Artificial intelligence as a tool for addressing racial and socioeconomic inequalities in cardiac arrest: a systematic review

INTRODUCTION: To determine whether artificial intelligence (AI) interventions have the potential to mitigate racial and socioeconomic disparities in out-of-hospital cardiac arrests (OHCA).OHCA remains a leading cause of mortality globally, with survival rates significantly influenced by systemic healthcare inequalities. Research has identified significant disparities in OHCAs, ethnic minorities have double the risk of OHCA mortality in comparison to Caucasians, and this data has not changed in over 30 years. The AI revolution has captivated physicians with its ability to improve outcomes. The question now pivots to whether it can ameliorate disparities which have been ever present in medical history. This study aims to explore the role of AI in identifying these disparities in the context of OHCA focused on potential benefits and limitations of AI when applied. MATERIALS AND METHODS: A systematic search was conducted in CINAHL and Medline databases, restricted to studies published in the last five years. The search found 164 papers, narrowed to 20 after applying inclusion and exclusion criteria. Seven studies were identified as relevant to the question, focused on AI’s role in addressing disparities in OHCA outcomes. RESULTS: Two overarching themes were identified to encapsulate current research that can be utilised to promote awareness and mitigate ensuring disparities in OHCAs: racial and socioeconomic disparities identified by AI, and implications of AI interventions in OHCA. AI demonstrates the capacity to identify disparities through geographic and demographic predictors but is limited in its ability due to gaps in implementation and ethical programming practices. CONCLUSIONS: The intersectionality of high-risk socioeconomic status and ethnic group minorities in OHCA prevalence is evident, where low-income black sociodemographic contend with incomparable vulnerability. AI presents many opportunities for emergency medicine. However, at present AI should not be implemented as it would only widen disparities and not mitigate due to a lack of relational ethics in the programming process and evolving AI research failing to acknowledge race and socioeconomic characteristics as influential variables on OHCA outcomes.