Artificial intelligence generated visual communication improves comprehension and adherence in cervical cancer...

Area of Science:

• Medical Informatics
• Health Communication
• Artificial Intelligence in Healthcare

Background:

• Cervical cancer screening faces challenges with Pap smear result comprehension and follow-up adherence, especially in low health literacy populations.
• Georgia's screening coverage below 20% necessitates innovative communication strategies.
• Artificial intelligence (AI) offers potential for enhanced patient communication via adaptive, visually expressive tools.

Purpose of the Study:

• To assess if AI-generated visual explanations combined with simplified text improve comprehension, satisfaction, and follow-up adherence post-cervical cancer screening.
• Comparison against conventional text reporting methods.

Main Methods:

• A randomized controlled trial involving 3,000 women aged 21-65 undergoing Pap smear testing.
• Participants were allocated to three groups: Control (standard text), Text-only (plain language), and Intervention (AI visuals + text).
• Outcomes measured included comprehension, satisfaction, and follow-up adherence, analyzed using chi-square tests and logistic regression.

Main Results:

• The AI visual intervention group showed significantly higher comprehension (90%) and satisfaction (90%) compared to Text-only (78%) and Control (65%) groups.
• Follow-up adherence was 75% in the AI group, a threefold increase in odds compared to Control (50%).
• Improvements were most notable for participants with abnormal Pap smear results (ASCUS, LSIL, HSIL).

Conclusions:

• AI-generated visual communication is a powerful tool for improving cervical cancer screening outcomes.
• This informatics solution demonstrates scalability for patient engagement but requires further research on long-term impact and integration.
• Challenges include adapting AI visuals for diverse populations and seamless incorporation into existing health systems.