Long-term immunity following yellow fever vaccination: a systematic review and meta-analysis

Long-term immunity following yellow fever vaccination remains controversial. We aimed to summarise the literature regarding the long-term protection (≥10 years) conveyed by a single dose of yellow fever vaccination.

In this systematic review and meta-analysis, we searched 11 databases from database inception to Aug 24, 2023. We included cohort and cross-sectional studies reporting immunogenicity outcomes for children or adults who received a single dose of yellow fever vaccination 10 or more years ago. Case series and single case reports were excluded. Participants who received more than one dose of yellow fever vaccination before measurement of the outcome were excluded. Identified records were reviewed by two independent reviewers. The primary outcome of the meta-analysis was the pooled seroprotection rate. Risk of bias was assessed with the Risk Of Bias In Non-randomized Studies of Interventions tool, and the Joanna Briggs Institute tool for analytical cross-sectional studies. Studies of moderate or good quality that reported seroprotection were included for random-effects meta-analysis and stratified by endemicity and specific risk groups. The study was registered with PROSPERO, CRD42023384087.

Of the 7363 articles identified by our search, 39 were eligible for inclusion for systematic review. These studies comprised 2895 individuals vaccinated 10-60 years ago. 20 studies were included in the meta-analysis. Pooled seroprotection rates were 94% (95% CI 86-99) among healthy adults in a non-endemic setting (mostly travellers) and 76% (65-85) in an endemic setting (all Brazilian studies). The pooled seroprotection rate was 47% (35-60) in children (aged 9-23 months at time of vaccination) and 61% (38-82) in people living with HIV. Reported criteria for seroprotection were highly heterogeneous.

The gathered evidence suggests that a single dose of yellow fever vaccination provides lifelong protection in travellers. However, in people living with HIV and children (younger than 2 years), booster doses might still be required because lower proportions of vaccinees were seroprotected 10 or more years post-vaccination. Lower observed seroprotection rates among residents of endemic areas were partly explained by the use of a higher cutoff for seroprotection that was applied in Brazil. Studies from sub-Saharan Africa were scarce and of low quality; thus no conclusions could be drawn for this region.

None.