Government health agencies are increasingly turning to real-world data (RWD) for public health innovation and exploring the transformative potential of artificial intelligence (AI) to improve public health outcomes. Tracy Hammonds, Ph.D., Director of Real-World Data Insights at HealthVerity, has a unique vantage point on this trend. Dr. Hammonds, who earned her Ph.D. in Experimental Psychology from Kent State University, leads initiatives that bridge real-world data with government programs. In a recent discussion, she noted that agencies across the U.S. Department of Health and Human Services (HHS) are cautiously evaluating AI to enhance operational efficiency, disease surveillance, and program integrity. Government agencies are eager to leverage AI for faster insights, but remain vigilant to ensure accuracy and public trust.
Dr. Hammonds’s insights highlight how major public sector efforts are harnessing AI and real-world data (RWD) for innovation: the Food and Drug Administration’s (FDA)push to streamline drug approvals, the National Institutes of Health’s (NIH) strategy to democratize AI for researchers, the Centers for Disease Control and Prevention’s (CDC) use of RWD to strengthen national HIV surveillance, and ARPA-H’s initiative to target maternal health and heart disease through data-driven interventions. Each case underscores the growing value of real-world data and AI applications – and how platforms like HealthVerity Marketplace and taXonomy provide the high-quality data foundation needed to realize these ambitions.
ARPA:-H: Targeting public health challenges with RWD
A striking example of AI and real-world data driving public health innovation comes from ARPA-H. The Advanced Research Projects Agency for Health is a new agency within HHS focused on accelerating high-impact health solutions. One of its first programs addresses severe obstetric complications and heart disease, two critical issues affecting underserved populations.
Working with HealthVerity and MITRE, ARPA-H developed a regional analytics platform using monthly real-world data reports at the ZIP3 level. These reports track complication rates and heart disease risk by geography and payer type. As Dr. Hammonds explained, “We worked with ARPA-H on a project around severe obstetric complications and heart disease. The goal was to identify at-risk areas and fund local solutions, so we built a platform that broke down regional health data by complication rates and payer type.”
"The goal was to identify at-risk areas and fund local solutions, so we built a platform that broke down regional health data by complication rates and payer type.”
- Dr. Hammonds, Ph.D. Director of Real-World Data Insights
Public health agencies use this data to apply for performance-based funding, proposing targeted interventions to improve outcomes. HealthVerity’s de-identified and linked datasets support ongoing monitoring to evaluate each program’s success. In essence, the agency created a “health accelerator” model: it invested in local solutions, but tied payments to measurable results. As Dr. Hammonds recounted, “local public health agencies used that data to apply for grants…Now the data is being used not just to launch programs, but to monitor performance.”
If agencies meet their benchmarks, they receive continued funding. This outcome-driven model reflects a new approach to public health that uses AI and real-world evidence to guide smarter, more accountable interventions.
CDC: Advancing HIV surveillance with real-world data
Another powerful example of HealthVerity real-world data at work in the federal landscape comes from the Centers for Disease Control and Prevention (CDC). In a newly published peer-reviewed study in Clinical Infectious Diseases, CDC researchers leveraged HealthVerity data to evaluate HIV testing patterns among individuals prescribed cabotegravir (CAB-LA), a long-acting injectable for preexposure prophylaxis (PrEP).1 The study is the largest U.S. real-world cohort of CAB-LA PrEP users to date, with follow-up periods extending up to seven months.
CAB-LA represents a major innovation in HIV prevention, but its clinical benefits come with surveillance challenges. Its mechanism of action can suppress viral replication, which may delay HIV detection and create risks around diagnosis timing and resistance. To mitigate this, the CDC's 2021 guidelines recommend a combination of antigen/antibody and RNA testing throughout CAB-LA use.
To understand whether these recommendations were being followed in practice, CDC researchers used HealthVerity Marketplace data, drawing from longitudinal prescription, laboratory, and payer records across more than 20 major sources. The richness of this integrated dataset enabled them to track real-world testing behaviors at scale, helping to assess how clinical guidance is being translated into patient outcomes.
This type of longitudinal, multi-source RWD is crucial for national surveillance programs. It allows agencies like the CDC to monitor prevention strategies in action, flag gaps in testing practices, and protect vulnerable populations with data-informed guidance. This work illustrates the growing role of public health surveillance data in ensuring that emerging prevention strategies like CAB-LA are effectively monitored in real-world settings.
As Dr. Hammonds noted in our discussion, agencies are actively exploring how RWD can “enhance disease surveillance and prediction” across public health domains. The CDC’s CAB-LA PrEP study underscores how high-quality data assets, like those available through HealthVerity, can accelerate these goals, equipping researchers and policymakers with the tools to monitor and optimize critical health interventions.
