Impact of birth rate, seasonality and transmission rate on minimum levels of coverage needed for rubella vaccination

Abstract

Childhood rubella infection in early pregnancy can lead to fetal death or congenital rubella syndrome (CRS) with multiple disabilities. Reduction of transmission via universal vaccination can prevent CRS, but inadequate coverage may increase CRS numbers by increasing the average age at infection. Consequently, many countries do not vaccinate against rubella. The World Health Organization recommends that for safe rubella vaccination, at least 80% coverage of each birth cohort should be sustained. The nonlinear relationship between CRS burden and infection dynamics has been much studied; however, how the complex interaction between epidemic and demographic dynamics affects minimum safe levels of coverage has not been quantitatively evaluated across scales necessary for a global assessment. We modelled 30-year CRS burdens across epidemiological and demographic settings, including the effect of local interruption of transmission via stochastic fadeout. Necessary minimum vaccination coverage increases markedly with birth and transmission rates, independent of amplitude of seasonal fluctuations in transmission. Susceptible build-up in older age groups following local stochastic extinction of rubella increased CRS burden, indicating that spatial context is important. In low birth-rate settings, 80% routine coverage is a conservative guideline, particularly if supplemented with campaigns and vaccination of women of childbearing age. Where birth and transmission rates are high, immunization coverage must be well above 80% and campaigns may be needed. Policy-makers should be aware of the potential negative effect of local extinction of rubella, since heterogeneity in vaccination coverage will shape extinction patterns, potentially increasing CRS burdens.

Fig. 1.

Dynamics of the relative congenital rubella syndrome (CRS) burden for R₀ = 12, and 40 births/1000 per year for three different levels of routine vaccination coverage (see key) for four different immunization strategies (arrows indicate timing of SIAs, or starting campaigns). The x-axis shows time in years, and the y-axis indicates ρ_y, the ratio of yearly summed CRS cases in the presence and absence of vaccination. (b–d) Lines corresponding to coverage higher than 50% lie close to the bottom of the plot where ρ_y = 0.

Fig. 2.

Effect of routine vaccination coverage levels on congenital rubella syndrome (CRS) burden over 30 years, for different levels of routine infant and young children vaccination coverage only, total number of CRS cases/1000 live births for a population of 750 000 across different birth rates for (a) 12 births/1000, (b) 30 births/1000 and (c) 40 births/1000). Three levels of R₀ are shown (in colour) under weak seasonality in transmission (α = 0·2; higher seasonality does not alter results). Horizontal lines indicate the number of CRS cases/1000 live births occurring in the absence of vaccination; points on each curve above the corresponding line correspond to negative outcomes of vaccination; arrows indicate the level of coverage required to avoid this. Note that the exact numbers on the y-axis in any particular context will depend on the precise pattern of the fertility curve, which may itself vary with birth rate. Here, it is assumed that the pattern over age follows that for Niger.

Fig. 3.

Minimum level of vaccination required to retain ρ, the ratio of cumulative congenital rubella syndrome incidence before and after vaccination <1 over 30 years (y-axis), across births/1000 per year (x-axis) and levels of R₀ (panels) for five different immunization strategies, (i) routine vaccination only, (ii) routine + supplementary immunization activities (SIAs) of 1- to 4-year-olds every 4 years, (iii) routine + SIAs of 1- to 4-year-olds every 4 years + a starting campaign in 1- to 4-year-olds, (iv) routine + SIAs of 1- to 4-year-olds every 4 years + a starting campaign in 1- to 14-year-olds, and (v) routine + SIAs of 1- to 4-year-olds every 4 years + a starting campaign in 1- to 14-year-olds + vaccination of women of childbearing age (WCB). The level currently recommended by WHO (80%) is shown (grey line).

Fig. 4.

The effect of reduced infected immigration on the burden of CRS in a stochastic situation. Stochastic simulations showing bi-weekly rubella incidence in the presence of some immigration of infected individuals [(a) left panel, on average three infected immigrants per year] or reduced immigration of infected individuals [(a) right panel, on average 0·3 infected immigrants per year]. Panel (b) shows the numbers of CRS cases/1000 live births [medians (inner horizontal bar), 25% and 75% quantiles (box) and 0·975 and 0·025 quantiles (outer lines) across 10 simulations], and corresponding levels in the deterministic situation (red asterisks). All simulations had a birth rate of 12/1000 and R₀ = 6 (set since the effect is expected to be strongest in communities with low transmission and/or birth rate) and population size of 50 000; burdens were assessed across 40 years.