Abstract Background COVID-19 vaccines provide strong protection against severe outcomes but do not fully prevent infection, and the extent to which they attenuate disease severity among breakthrough infections remains unclear. Existing evidence is limited by small sample sizes, potential selection bias, and inconsistent findings. Methods We conducted a systematic selection and meta-analysis of studies identified from a living systematic review of global observational studies reporting vaccine effectiveness (VE) against symptomatic infection and hospitalization. We computed the relative VE (rVE) against hospitalization versus symptomatic infection for matched VE pairs, representing vaccine-induced attenuation of severity. Multilevel meta-analysis and meta-regression were used to pool estimates and assess effect modifiers. Results A total of 184 VE pairs from 37 studies were included. The pooled rVE was 51% (95% CI: 39%, 60%) for broadly defined symptomatic infection and 19% (95% CI: 12%, 25%) for medically attended symptomatic infection, indicating significant attenuation of disease severity among vaccinated cases. Given substantial heterogeneity, interpretation focused on stratified and adjusted analyses. Attenuation was lower in older adults aged ≥60 years and showed no significant waning over time. Variant patterns differed by endpoint definition: attenuation was weaker for Omicron subvariants when broadly defined symptomatic infection was the comparator. No significant differences were observed between vaccine platforms. Conclusions COVID-19 vaccination markedly reduces the risk of progression from symptomatic infection to hospitalization. Attenuation of severity was sustained over time and did not differ significantly by vaccine platform, but was lower in older adults and varied by symptomatic-infection definition and variant period.
Liu et al. (Mon,) studied this question.