High pedagogical quality does not ensure evidence-based grounding in videos disseminated on social media by health professionals.

OBJECTIVE: Social media platforms have become influential environments for health education, patient counseling, and health communication. Short-form videos produced by health professionals are widely consumed and perceived as credible, yet whether their pedagogical quality coincides with evidence-based grounding remains unclear. This study evaluated the pedagogical quality and evidence-based grounding of short-form health education content produced by health professionals on social media and determined the extent to which these two dimensions are associated across professional groups. METHODS: This cross-sectional study analyzed 2000 videos from 100 health professionals (digital health influencers) in nursing, pharmacy, medicine, nutrition, and dentistry on TikTok and Instagram in Brazil, selected by a composite engagement rate. Pedagogical quality was assessed with the ECVIH, an educational-content instrument (18 items, 0-36), and evidence-based grounding with a six-dimension content-analysis coding framework (0-12) derived from established health-information quality criteria. Both were scored as continuous measures by trained reviewers with consensus adjudication and inter-rater reliability testing. RESULTS: Pedagogical scores were uniformly high (profession means 28.2-32.0), whereas evidence-based grounding was lower and more variable (means 7.2-9.3). Pedagogical quality explained only 24.0% of the variance in evidence-based grounding (r = 0.490; 31.8% with profession added), indicating that pedagogical quality is an unreliable predictor despite a positive, moderate correlation. The criterion capturing evidence limitations and uncertainties was among the weakest items, with zero-score rates of 44.64% in nutrition and 36.75% in medicine; even among the most pedagogically polished videos (highest ECVIH quartile), 18.8% provided no such caveats. CONCLUSION: Short-form health content is pedagogically strong, but its educational polish does not reliably predict evidence-based grounding. Polished delivery may obscure evidentiary gaps, fuel health misinformation, and increase patient informational vulnerability. PRACTICE IMPLICATIONS: Professional training and regulatory approaches in digital health education should explicitly incorporate competencies for evidence-based scientific translation.