Using conversational agents to explain medication instructions to older adults

Area of Science:

• Human-Computer Interaction
• Gerontology
• Health Informatics

Background:

• Developing effective computer agent (CA) based systems for patient-centered communication is crucial, especially for older adults receiving medication instructions.
• Understanding how CA design and message framing influence patient responses is essential for optimizing health information delivery.

Purpose of the Study:

• To evaluate older and younger adults' responses to medication information presented by computer agents (CAs) with varying appearances and message framing.
• To assess the impact of CA realism, expressiveness, and message framing (gains vs. losses) on patient memory, affect, and perception of teaching effectiveness.

Main Methods:

• 360 older and younger adults were exposed to medication information delivered by CAs differing in appearance (photorealistic, cartoon, emoji) and message framing (gain vs. loss).
• Evaluated outcomes included patient memory for medication information, affective responses, perceived CA teaching effectiveness, expressiveness, and similarity.

Main Results:

• Computer agents (CAs) significantly impacted older adults' learning of medication information.
• Social responses engendered by CAs played a key role in enhancing learning outcomes for older adults.
• CA appearance and message framing influenced patient perceptions and memory, suggesting design implications for health communication systems.

Conclusions:

• Computer agents (CAs) can be effective tools for delivering medication information to older adults, particularly by leveraging social interaction principles.
• Tailoring CA design and communication strategies to elicit positive social responses can enhance patient understanding and adherence to medication instructions.