Human-to-Computer Interactivity Features Incorporated Into Behavioral Health mHealth Apps: Systematic Search

Area of Science:

• Digital health
• Human-computer interaction
• Behavioral science

Background:

• Many behavioral health apps suffer from low user retention, limiting their effectiveness.
• Increasing user interaction types may enhance engagement and app stickiness.

Purpose of the Study:

• To systematically categorize user interactions in behavioral health apps.
• To determine if increased interactivity correlates with user satisfaction and app metrics.

Main Methods:

• A modified PRISMA methodology was used to identify 76 behavioral health apps with interactive features.
• 34 apps were analyzed for 6 types of human-machine interactivities.
• User ratings and app visibility data were collected.

Main Results:

• Apps averaged 2.53 interactivity features, with human-to-data (100%) and human-to-algorithm (44.2%) being most common.
• Human-artificial intelligence interactivity was least common (20.5%).
• No significant association was found between interactivity features and user ratings or app visibility.

Conclusions:

• Behavioral health apps should incorporate more diverse interactivity features to leverage smartphone capabilities.
• Increased user engagement through varied interactivity can maximize therapeutic benefits.
• Developers should explore a wider range of interactive modalities to improve app stickiness and user outcomes.