Meta-analyses are powerful tools for synthesizing evidence on vaccine effectiveness, though results may be affected by heterogeneity and bias. This study examined the effectiveness of the Janssen COVID-19 vaccine in preventing coronavirus disease across multiple regions. A meta-analysis of 20 studies involving 39,321 participants (19,630 vaccinated; 19,691 placebo) was conducted. Standardized mean differences were pooled using both fixed- and random-effects models. Heterogeneity was assessed with Cochran’s Q and the I² statistic, while potential publication bias was evaluated using Egger’s regression test and a funnel plot. Substantial heterogeneity was observed among studies (Q = 85.45, p < 0.001; I² = 77.8%), indicating variation in study outcomes. The random-effects model produced a pooled effect size of 0.008 (95% CI: –0.011 to 0.027), suggesting no consistent overall effect. Egger’s test (p = 0.240) and funnel plot inspection revealed no evidence of publication bias. Although the pooled effect was negligible, the high heterogeneity indicates that vaccine effectiveness likely varies by setting, population, and circulating variants. Importantly, absence of publication bias strengthens the reliability of the analysis. In the broader context of vaccine literature, the Janssen vaccine has been shown to reduce hospitalization and severe outcomes, particularly among older adults, though its effectiveness is lower than that of mRNA vaccines. These findings underscore the need for cautious interpretation of pooled estimates and highlight the importance of context-specific vaccine policies. Further subgroup analyses are recommended to clarify sources of variability and guide targeted immunization strategies.
Isyaku et al. (Mon,) studied this question.