborating to ensure members are receiving the best and right care to optimize member outcomes from the start.

Photo: Andriy Onufriyenko, 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.