Artificial Intelligence and Automated Note Systems Increased Insurance Costs

According to recent research published by the Blue Cross Blue Shield Association, artificial intelligence systems and automated note-taking technologies used in the healthcare sector have increased secondary disease billing, raising insurance companies' costs by nearly $1 billion.

Research Results and Costs

According to the study published by the Blue Cross Blue Shield Association, healthcare providers billing for more secondary diseases and additional health problems generated an additional cost of $653 million for companies affiliated with BCBSA in the years 2024 and 2025.

The research stated that the impact of cases noted as more complex and requiring intensive care compared to 2023 reached a total cost of $942 million over the two-year period.

The Relationship Between Artificial Intelligence and Billing

Healthcare organizations have started recording more secondary disease diagnoses thanks to artificial intelligence systems that scan patient records and technologies that listen to doctor-patient conversations to automatically generate notes.

These diagnoses, which refer to additional health problems accompanying the primary treated illness, lead to the patient's condition being recorded as more complex and prompt insurance companies to make higher payments.

Disconnect Between Diagnoses and Treatments

BCBSA Senior Vice President of Product and Data Science Luke Chalker stated that if patients were truly in a more severe condition, they would expect to receive more treatments.

However, according to the research, it was determined that despite the recording of more complex diagnoses, there was no corresponding increase in certain treatments.

Expert Opinions and Examples

BCBSA Vice President of Clinical Affairs Dr. Razia Hashmi said that although the number of patients undergoing bowel surgery or diagnosed with anemia increased, common treatments such as blood transfusions did not rise to the same extent.

Luke Chalker evaluated this situation as suggesting that the disconnect between diagnoses and treatments indicates artificial intelligence is identifying more billable conditions rather than sicker people.