McDonald · NPJ vaccines 2021 · systematic review and meta-analysis · n=55 studies (32 human studies, 23 non-human primate studies)

Comparative systematic review and meta-analysis of reactogenicity, immunogenicity and efficacy of vaccines against SARS-CoV-2.

Cited 265 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of human clinical trials and preclinical studies

PubMed 33986272 · doi:10.1038/s41541-021-00336-1 · record verified 2026-08-31

What was done

A systematic literature review and meta-analysis was conducted across Phase I, II, and III human trials as well as non-human primate (NHP) studies available up to March 22, 2021. The review evaluated and compared reactogenicity (adverse events), immunogenicity (spike neutralizing antibody titers and T-cell responses), and clinical efficacy across different SARS-CoV-2 vaccine platforms. A total of 32 human studies and 23 NHP studies were included.

What was found

Vaccines across platforms produced mostly mild, self-limiting adverse events. mRNA-1273 and adjuvanted NVX-CoV2373 generated the highest spike neutralizing antibody responses, while ChAdOx-SARS-CoV-2 induced the highest T-cell ELISpot responses. Pre-existing anti-vector neutralizing antibodies halved immunogenicity for AdH-5-SARS-CoV-2. Optimal immunogenicity for mRNA vaccines required booster doses, particularly in older adults. Clinical efficacy was >94% for BNT162b2 and mRNA-1273, >91% for rAd26/5, and >66.7% for ChAdOx-SARS-CoV-2.

Why it matters

This review provides a unified comparative benchmark of the immunological trade-offs, reactogenicity profiles, and protective efficacies of major first-generation COVID-19 vaccine platforms.

Limits

The literature search was restricted to evidence available through March 2021, excluding later real-world post-marketing effectiveness and subsequent viral variant performance. Synthesizing early-phase trials with animal data introduces indirectness, and the abstract does not report confidence intervals or statistical heterogeneity metrics for the meta-analytic estimates.

Cited by