NIH: Simplifying AI for researchers with low-code tools
While the FDA focuses on regulatory uses of AI, the National Institutes of Health (NIH) is concentrating on making AI accessible to a broad community of researchers. NIH leadership recently released strategic goals for AI implementation that place a strong emphasis on accessibility. A core theme is the development of low-code or no-code AI platforms that empower scientists who may not be software engineers. In other words, NIH wants,“to make research more accessible to the researchers and the scientists… so they’re looking at how to deploy AI with low code or no code solutions for research,” as Dr. Hammonds observed.
"[The NIH wants] to make research more accessible to the researchers and the scientists… so they’re looking at how to deploy AI with low code or no code solutions for research."
- Dr. Hammonds, Ph.D. Director of Real-World Data Insights
This push for accessibility aligns closely with HealthVerity’s recent partnership with Medeloop, innovators in AI-driven analytics. The strategic collaboration combines the extensive HealthVerity real-world data ecosystem with Medeloop’s advanced, no-code AI analytics platform, empowering researchers and healthcare professionals to directly pose complex research questions and generate actionable insights rapidly without requiring programming expertise. Medeloop’s intuitive platform translates simple written queries into comprehensive research studies in minutes, significantly accelerating NIH’s vision for accessible, impactful research.
FDA: Using AI and real-world data to accelerate drug approvals
One prominent example of government action is the FDA’s new draft guidance on AI in healthcare. In January 2025, the FDA issued its guidance for using AI to support drug and biologic development.2 This draft guidance lays out a risk-based framework for sponsors to assess and ensure the credibility of AI models in the regulatory process – a critical step toward using AI to speed up drug reviews without compromising safety or efficacy. In practical terms, the FDA AI in government healthcare helping to analyze the massive datasets (including RWD from sources like electronic health records and insurance claims) that inform drug approvals.
According to Dr. Hammonds, government agencies tend to be conservative in adopting AI at first – “Because they don’t want to put out wrong information”. However, the FDA stands out for its willingness to innovate. It is “particularly eager to streamline the drug approval process using AI”, Dr. Hammonds explained. “The FDA’s goal with AI is to make the approval process more efficient and get quicker insights from data so they can get drug products out faster,” she said.
“The FDA’s goal with AI is to make the approval process more efficient and get quicker insights from data so they can get drug products out faster."
- Dr. Hammonds, Ph.D. Director of Real-World Data Insights
At the heart of the FDA’s guidance is the message that data quality and traceability determine the trustworthiness of AI-driven outcomes. This is where HealthVerity provides critical support. Through the HealthVerity Marketplace, FDA scientists and sponsors can tap into a vast, privacy-compliant repository of real-world data. The Marketplace offers access to more than 75+ distinct data sources and 340+ million de-identified patients in a single, unified environment. Equally important, every dataset’s provenance is transparent and auditable – a key requirement in the FDA’s framework for AI credibility.
Our solution: Building a foundation for AI-powered public health
Across the FDA, NIH, CDC, and ARPA-H examples, a common thread is clear: trusted, representative real-world data is the foundation for meaningful public health innovation. Whether accelerating drug approvals, enhancing disease surveillance, democratizing research, or funding outcome-based programs, government leaders are increasingly recognizing that credible data is what enables smarter, faster, and more equitable decisions.
Dr. Hammonds’s experiences reinforce this point. The role of HealthVerity was to provide that strong data foundation, enabling AI and analytics to produce credible, actionable results. The HealthVerity Marketplace and taXonomy are key tools in this space, giving agencies a convenient way to access high-quality, linked and synchronized data at scale. In fact, HealthVerity Marketplace is the nation’s largest healthcare and consumer data ecosystem and taXonomy is the most comprehensive claims solution for real-world evidence.
By using these platforms, public sector organizations can obtain HIPAA-compliant data on hundreds of millions of patients. Transparency, interoperability, and timeliness are all included under a single, secure contract.
If you want to learn more, please reach out to the Dr. Hammonds and the government team here: https://healthverity.com/government-request/
References
- Zhu W, Delaney K, Huang YLA, Patel RR, Kourtis AP, Hoover KW. Real-world human immunodeficiency virus rna and antigen/antibody testing among people who use long-acting injectable preexposure prophylaxis. Clinical Infectious Diseases. Published online April 3, 2025.
- US Food and Drug Administration. Considerations for the use of artificial intelligence to support regulatory decision-making for drug and biological products. Draft Guidance for Industry. Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological January 2025.
