Racial and ethnic bias found in health algorithms

Algorithms commonly used in electronic health records, clinical guidelines and health care decision tools may introduce bias into care delivery and outcomes, according to an Agency for Healthcare Research and Quality-funded study. Published in JAMA Health Forum, the authors called for increased awareness for clinicians and patients, as well as the development of transparent approaches for algorithm creation and implementation to address racial and ethnic biases and reduce health inequities. The authors undertook a qualitative analysis of 42 responses to AHRQ’s request for information in the Federal Register on potential algorithm-related bias. The respondents identified 18 algorithms in use with the potential for bias. (AHRQ newsletter, 6/1/23)