Algorithmic bias

Digital Health and AI

Definition. Algorithmic bias is a systematic difference in how an algorithm performs or allocates resources across groups, producing unfair outcomes for some patients.

Explanation

Bias can enter through unrepresentative training data, biased labels, or proxies that encode social disadvantage. A widely cited example showed that an algorithm using past costs as a proxy for need underestimated the illness of Black patients (Obermeyer et al., 2019). Theories of digital medicine stress that such effects are built into data and design choices, not only into code (Yang et al., 2026).

Fair machine learning techniques can reduce bias in real-world health data, but no single method suits all settings (Huang et al., 2024). Health organisations should test performance by subgroup, document data sources and keep human review for high-stakes decisions.

Related concepts

Artificial intelligence (AI) Machine learning Health equity AI governance Data governance

Read on The Health Topics

References

Huang, Y., Guo, J., Chen, W.-H., Lin, H.-Y., Tang, H., Wang, F., Xu, H., & Bian, J. (2024). A scoping review of fair machine learning techniques when using real-world data. Journal of Biomedical Informatics, 151, 104622. https://doi.org/10.1016/j.jbi.2024.104622

Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019). Dissecting racial bias in an algorithm used to manage the health of populations. Science, 366(6464), 447–453. https://doi.org/10.1126/science.aax2342

Yang, Y., Liu, X., Cuyubamba Dominguez, J. L., Fang, Y., Xie, W., Shen, B., & Siau, K. L. (2026). Analysis theories on artificial intelligence, ChatGPT, data science, and metaverse: The case of digital medicine. Journal of Organizational and End User Computing, 38 (1). https://doi.org/10.4018/JOEUC.400122

Cite this entry: Kurutkan, M. N. (2026). Algorithmic bias. In The Health Topics glossary. https://healthtopic.org/glossary/algorithmic-bias/

← Back to the glossary