Published by HealthLeaders, November 2025 | Author: Jim Bohnsack, Chief Strategy and Client Officer, Aspirion
Hospitals face a growing threat that’s quietly eroding their financial stability: payers using AI to retroactively deny claims based on clinical appropriateness, even after care has been approved and delivered. According to Jim Bohnsack, chief strategy and client officer at Aspirion, these aren’t simple coding errors—payers are applying advanced analytics to question medical necessity long after treatment, turning approved claims into moving targets.
The Shifting Denial Landscape
While payers publicly claim they’re reducing prior authorization requirements, they’ve actually shifted scrutiny to the backend. They’re requesting more medical records, feeding them into large language models, and applying automated rule sets to deny or downgrade claims retroactively. The result is the same: reduced claim payments. This leaves hospitals scrambling to respond with credentialed experts who must manually review entire medical records to build appeals—a process most providers can’t sustain at scale.
AI Levels the Playing Field
Aspirion developed Doc IQ, an LLM-powered platform that replicates what their nurses and lawyers traditionally did manually. The system extracts clinical evidence from hundreds or thousands of pages of medical records, matches it against payer policies and coding guidelines, and generates comprehensive appeals. The platform began by targeting a specific high-impact denial type: DRG downgrades related to sepsis. After proving success there, Aspirion expanded coverage to all DRGs, multiple disputed codes, and patient type determinations.
Measurable Results
The impact has been substantial and rapid. In January, Doc IQ processed about 2,000 of Aspirion’s 20,000 clinical appeals. By the following month, it handled over 14,000—representing two-thirds of their clinical volume. The platform now covers roughly 70% of clinical denial types, with plans to reach over 90% next quarter.
Clients are seeing concrete improvements: a 15% decrease in appeals required per case, a 10-20% increase in success rates, and approximately 40 days reduced from placement to payment—directly improving accounts receivable days and cash flow.
Implementation Advice
Bohnsack emphasizes that finance leaders should start by clearly defining their problem: increasing throughput or improving success rates. AI solutions don’t need to be fully automated from day one. Aspirion’s approach was incremental—first structuring medical records for easier review, then teaching the system to pre-answer standard questions, validating quality at each step before expanding functionality. This measured approach allows organizations to see ROI quickly while building toward comprehensive automation.
Read the full article here.