Objectives: Leveraging artificial intelligence (AI) in conjunction with electronic health records (EHRs) holds transformative potential to improve healthcare. Yet, addressing bias in AI, which risks worsening healthcare disparities, cannot be overlooked. This study reviews methods to detect and mitigate diverse forms of bias in AI models developed using EHR data. Methods: We conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines, analyzing articles from PubMed, Web of Science, and IEEE published between January 1, 2010, and Dec 17, 2023. The review identified key biases, outlined strategies for detecting and mitigating bias throughout the AI model development process, and analyzed metrics for bias assessment. Results: Of the 450 articles retrieved, 20 met our criteria, revealing six major bias types: algorithmic, confounding, implicit, measurement, selection, and temporal. The AI models were primarily developed for predictive tasks in healthcare settings. Four studies concentrated on the detection of implicit and algorithmic biases employing fairness metrics like statistical parity, equal opportunity, and predictive equity. Sixty proposed various strategies for mitigating biases, especially targeting implicit and selection biases. These strategies, evaluated through both performance (e.g., accuracy, AUROC) and fairness metrics, predominantly involved data collection and preprocessing techniques like resampling, reweighting, and transformation. Discussion: This review highlights the varied and evolving nature of strategies to address bias in EHR-based AI models, emphasizing the urgent needs for the establishment of standardized, generalizable, and interpretable methodologies to foster the creation of ethical AI systems that promote fairness and equity in healthcare.

Author summary Artificial Intelligence (AI) creates opportunities for accurate, objective and immediate decision support in healthcare with little expert input–especially valuable in resource-poor settings where there is shortage of specialist care. Given that AI poorly generalises to cohorts outside those whose data was used to train and validate the algorithms, populations in data-rich regions stand to benefit substantially more vs data-poor regions, entrenching existing healthcare disparities. Here, we show that more than half of the datasets used for clinical AI originate from either the US or China. In addition, the U.S. and China contribute over 40% of the authors of the publications. While the models may perform on-par/better than clinician decision-making in the well-represented regions, benefits elsewhere are not guaranteed. Further, we show discrepancies in gender and specialty representation–notably that almost three-quarters of the coveted first/senior authorship positions were held by men, and radiology accounted for 40% of all clinical AI manuscripts. We emphasize that building equitable sociodemographic representation in data repositories, in author nationality, gender and expertise, and in clinical specialties is crucial in ameliorating health inequities.