AI in Health Care: Closing the Equity Gap or Widening It?

Author: Dr. Shakira J. Grant

February 25, 2025

A Startling Reality: When AI Gets It Wrong

Imagine being diagnosed with cancer and needing essential tests and support services. But instead of receiving the required resources, an AI system erroneously determines that you are not among those who would benefit most. This is not just hypothetical—flawed AI decision-making has already impacted actual patients, highlighting the urgent need for equitable and accountable AI in health care.

In a striking example from the U.S., a widely used AI-driven risk calculator was found to disproportionately deny Black patients access to high-risk care management programs because it based predictions on prior health care spending rather than actual medical need.[1] Similarly, a sepsis detection AI-tool deployed across two American hospitals, failed to identify the condition in 67% of cases where timely treatment could have been life-saving.[2]

The Role of AI in Transforming Health Care

AI is revolutionizing health care, offering improved diagnostics, streamlined operations, and personalized treatments. But an urgent question remains: will AI bridge existing health disparities or deepen them?

The Potential of AI to Improve Health Equity

AI is making health care more inclusive and accessible through key innovations:

·       Early Disease Detection: AI-powered radiology improves accuracy in detecting diseases like cancer at early stages, leading to better outcomes.

·       Diverse Clinical Trials: AI helps match patients to suitable clinical trials, increasing diversity and improving the reliability of medical research.

·       Personalized Medicine: By analyzing genetic and clinical data, AI tailors treatment plans to individual patients, optimizing health outcomes.

These tools promise to eliminate traditional barriers and enhance global health care accessibility.

The Challenges: Can AI Unintentionally Perpetuate Bias?

While AI holds immense promise, it can also reinforce existing health disparities if not implemented carefully:

A man is profiled by an AI-powered robot on a busy city street, showcasing contemporary surveillance technology in urban settings.

·       Bias in AI algorithms: AI models trained on non-representative data can perpetuate systemic biases, leading to disparities in diagnostics and treatment recommendations.

·       Lack of diversity in training data: AI tools may be less accurate for underrepresented populations, as seen in cases where algorithms struggle to detect conditions like melanoma (skin cancer) in darker skin tones.

·       Risk of marginalization: AI-driven decision-making in areas such as insurance approvals, hiring, and patient triage may inadvertently exclude vulnerable communities from essential health services.

·       High implementation costs: The expense of adopting and maintaining AI-driven tools can be prohibitive for lower-resourced health care systems, potentially widening the gap in access to advanced medical technologies.

Ethical and Regulatory Considerations

As AI becomes more integrated into health care, key ethical and regulatory issues arise:

• Accountability: Who is responsible when AI-generated diagnoses are incorrect?

• Transparency: How can we ensure AI decision-making is clear and fair?

• Policy safeguards: What regulations are needed to prevent AI from deepening inequities?

Solutions: How AI Can Be Designed for Equity

To ensure AI advances health equity, we must take intentional steps:

·       Inclusive AI Development: Train AI models using diverse, representative data sets to minimize bias and improve accuracy for all populations.

·       Ethical Governance: Establish clear frameworks for transparency, fairness, and accountability in AI decision-making.

·       Community Engagement: Actively involve policymakers, researchers, and affected communities in AI development and deployment to ensure solutions address real-world needs.

·       Equitable Resource Allocation: Create funding mechanisms and public-private partnerships to support AI adoption in lower-resourced health care settings, reducing cost barriers.

·       Workforce Training: Educate health care professionals on AI biases and best practices to ensure AI tools enhance care rather than reinforce disparities.[3]

Looking Ahead: A Call to Action

AI in health care presents an unparalleled opportunity to create a more equitable system—but only if we develop and deploy it responsibly. The challenge remains: How do we ensure AI benefits everyone, not just a select few?

At CROSS Global Research & Strategy, we are dedicated to studying AI’s impact on health equity and finding ways to use it as a force for good. We invite you to be part of this mission. Explore our research, engage in discussions, and schedule an initial consultation with us to help shape AI-driven solutions that serve all communities fairly.

Want to stay on top of the latest insights in health care AI? Join our community to receive updates on our blog series and other resources.

Subscribe Here

References

1.         Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. Oct 25 2019;366(6464):447-453. doi:10.1126/science.aax2342

2.         Wong A, Otles E, Donnelly JP, Krumm A, McCullough J, DeTroyer-Cooley O, Pestrue J, Phillips M, Konye J, Penoza C, Ghous M, Singh K. External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients. JAMA Intern Med. 2021 Aug 1;181(8):1065-1070. doi: 10.1001/jamainternmed.2021.2626. Erratum in: JAMA Intern Med. 2021 Aug 1;181(8):1144. doi: 10.1001/jamainternmed.2021.3907. PMID: 34152373; PMCID: PMC8218233.

3.         Matheny ME, Goldsack JC, Saria S, et al. Artificial Intelligence In Health And Health Care: Priorities For Action. Health Affairs. 2025-02-01 2025;44(2):163-170. doi:10.1377/hlthaff.2024.01003

Image Credits

Images licensed from Adobe Stock.

Logo for CROSS Global Research & Strategy, depicting hands cradling a heart with a medical cross, symbolizing care and health-focused initiatives.