A Meta Analysis of Social MEDIA COVID-19 Vaccine Misinformation and COVID-19 Vaccine Hesitancy

Introduction: Even prior to the COVID-19 pandemic, the World Health Organization (WHO) had identified reluctance to vaccinate as a major global health threat, listing it among the top ten concerns in 2019 (WHO, 2019). Vaccine hesitancy or anti vaccine sentiment fueling through social media platforms being one of the cardinal limiting factors of the successful implementation of Covid-19 vaccination program worldwide.

Objective: The objective of this study was to synthesize data from ten cross-sectional empirical studies to measure how exposure to Covid-19 vaccine misinformation from social platforms influences Covid-19 vaccine hesitancy rates by quantifying both the strength and consistency of this association across multiple research contexts.

Methods: The total sample size consisted of 27,483 participants (age ≥18) were across ten countries (United Kingdom, Italy, Pakistan, Nigeria, United States of America, Greece and Canada). A random-effects models calculated pooled odds ratios (ORs) to understand between study and withing study variability, with subgroup analyses by country, sample size, mean age, gender and study quality. Test of Heterogeneity was conducted using Cochran Q and I² statistic in STATA for Windows, version 17 (STATA Corp). Publication bias was assessed using Egger's test Begg’s test and trim-and-fill analysis. A meta regression analysis was performed to understand the how moderators (e.g., sample size, demographic factors, and study quality) might explain the substantial heterogeneity observed in the association between social media misinformation and vaccine hesitancy.

Results: The meta-analysis found that social media misinformation significantly increased COVID-19 vaccine hesitancy (OR=1.62, 95%CI[0.73-2.51], p<0.001), with extreme heterogeneity (I²=99.9%). Effects were strongest in Nigeria (OR=5.20) and younger populations (OR=5.20, age <45). Study quality scores were found to marginally moderate effects (β=15.00,p=0.079). No publication bias was detected (Egger's p=0.941; Begg's p=0.721). The results from the meta-regression analysis revealed that study quality scores marginally moderated the association (β = 15.00, p = 0.079), with higher-quality studies showing stronger effects, while other covariates (sample size, age, gender) did not significantly explain heterogeneity (all p > 0.05), leaving 80.8% of variance unaccounted for (I² = 96.68%). Findings remained robust in sensitivity analyses.

Conclusion: The findings of the meta-analysis confirmed that social media misinformation increases COVID-19 vaccine hesitancy significantly, particularly among younger populations in specific regions who are more prone to social media usage. The robust association reinforced the urgent need for evidence based targeted digital public health messaging to counter vaccine misinformation, while future research should include important contextual moderators to explain residual variability in effects.