Virtual Assistants' Response to Queries About Nicotine Replacement Therapy: A Mixed-Method Analysis.

Area of Science:

• Digital Health
• Artificial Intelligence in Healthcare
• Public Health Interventions

Background:

• Smoking cessation remains a global health priority.
• Nicotine Replacement Treatment (NRT) is a key intervention.
• The role of Artificial Intelligent-powered Virtual Assistants (VAs) in delivering health information is emerging.

Purpose of the Study:

• To evaluate the effectiveness of VAs (Amazon Alexa, Apple Siri, Google Assistant) in providing information on Nicotine Replacement Treatment (NRT).
• To assess voice recognition accuracy and response quality of VAs for NRT-related queries.

Main Methods:

• Cross-sectional mixed-method design.
• 40 NRT-related questions posed to 3 VAs, totaling 360 responses.
• Analysis using SAS for inter-rater reliability and response differences; NVivo for qualitative assessment.

Main Results:

• Google Assistant achieved 100% voice recognition, followed by Apple Siri (97.5%) and Amazon Alexa (83.3%).
• Significant differences in responses were noted between Amazon Alexa and the other two VAs.
• VAs exhibited limitations in understanding context, providing in-depth information, and avoiding generic or irrelevant answers.

Conclusions:

• VAs demonstrate potential as tools within smoking cessation interventions and tobacco control programs.
• Improvements in VA contextual understanding, information depth, and response relevance are necessary for optimal integration.