Bridging the Gap: Evaluating Publicly Available Education on SSRI Use in Pregnancy in the Age of Artificial Intelligence ID 3729

INTRODUCTION: In the context of declining literacy rates and July 2025 FDA scrutiny of selective serotonin reuptake inhibitor (SSRI) use in pregnancy, this study compares the readability of relevant patient education across medical societies, educational websites, and social media, and evaluates large language models (LLMs) as tools for improving patient understanding. METHODS: Patient-facing content on SSRI use in pregnancy from the American Academy of Family Physicians, the American College of Nurse-Midwives, the American College of Obstetricians SD 1.25), and ACN-M’s (17.50, SD 1.33), which were all college graduate level. Responses from Gemini (13.99, SD 1.37) and Chat GPT (15.3, SD 1.37) were more readable than ACOG FAQs (10.79, SD 1.72). The ACOG FAQs rewritten by LLMs were the most readable (Chat GPT 8.95, SD 1.18; Gemini 8.78, SD 1.69), although still not at the requested sixth-grade level. CONCLUSIONS/IMPLICATIONS: More readable patient education on SSRI use in pregnancy is needed, as most Americans read at or below a seventh-grade level. Large language models may be useful in addressing this gap.