Elevance Health AI Slashes Medical Record Review Time by 50%

**TL;DR**: An AI tool developed by Elevance Health's Carelon Payment Integrity division cut medical record review time by 50%, saving 20 minutes per case across nearly 23,000 annual reviews.

Fraud, waste, and abuse cost the US healthcare system tens to hundreds of billions of dollars annually. These inefficiencies range from intentional fraud—such as providers billing for services never rendered—to unintentional waste like duplicate testing. Identifying these issues requires painstaking review of medical records and claims data, work that traditionally consumes significant human hours.

Elevance Health's Carelon Payment Integrity division deployed an AI program that reduced average medical record review time from 40 minutes to 20 minutes per case. The tool processed nearly 23,000 reviews this year, saving approximately 380 human hours annually. The AI identifies patterns that human investigators often miss, including subtle billing anomalies and coding discrepancies that suggest potential upcoding or unnecessary services.

Matt Glynos, Vice President of Carelon Payment Integrity, said AI identifies patterns humans miss while human investigators provide context—combining both protects members and reduces costs. This hybrid approach allows the AI to flag potential issues rapidly while human reviewers apply judgment to determine whether flagged cases represent genuine fraud, acceptable variation, or documentation errors.

The 50% time reduction addresses a key bottleneck in payment integrity operations. Human reviewers can now investigate more cases without expanding staff, potentially increasing the total fraud, waste, and abuse identified each year. However, the tool's impact on actual cost recovery remains to be measured. The efficiency gains demonstrate AI's ability to augment rather than replace skilled investigators, handling initial screening while humans focus on complex adjudication.

Healthcare payers facing similar resource constraints may evaluate comparable AI-assisted workflows. The broader question—whether increased detection efficiency translates to proportional savings—will shape future investment decisions across the industry.