Parental Health Information Needs, Child-Health Literacy, and Use of Generative AI Among Parents of Children Aged 0 to 4 Years in France: A Cross-Sectional Study

Key Findings

• — 12,634 parents responded (5,410 to the needs questionnaire, 7,224 to the literacy-and-AI questionnaire); 92.9% of AI-questionnaire respondents were mothers
• — 54.2% (n=3,919) had used a conversational AI tool for a child-health question: 30.2% once or twice, 12.8% several times per month and 11.2% several times per week
• — Among users, the most common query topic was symptoms and illnesses (59.9%); 29.6% of respondents named nighttime triage as a situation where an AI assistant would be useful
• — Of 3,089 respondents able to report a direction of change, 68.3% had increased their AI use for child-health questions over the preceding 12 months and 6.2% had decreased it
• — Trust most often depended on supervision by physicians (76.6%); among AI users, 66.8% reported no change in how often they consulted a professional and 6.9% consulted less often
• — 35.5% of respondents had raised something with an AI tool that they had shared with neither a health professional nor their personal network, most often about their own well-being (25.2%); the figure was 64.1% among regular users

Methodology Notes

Convenience sample recruited through in-app placements (a marketing card and a full-screen prompt) in the Doctolib app; no incentive; anonymous Typeform questionnaires; the two questionnaires went to distinct respondents and cannot be linked. Questionnaire 1 ran 2026-05-12 to 05-22 and Questionnaire 2 2026-05-22 to 06-11. Items were not piloted or psychometrically validated. Multivariable logistic regression with multiple imputation. All authors are Doctolib employees (one is its Chief Medical Officer); the paper states no personal financial benefit. Self-report, digital-recruitment selection bias and few fathers are stated limitations. Not peer reviewed. Verified from the full PDF on medRxiv on 2026-10-22.

Sources