Comparison of OneChoice AI-based clinical decision support recommendations with infectious disease specialists and non-specialists for bacteremia treatment in Lima, Peru

Comprehensive Summary

This study evaluated the effectiveness of an AI-based clinical decision support system, OneChoice, in guiding antibiotic treatment for bacteremia in Lima, Peru. Researchers compared the system’s recommendations to those made by 94 physicians, including infectious disease specialists and non-specialists, across 366 simulated clinical cases. The results showed high agreement between the AI system and physicians, with a 96.1% match when considering any suggested treatment and 74.6% agreement for the top recommendation. Infectious disease specialists had higher agreement with the AI than non-specialists. The AI system also helped reduce inappropriate antibiotic use, especially unnecessary use of strong antibiotics like carbapenems.

Outcomes and Implications

This study suggests that AI-based clinical decision tools can improve antibiotic prescribing practices and support antimicrobial stewardship, particularly in low-resource settings. By helping standardize treatment decisions and reduce misuse of antibiotics, these tools could play a key role in combating antibiotic resistance and improving patient outcomes globally.