Higher Population Coverage with Typhoid Conjugate Vaccine is Needed to Induce Herd Protection: Evidence from a Cluster-Randomized Trial in Urban Bangladesh

Technical Summary: Higher Population Coverage with Typhoid Conjugate Vaccine is Needed to Induce Herd Protection

Problem Statement
Typhoid fever, caused by Salmonella enterica serotype Typhi, remains a significant public health burden in low- and middle-income countries. While the World Health Organization (WHO) recommends typhoid conjugate vaccines (TCVs), specifically Vi-tetanus toxoid (Vi-TT), for children aged 6 months to <16 years, the extent of indirect (herd) protection remains a critical question. A previous cluster-randomized trial (CRT) in Dhaka, Bangladesh, demonstrated that Vi-TT conferred 85% direct protection to vaccinees but failed to show significant herd protection for unvaccinated individuals. This absence of herd effects was surprising, particularly when contrasted with earlier trials of the less protective Vi-polysaccharide (Vi-PS) vaccine in Kolkata, which showed significant herd protection when administered to both children and adults. The authors hypothesized that the lack of herd protection in the Bangladesh trial might be due to insufficient vaccine coverage of the entire population, as the trial restricted vaccination to children, leaving adults unvaccinated and potentially acting as reservoirs for transmission.

Methodology
This study is a post-hoc reanalysis of data from a participant and observer-blind CRT conducted in Mirpur, Dhaka, involving 150 clusters.

• Study Design: The original trial randomized clusters to receive either a single dose of Vi-TT or a Japanese encephalitis (JE) vaccine (control). The target population for vaccination was children aged 9 months to <16 years.
• Surveillance: Passive surveillance for typhoid fever was conducted over a 23-month period across the entire population (all ages) in designated treatment facilities. Cases were confirmed via blood culture.
• Analytical Approach: The authors stratified the 150 clusters into four quartiles based on ascending levels of vaccine coverage. Two distinct coverage metrics were calculated for each cluster:
  1. Entire Population Coverage: The proportion of all residents (all ages) who were vaccinated.
  2. <16-Year Coverage: The proportion of children aged 9 months to <16 years who were vaccinated.
• Statistical Models: Using Cox proportional hazard models with robust variance (accounting for cluster randomization), the authors estimated Hazard Ratios (HR) to calculate indirect, total, and overall protection. Analyses were stratified by the two coverage metrics and by age groups (<16 years and ≥16 years). Covariates included geographic ward, distance to clinics, and baseline demographic/hygiene factors.

Key Contributions

• Re-evaluation of Herd Protection: The study provides evidence that herd protection from Vi-TT is not absent but is contingent upon the level of vaccine coverage across the entire population, rather than just the targeted pediatric population.
• Age-Stratified Transmission Dynamics: The analysis reveals that indirect protection is primarily observed in adults (≥16 years) in clusters with higher overall population coverage, whereas no significant indirect protection was found in children (<16 years), even in clusters with the highest pediatric coverage.
• Policy Implications for TCVs: The findings challenge the current WHO recommendation to vaccinate only children, suggesting that achieving herd immunity may require broader age-targeting strategies that include adults.

Results

• Coverage Levels: The average vaccine coverage across all clusters was 69% among the targeted children (<16 years) but only 20% in the entire population.
• Indirect Protection by Coverage Metric:
  • Entire Population Coverage: Significant indirect protection (47%; 95% CI: 3%, 71%) was observed only in the highest quartile of entire population coverage (21.5%–26.0%). No significant indirect protection was found in lower quartiles.
  • <16-Year Coverage: No significant indirect protection was observed in any quartile of child-specific coverage, including the highest quartile (72.1%–78.9%), where indirect protection was estimated at -8% (95% CI: -108%, 44%).
• Age-Specific Indirect Protection:
  • Adults (≥16 years): Indirect protection increased with higher entire population coverage, reaching 58% (95% CI: 18%, 78%) in the highest quartile.
  • Children (<16 years): No evidence of indirect protection was found in children across any coverage quartile.
• Overall Protection: Overall protection for the entire population increased with higher entire population coverage, reaching 66% (95% CI: 45%, 78%) in the highest quartile.

Significance and Claims
The paper concludes that the failure to observe herd protection in the original analysis was likely due to the low coverage of the total population, as the vaccination strategy excluded adults. The authors claim that:

1. Threshold for Herd Effects: Herd protection from TCVs is evident only when vaccine coverage of the entire population reaches a specific threshold (observed here at >21.5%), even if coverage among children is high.
2. Role of Adults: Adults may be a primary source of typhoid transmission in this setting, potentially due to food preparation roles or chronic fecal excretion, necessitating their inclusion in vaccination campaigns to interrupt transmission.
3. Policy Revision: To achieve substantial reductions in the total typhoid burden and maximize public health impact, vaccination strategies may need to evolve from targeting only children to including adults. However, the authors note that this is a hypothesis requiring further validation in future studies before WHO recommendations are revised.
4. Limitations: The authors acknowledge the study is a post-hoc analysis conducted in a high-transmission urban slum, and the findings may not generalize to lower-incidence settings. The effects of vaccinating both children and adults simultaneously remain to be demonstrated in a prospective trial.