The risk of coding racism into pediatric sepsis care: the necessity of anti-racism in machine learning

Machine learning (ML) holds the possibility of improving racial health inequalities by compensating for human bias and structural racism. However, unanticipated racial biases may enter during model design, training, or implementation and perpetuate or worsen racial inequalities if ignored. Pre-existing racial health inequalities could be codified into medical care by ML without clinicians being aware. To illustrate the importance of a commitment to anti-racism at all stages of machine learning, we examine ML in predicting severe sepsis in Black children, focusing on the impacts of structural racism that may be perpetuated by ML and difficult to discover.

source https://www.jpeds.com/article/S0022-3476(22)00339-0/fulltext?rss=yes