AI arms race leading to prior auth problems, reimbursement cuts

AI worsens existing prior auth problems

There are fundamental issues with prior authorizations that need to be addressed before AI can optimize the process, the key stakeholders agreed. The application of AI on both sides has only exposed these "deeper structural limitations that technology alone cannot resolve," the report stated.

Prior authorization has become a top use case for AI in healthcare, and the report indicated that real-time authorization at the point of care is an emerging model. However, current proofs of concept are still narrow and not yet scalable.

Still, AI-driven prior authorizations are reducing process costs for individual organizations, though not for the overall system. Providers reported AI supporting medical necessity justification drafting, submission form completions and appeal generations, while payers have used AI to triage requests and provide decision support.

AI is allowing providers to "submit more complete requests with less effort," and payers to process more submissions at a lower cost per decision. However, evidence is lacking that the average cost per claim is now lower.

"As such, participants raised concern that optimizing each side of the transaction risks making the overall process more activity-intensive, rather than more efficient," the report explained.

Participants recommended alternatives to prior authorization that could help meet utilization management goals. Rather than layering technology on a flawed system, they suggested prepayment review or offering a discounted payment rate for providers to bypass upfront review requirements.

AI for medical billing inflates healthcare spending

PHTI reported that a key takeaway is that provider use of AI for medical billing is increasing coding intensity, leading to inflated healthcare spending.

AI scribes and automated coding tools are enabling providers to more completely capture patient complexity while modestly saving them time on documentation.

However, a pair of studies released last month confirmed that AI is intensifying coding and, therefore, revenue capture, at least for maternity admissions and hospital outpatient care. One study even estimated an additional $2.3 billion in healthcare spending due to more aggressive billing practices enabled by AI.

Payers are noticing this shift in provider medical billing and responding by across-the-board downcoding and other reimbursement cuts, the PHTI report said. They reportedly use AI themselves to identify and automatically adjust outlier high-complexity codes for evaluation and management (E/M) services that appear inconsistent with clinical documentation.

The report also indicated that payers are reducing reimbursement for certain modifiers and comparing submitted E/M levels with those of peer providers with similar patient populations. Although PHTI noted that these payer responses vary and are not yet well documented.

These strategies may widen the gap between the haves and have-nots, the report added. As reimbursement decreases in response to AI-driven billing, providers who have not adopted the tools could be disproportionately harmed.

Stakeholders agreed that reimbursement policy is "the strongest level to drive administrative efficiencies and system-level cost savings." They suggested possible policy solutions, including requiring the disclosure of AI tools for coding and implementing oversight frameworks (e.g., cost growth targets, audits).

However, future research is critical to understand the impact AI has on medical inflation, PHTI stressed.

Jacqueline LaPointe is a graduate of Brandeis University and King's College London. She has been writing about healthcare finance and revenue cycle management since 2016.