Examining how information presentation methods and a chatbot impact the use and effectiveness of electronic health record patient portals: An exploratory study

Patient Educ Couns. 2024 Feb:119:108055. doi: 10.1016/j.pec.2023.108055. Epub 2023 Nov 5.

Abstract

Objectives: Examining information presentation strategies that may facilitate patient education through patient portals is important for effective health education.

Methods: A randomized exploratory study evaluated information presentation (text or videos) and a chatbot in patient education and examined several performance and outcome variables (e.g., search duration, Decisional Conflict Scale, and eye-tracking measures), along with a simple descriptive qualitative content analysis of the transcript of chatbot.

Results: Of the 92 participants, those within the text conditions (n = 46, p < 0.001), had chatbot experiences (B =-74.85, p = 0.046), knew someone with IBD (B =-98.66, p = 0.039), and preferred to engage in medical decision-making (B =102.32, p = 0.006) were more efficient in information-searching. Participants with videos spent longer in information-searching (mean=666.5 (SD=171.6) VS 480.3 (SD=159.5) seconds, p < 0.001) but felt more informed (mean score=18.8 (SD=17.6) VS 27.4 (SD=18.9), p = 0.027). The participants' average eye fixation duration with videos was significantly higher (mean= 473.8 ms, SD=52.9, p < 0.001).

Conclusions: Participants in video conditions were less efficient but more effective in information seeking. Exploring the trade-offs between efficiency and effectiveness for user interface designs is important to appropriately deliver education within patient portals.

Practice implications: This study suggests that user interface designs and chatbots impact health information's efficiency and effectiveness.

Keywords: Eye-tracking; Health-information seeking; Inflammatory Bowel Disease (IBD); Patient education; Usability.

MeSH terms

  • Artificial Intelligence*
  • Clinical Decision-Making
  • Electronic Health Records*
  • Humans
  • Patient Portals*
  • Software