| Sumario: | Background & Significance: Prostate cancer (PC) is the most frequently diagnosed cancer among men in the United States. PC survivors who are often older adults face complex decisions about treatment while navigating barriers such as limited health literacy and restricted access to care. Artificial intelligence (AI)-powered chat bots provide timely and personalized information but raise concerns about accuracy and trustworthiness. Few studies have applied frameworks to evaluate chat bot communication in oncology. Johnson's Comprehensive Model of Information Seeking (CMIS) highlights how antecedents and information carrier characteristics shape evaluations, while the Technology Acceptance Model (TAM) emphasizes usefulness and ease of use in adoption. Previous literature has focused on telehealth or mobile health, leaving no framework to guide understanding of AI-powered chat bot communication in PC care. Purpose: This study develops a theoretical framework integrating CMIS and TAM to explain how healthcare professionals for PC patients evaluate AI chat bot communication compared with physicians. Using this framework, we aim to map the existing literature applying CMIS and TAM framework for cancer survivors including PC. Methods: This study employs a scoping review design and uses a theory-building approach to to adapt and integrate CMIS and TAM into a unified framework for evaluating communication facilitated by AI chat bots and physicians. Key constructs will be drawn from prior applications of CMIS in cancer information seeking and TAM in technology adoption. Guiding by Arksey and O'Malley's framework, we will systematically searched PubMed, CINAHL, PsycINFO, and Web of Science for studies applied either CMIS or TAM in cancer survivors. Findings and Interpretations: In figure 1, the integrated CMIS-TAM framework positions antecedents as precursors to information needs, which inform evaluations of accessibility, appropriateness, clarity, comprehensiveness, and trustworthiness. Attitudes reflect perceptions of physicians versus chat bot responses, while action is conceptualized as comparative evaluations and behavioral intentions toward adopting these sources. This review will map prior cancer research that applied CMIS and TAM. Discussion: This integrated framework provides a novel foundation for assessing information seeking characteristics for decision-making in PC. Our empirical testing will evaluate how healthcare professionals judge chat bot versus physician generated responses to PC related questions. The framework advances theoretical application in oncology and informs the design of AI tools are clinically meaningful.
|
|---|