OHSU reduces coder workload 28% and cuts radiology coding denials 70% with autonomous AI coding

OHSU faced a persistent shortage of medical coders, causing growing case backlogs and unsustainable coder overtime. The backlog led to missed payer timely filing deadlines, resulting in lost reimbursement. Prior computer-assisted coding tools failed to make a meaningful impact, and planned bed expansion threatened to worsen the volume problem further.

OHSU deployed CodaMetrix's AI-driven autonomous coding platform for radiology, fully integrated with their Epic EHR. The system automates end-to-end radiology coding without human intervention for the majority of cases, routing only complex or edge cases to human coders. This eliminated the backlog while freeing coders to focus on higher-complexity work.

The radiology automation rate reached 92%, reducing coder workload by nearly 28%. Coding-related denials for autonomously coded radiology cases dropped 70% compared to manual coding (0.33% vs. 1.09% denial rate). MR case denials — a high-cost imaging category — fell 65%, with an automated denial rate of 0.48% versus 1.38% for manual coding.

• Change management is critical: coders feared job loss from automation, requiring transparent communication that AI handles volume while humans handle complexity.
• Prior failed automation (computer-assisted coding) created institutional skepticism that needed to be overcome before AI adoption could succeed.
• Proving ROI in one service line (radiology) created momentum to expand autonomous coding to additional service lines.