The integration of artificial intelligence (AI) into the U.S. healthcare system is expanding beyond patient diagnosis and treatment, diving deep into administrative functions such as hospital billing, insurance reimbursements, and claims management. While AI has been heralded for its potential to enhance productivity and create efficiencies, emerging evidence is prompting skepticism about whether these technologies might actually escalate costs in an already pricey system.
The Blue Cross Blue Shield Association (BCBSA) recently reported that the incorporation of AI in medical coding has led to nearly $1 billion in added expenses for its health plans from 2023 to 2025. A significant portion of these costs stemmed from secondary diagnoses that qualify patients for higher reimbursement tiers, with BCBSA noting that many of these diagnoses may emerge from isolated lab results, a scenario where AI is particularly effective.
As healthcare providers and insurers embrace AI for billing and claims processing, a critical question arises: can the reduction of administrative burdens lead to heightened financial incentives inherent in the system? Christopher Whaley, a health economist from Brown University specializing in hospital coding, believes AI is making it easier to leverage existing billing incentives, suggesting an acceleration of current trends.
Whaley clarified that while the additional diagnoses identified by AI can often be legitimate, some may not significantly impact patient care yet still allow hospitals to bill for more services. This shift raises concerns about the growing disconnect between clinical treatment and the complexities of coding.
BCBSA's analysis highlighted a striking statistic: approximately 70%, or around $653 million, of the additional billing was linked to new diagnoses that did not result in any changes to patient care, according to Luke Chalker, BCBSA's senior vice president of product and data science. However, Chalker emphasized that AI is just one piece of a larger puzzle. "Various factors drive coding intensity," he noted, suggesting that AI's role, while significant, isn’t the sole contributor.
Chalker also expressed genuine consumer concerns, as heightened coding complexity could lead to increased reimbursement rates without corresponding care improvements. This scenario might ultimately manifest in higher premiums and out-of-pocket expenses for patients. The latest projections from benefits consulting firm Marsh indicate an anticipated average rise of 8.2% in health coverage costs per employee by 2027, marking the steepest increase since 2003.
In contrast, the American Hospital Association (AHA) countered BCBSA’s claims, stating that today’s patients are often more clinically complex, necessitating accurate condition representation for effective care planning. According to the AHA, BCBSA’s assessment fails to account for the broader implications of AI on healthcare quality, patient access, and overall spending.
"It's particularly concerning for insurers to question provider coding while employing automated methods that can deny necessary coverage," an AHA representative commented, pointing out potential administrative inefficiencies caused by such practices.
Chalker acknowledged that Blue Cross Blue Shield companies also utilize AI in claims processing, assuring that all clinical denial decisions involve a review by qualified clinicians.
Whaley described the growing use of sophisticated AI tools within the healthcare sector as an "administrative arms race," noting that these technologies, while costly to implement, do not directly enhance patient care delivery. The financial burden of these advanced systems often translates into increased costs for consumers, including elevated premiums and taxes.
Marisa Greenwald, a partner at Oliver Wyman's Health and Life Sciences division, which provides consultancy on strategies related to AI and operations for hospitals and insurers, argued that AI is delivering tangible advantages to healthcare systems. By minimizing administrative tasks, physicians can dedicate more time to patient interactions and regain a semblance of work-life balance. Greenwald noted that health systems are leveraging AI not only for efficiency improvements but also to enhance the overall work experience for doctors.
