Training an Artificial Intelligence Chatbot to Manage Health in Underserved Populations: A Methodological Approach
ABSTRACT
Background:
Health disparities such as maternal morbidity and mortality amongst childbearing women remain high in the United States, especially amongst those with risks associated with the criminal legal system. Underserved childbearing women, specifically those with criminal legal oversight such as community supervision, are a group that could benefit from mHealth apps. Underserved populations are utilizing mobile health (mHealth) applications on smartphones to access health information more frequently than other potential pathways. MHealth apps that use artificial intelligence (AI) chat-bots may bridge access to care in an efficient and time sensitive manner to manage health in vulnerable populations.
Objective:
Using the Information Systems Research (ISR) framework we discuss approaches for training a chat-bot to enhance its AI capacities. Particularly, we will focus on strategies to address health domains of underserved childbearing women with criminal legal system involvement.
Methods:
Using a case study approach, we applied the ISR framework to refine a MHealth app.
Results:
Our strategies and findings outline in detail how iteratively applying the three cycles (relevance, design, and rigor) of the ISR framework in initial design, implementation, and testing of a mHealth app may enhance the AI chat-bot’s algorithm to provide tailored feedback to address the complex needs of underserved women.
Conclusions:
We illustrate the methodological development of an AI chatbot within a mHealth app to provide health and safety-related support for underserved childbearing women. This methodological approach to designing and testing an AI chat-bot poses possibilities for further development to tailor interventions for populations with risks to their health and safety. Clinical Trial: NCT06636110
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.