# Using secure artificial intelligence agents integrated within the electronic medical record for the evaluation of blood culture appropriateness—Northern California, 2025

## TL;DR

This paper explores using AI agents in medical records to evaluate blood culture tests, finding high sensitivity but low specificity.

## Contribution

The study introduces secure AI agents integrated into electronic medical records for assessing blood culture appropriateness.

## Key findings

- LLM agents showed high sensitivity but low specificity in evaluating blood culture appropriateness.
- Performance was influenced by prompt phrasing, sycophantic behavior, and semantic triggers.
- The study highlights both the potential and limitations of LLMs in clinical decision support.

## Abstract

We evaluated large language model (LLM)-based agents integrated with the electronic medical record to assess blood culture appropriateness. While sensitivity was high, specificity remained low. Performance was shaped by prompt phrasing, sycophantic behavior, and semantic triggers, reflecting both the potential and limitations of LLMs in real-world clinical decision support.