The healthcare sector is grappling with a pressing question posed by insurers: are artificial intelligence-supported billing tools contributing to soaring costs within the industry?
Technologies such as autonomous medical coding, AI-driven documentation assistance, and ambient listening for notetaking are designed to alleviate the administrative workload in healthcare settings. Nevertheless, recent research from the Blue Cross Blue Shield Association (BCBSA) indicates that these tools have resulted in an additional expenditure of approximately $942 million over two years.
BCBSA points out that this surge in costs isn't matched by a rise in patient treatment levels, which would typically be expected if patients were experiencing a genuine decline in health. Hospitals have increasingly classified inpatient stays as more complex, citing that a larger number of patients are presenting multiple medical diagnoses that necessitate higher reimbursement.
According to BCBSA, this suggests that AI-driven billing innovations might actually be exacerbating healthcare expenses. Luke Chalker, BCBSA's senior vice president of product and data science, emphasized, “If patients are truly sicker, we'd expect to see more treatment.” He noted that there has been a notable increase in diagnoses like anemia at hospitals, yet this hasn’t coincided with a rise in necessary treatments such as blood transfusions. This disparity implies that AI is identifying more billable conditions rather than reflecting the true health status of patients.
On the flip side, developers of these billing technologies argue against this assertion, claiming that their AI systems are simply ensuring that providers receive the reimbursements they deserve according to existing healthcare regulations. Dr. Travis Bias, deputy chief medical officer of health information systems at Solventum, a coding platform utilized by over 80% of U.S. hospitals, acknowledged that he hadn't reviewed the BCBSA analysis in detail but reiterated the importance of the current fee-for-service system, which compensates based on volume rather than value. He noted, "Yes, if you collect more codes and do more procedures, the payments will increase."
The issue of rising costs is not unique to BCBSA, with analysts from PwC projecting a 9% hike in medical costs for insurers next year, attributing it to AI-enhanced billing technologies. Payers are particularly concerned that these systems enable 'upcoding,' where inflated diagnostic codes are submitted for higher reimbursements.
However, Hamid Tabatabaie, CEO of Codametrix—a healthcare tech firm that employs AI for coding across over 500 hospitals—agreed that although costs are rising, attributing these increases solely to AI is misleading. While some costs may arise from the technology, he argues that they are largely due to more complete data capture rather than malpractice. He pointed out, “Last year, if they weren’t using a valid system, they were missing it... They have now addressed that.”
In the current fee-for-service ecosystem, better documentation can lead to increased costs, as excess care is rewarded. This system incentivizes providers to offer more services for revenue while payers try to limit expenses. Tabatabaie argues that the true purpose of medical codes should be to represent clinical realities rather than serve billing purposes, which has allowed payers to streamline claims processing.
Both executives assert that their technologies aim to provide a fuller representation of a patient’s health status, which can also help public health efforts. They emphasize the strong compliance measures in their systems to ensure accurate coding.
Looking ahead, the challenges raised by payers may not persist indefinitely. Bias suggests that shifts in provider behavior and new incentive structures could mitigate some costs linked to AI billing platforms. The healthcare landscape has been shifting towards value-based care since the Affordable Care Act in 2010, which rewards providers based on patient outcomes instead of service volume.
Despite not yet overtaking fee-for-service models, there is a growing recognition that they should be a priority for payers. Enhanced value-based payment systems would reduce the financial burden tied to each submitted code, aligning more closely with medical necessity and appropriate resource use.
As AI technology continues to evolve, both providers and insurers may adapt their approaches to documentation, coding, and payments. Bias indicated that this evolution could significantly decrease administrative inefficiencies, but it will require a systemic overhaul of incentives.
While the discourse surrounding AI's impact on billing remains contentious, it is evident that these technologies are becoming integral to healthcare operations, raising questions about who ultimately bears the financial burden. As increased spending leads to higher premiums, the financial strains extend to employers and patients alike, potentially intensifying administrative challenges such as audits and appeals. Bias has noted that major health systems are currently expending vast amounts merely on disputing denial claims.
As the debate regarding coding accuracy unfolds between vendors and payers, the financial impact continues to accumulate.
