Health equity, as defined by the Centers for Medicare and Medicaid Services (CMS), is the attainment of the highest level of health for all people, ensuring everyone has a fair and just opportunity to achieve optimal health regardless of race, ethnicity, disability, sexual orientation, gender identity, socioeconomic status, geography, preferred language, or other factors affecting access to care and health outcomes. The goal is to eliminate disparities and provide comprehensive, inclusive, and effective healthcare to all communities. Akin Demehin, Senior Director of Quality and Patient Safety Policy at the American Hospital Association, declared, “Health equity is a team sport, and health equity is in the DNA of CMS.”
In pursuit of these goals, CMS and the entire healthcare sector are increasingly exploring the use of innovative technologies like artificial intelligence (AI) and large language models (LLMs). These advancements are bringing forth new capabilities that can revolutionize how healthcare is delivered, making it more personalized, efficient, and equitable. The integration of AI into healthcare also brings challenges that must be addressed to ensure these technologies contribute positively to health equity.
At the 2024 CMS Health Equity Conference, held May 29-30, Dr. Kenyon Crowley of Accenture Federal Services, Dr. Ritu Agarwal of John Hopkins University, Deelip Mhaske of the National Minority Quality Forum, and Meagan Khau of the CMS Office of Minority Health spoke on a panel moderated by Demehin on generative AI and LLMs in enhancing health equity. The panel focused on applications, ethical considerations, and how CMS can address programmatic needs through these technologies.
Understanding Generative AI and Large Language Models
Generative AI, with its ability to create new content such as images, code, and text, and streamlining tasks , presents promising opportunities and complex challenges in the healthcare domain. Trained on massive datasets, these models, like ChatGPT, Llama, and Claude, demonstrate immense potential to enhance patient care, assist physicians in research and scientific writing, and augment medical education. By leveraging AI, healthcare providers can generate predictive insights, particularly for underserved populations, driving proactive healthcare interventions and personalized care delivery.
Dr. Crowley emphasized the importance of responsible AI deployment. “Within these potential ideas to use generative AI, responsible AI must be at the core of any of these programs,” he said. “By responsible AI, we mean taking intentional actions to design, deploy, and use AI to create value and build trust by protecting from the potential risks of AI.”
Fostering Transparency and Accountability
In tandem with technological advancements, regulatory frameworks must adapt to ensure the responsible development and deployment of AI in healthcare. Dr. Crowley discussed recent guidelines, such as Executive Order 14091 and accompanying OMB guidance, that underscore the importance of transparency, community engagement, and accountability. By promoting fair and transparent AI practices, CMS can mitigate risks, build trust, and uphold the principles of health equity.
Khau highlighted the December 2023 “Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Healthcare” paper from the Agency for Healthcare Research and Quality (AHRQ). “The first principle of the framework is to promote health and healthcare equity during all healthcare algorithm lifecycle phases.” In terms of health equity, Khau said this means “ensure healthcare algorithms and their use are transparent and explainable.” Khau also stressed the importance of engaging with patients and communities to understand their needs and build trust. “Do you know the needs of your patient, of your communities, of your role? Do they trust you to share their needs without knowing their needs?” she asked.
Ethical Considerations and Mitigation of Bias
CMS serves almost 150 million Americans in one, or more, of its programs and therefore boasts one of the most robust data portfolios in the federal government. Despite these vast datasets, Mhaske, discussed the lack of data for marginalized populations. “There is a vast majority of people throughout the world who are not part of an AI system or the infrastructure at all,” he said. This absence of data means that many global health initiatives fail to address the needs of the most vulnerable. Mhaske’s organization, the National Minority Quality Forum (NMF), aims to change this by increasing clinical participation among minority populations and using AI to predict health outcomes. “If we lose this population and they are not a part of discussion, it’s going to be really another catastrophe,” he warned.
Dr. Agarwal emphasized the need for inclusive data that considers the diverse cultural and linguistic backgrounds of marginalized populations. “Every time this population has been always left behind,” she stated. Dr. Agarwal also discussed the importance of transparency and accountability in AI algorithms to avoid perpetuating biases. “We are still working to improve our collection, data collection on race and ethnicity,” she noted, advocating for more granular data to ensure equitable healthcare.
Collaborative Innovation for Equity
As CMS endeavors to embrace AI-driven innovations, collaboration across sectors and stakeholders is paramount. Initiatives aimed at increasing clinical participation among minority populations, such as those led by the National Minority Quality Forum, exemplify the collaborative spirit needed to address healthcare inequalities globally. By bridging data gaps and ensuring inclusivity in AI deployment, CMS can pave the way for equitable healthcare delivery worldwide.
Moving Forward: A Call to Action
Generative AI and large language models present a transformative opportunity to advance health equity within CMS and beyond. By embracing responsible AI practices, fostering transparency, and prioritizing community engagement, CMS can harness the full potential of AI to drive meaningful change in healthcare delivery. As technology continues to evolve, so too must the commitment to equity, ensuring that AI serves as a catalyst for positive and inclusive healthcare outcomes.
To learn more about the role of generative AI and healthcare, click here.