VCU news

By Maggie Christ

Virginia Commonwealth University researchers are examining how artificial intelligence can enhance public health. Sunny Jung Kim, Ph.D., assistant professor of social and behavioral sciences at VCU’s School of Public Health, led a group of researchers in a systematic review of recent studies that assessed the impact of AI-powered interactive technologies in cancer prevention and control as well as substance use. Their results were published on March 1 in Translational Behavioral Medicine.

Co-authors from the School of Public Health include alum Viktor Clark, Ph.D., now at University of Rochester; Vanessa Sheppard, Ph.D., professor and founding interim dean of the school; and Emmanuel Taylor, Dr.PH., professor of health policy.

Their findings showed that AIMC – AI-mediated communications – has promise in enhancing health behaviors, but the researchers outlined the need for further exploration regarding privacy risks, biases, safety concerns, chatbot features and serving underserved populations.

We spoke with Kim for more insight into the findings and the opportunities in further research.

First off, what do AI-mediated communications look like from the patient perspective, and what information can they provide?

AIMC deployed in a technology-based intervention could be presented as a chatbot, an avatar or virtual agents. These AI technologies can empower patients with instant, personalized support and information while also improving engagement and access to care — making health care more interactive, convenient and patient-centered. 

What did this review bring to light?

We explored whether AI-powered interactive technologies could help tackle chronic health challenges like drug addiction and cancer care. We conducted a systematic review to examine how emerging technologies such as chatbots could deliver real, accessible interventions, breaking down barriers like high costs, stigma and social isolation, while making care more reachable for everyone. We found that interventions using AIMC, like chatbots, showed promise in improving health behaviors such as substance use recovery, physical activity and dietary habits, with high retention rates.

What are the current and potential benefits of using AI in public health communications?

AI has the potential to revolutionize public health communications by making them more efficient, personalized and effective, while opening opportunities for innovation in how we tackle chronic diseases, prevent health crises and improve health equity. Fully developed, evidence-based AI could offer personalized messaging, scalable outreach, real-time feedback and valuable insights into our own health behaviors.

What concerns did this study present regarding using AI in this way? What might address those concerns?

We identified a critical need for developing and training AIMC systems with clinically valid and accurate health information. Providing users with incorrect or misleading responses can be harmful when they rely on these tools for reliable health guidance. Ensuring accuracy in AI-mediated health communication is essential to protect patients and support effective decision-making.

Researchers developing AIMC should also prioritize privacy, data safety and confidentiality when handling information exchanged during interactions with AI chatbots or agents. Protecting sensitive health data is essential for maintaining trust and ensuring compliance with ethical and legal standards in health care. Safeguarding this information collected during the interactions builds confidence in AI systems.

We also found that participants in studies on AIMC interventions were predominantly female. To ensure equitable access and better represent diverse populations, we, as researchers, should implement recruitment and engagement strategies that target more representative samples, benefiting all.

What’s next in this area of study?

We’re excited about the next steps in the AI era for health innovation. Our team has developed an aggregated resource base combining clinical guidelines and expert insights for cancer survivors. Building on this, we’ve created a beta version of a chatbot designed to support them. We aim to launch an evidence-based, longitudinal clinical trial that uses interactive AIMC components to offer real-time support, starting with cancer survivors and potentially expanding to other patient populations needing continuous care, self-management and support.

Other co-authors of the systematic review include Jeff Hancock, Ph.D., at Stanford University; Reza Rawassizadeh, Ph.D., at Boston University; and Hongfang Liu, Ph.D., at the University of Texas Health Science Center at Houston.