There are numerous ongoing and emerging threats to health equity, such as climate change, artificial intelligence, lack of access to insurance, and disparities in exposure to adverse childhood experiences which can impact health for a lifetime. The P4HE Collaborative emphasizes the importance of understanding these threats to health equity and addressing them through cross-sector collaboration. This blog examines the intersection of AI and health equity and discusses how different collaborators can work together to address the challenges these technologies present to health equity.
Predictive technologies, such as machine learning and artificial intelligence (AI), have the potential to significantly advance health equity by enabling early detection of opioid overdoses, enhancing syndromic surveillance, and improving monitoring of foodborne illnesses. However, to fully harness this potential, it is crucial to address the issue of algorithmic bias.
Algorithmic bias occurs when AI systems, trained on historical data, make decisions that reflect the prejudices inherent in that data, leading to unfair outcomes, particularly for Black communities. This can perpetuate and even amplify existing disparities in healthcare. For example, an algorithm used to predict patient risk scores was found to systematically underestimate the health needs of Black patients compared to White patients. Another instance is the use of race-based correction factors in clinical algorithms, which can result in patients of color receiving different treatment recommendations than their white counterparts. Additionally, AI systems used in medical imaging have shown higher error rates for women and people with darker skin tones.
For AI to become a reliable tool in promoting health equity, it is essential to use diverse and comprehensive training data sets that accurately represent all populations, including those historically underrepresented in public health efforts. By understanding and mitigating these biases, we can ensure that the benefits of AI and machine learning are equitably distributed, thereby minimizing the risks and fostering a more just healthcare system.
While not perfect, AI can be an important diagnostic tool. Technological innovation for health has enormous promise for health equity and medicine more broadly. For example, tuberculosis (TB) is the deadliest infectious disease, infecting an estimated 10.8 million people in 2023. The U.S. Centers for Disease Control and Prevention (CDC) has leveraged AI to assist with the fight against TB by employing it to analyze lung x-rays in low-resource settings. Without the use of AI in these areas, these cases may have gone undiagnosed due to a lack of radiologists to manually screen these x-rays. Another application of AI for health predicting protein structure predictions, which can help address health equity by improving protein design for malaria vaccines for more equitable global health or detecting osteoporosis earlier to improve women’s health as they age.
To effectively overcome the challenges posed by predictive technologies and advance health equity, various sectors can work together in the following ways:
- Advocates can push for comprehensive policies aimed at ensuring transparency and fairness in the use of AI and machine learning.
- Researchers can work towards developing unbiased algorithms and ensure diverse data sets that accurately represent communities facing health inequities are used.
- Communities can raise awareness about the potential biases in predicative technologies and advocate for fair practices.
- Philanthropists can support initiatives that promote ethical AI development and ensure equitable use of these technologies.
- Policymakers can pass regulations that require transparency and fairness in AI applications.
By working together, these groups can harness the potential of AI while minimizing its risks.
Interested in learning more?
- Subscribe to our P4HE Collaborative Newsletter to get notified when we post our resources from the webinar, Overcoming Threats to Health Equity: Today and Tomorrow.
- Stay tuned to our Learn page for updates on our upcoming programming.