Rubella vaccination in India: identifying broad consequences of vaccine introduction and key knowledge gaps

Abstract

Rubella virus infection typically presents as a mild illness in children; however, infection during pregnancy may cause the birth of an infant with congenital rubella syndrome (CRS). As of February 2017, India began introducing rubella-containing vaccine (RCV) into the public-sector childhood vaccination programme. Low-level RCV coverage among children over several years can result in an increase in CRS incidence by increasing the average age of infection without sufficiently reducing rubella incidence. We evaluated the impact of RCV introduction on CRS incidence across India's heterogeneous demographic and epidemiological contexts. We used a deterministic age-structured model that reflects Indian states' rural and urban area-specific demography and vaccination coverage levels to simulate rubella dynamics and estimate CRS incidence with and without RCV introduction to the public sector. Our analysis suggests that current low-level private-sector vaccination has already slightly increased the burden of CRS in India. We additionally found that the effect of public-sector RCV introduction depends on the basic reproductive number, R 0, of rubella. If R 0 is five, a value empirically estimated from an array of settings, CRS incidence post-RCV introduction will likely decrease. However, if R 0 is seven or nine, some states may experience short-term or annual increases in CRS, even if a long-term total reduction in cases (30 years) is expected. Investment in population-based serological surveys and India's fever/rash surveillance system will be key to monitoring the success of the vaccination programme.

Keywords: CRS; Congenital rubella syndrome; India; rubella; vaccine.

Fig. 1.

State-level covariates: (a) rural private-sector routine RCV coverage (as a proportion). (b) Urban private-sector routine RCV coverage (as a proportion). (c) Rural public-sector routine MR coverage (as a proportion). (d) Urban public-sector routine MR coverage (as a proportion). All coverage estimates were extracted from the Rapid Survey on Children 2013–14 [5]. To estimate private-sector routine RCV coverage, we assumed that any child who received their vaccinations in a private healthcare centre received RCV. To estimate public-sector routine MR coverage, we assumed that current MCV1 coverage estimates reflect future routine MR coverage estimates. See Supplemental Table S1 for a full list of coverage estimates for each simulated area.

Fig. 2.

Results of simulated rubella dynamics assuming an R₀ of 5 and private-sector vaccination since 1993: (a) Rural estimated 2016 number of CRS cases by state determined by ‘private-sector vaccine’ scenario. (b) Urban estimated 2016 number of CRS cases by state determined by ‘private-sector vaccine’ scenario. (c) Rural estimated 2016 CRS incidence per 100 000 live births by state determined by ‘private-sector vaccine’ scenario. (d) Urban estimated 2016 CRS incidence per 100 000 live births by state determined by ‘private-sector vaccine’ scenario. Broadly, rural areas experience higher burdens of CRS cases because they have larger populations, and urban areas have higher CRS incidence per 100 000 because they have higher private-sector coverage and lower birth rates. See Supplemental Table S2 for a full list of estimated CRS cases and incidence for each simulated area.

Fig. 3.

Results of simulated rubella dynamics assuming an R₀ of 5. The number of CRS cases by year if private-sector vaccination is or is not taken into account, India 1991–2016.

Fig. 4.

Results of simulated rubella dynamics by assumed R₀ values across columns for (a) urban Kerala (high coverage, low birth rates) in row 1, (b) urban Gujarat (somewhat average coverage and birth rate) in row 2 and (c) rural Uttar Pradesh (low coverage, high birth rate) in row 3. The figures show CRS incidence per 100 000 live births over time for four vaccination scenarios by assumed R₀ values across columns. The solid black lines represents the CRS incidence in the ‘no vaccine’ scenario. The dashed red line represents the CRS incidence in the ‘private-sector vaccine’ scenario; the estimated private-sector RCV coverage (as a proportion) is displayed in the legend for each area per [5]. The dotted blue line represents the CRS incidence in the ‘60% catch-up + routine vaccine’ scenario; the estimated public-sector routine MR coverage (as a proportion) is displayed in the legend for each area per [5]. The dashed and dotted green line represents the hypothetical ‘80% catch-up + 80% routine vaccine’ scenario, which is the critical RCV coverage threshold estimated per [12].

Fig. 5.

Results of simulated rubella dynamics taking into account private-sector vaccination since 1993: 30-year CRS incidence ratio (IR) of ‘60% catch-up + routine vaccine’ scenario compared with ‘private-sector vaccine’ scenario across all states by rural and urban areas and R₀. Shades of blue represent a CRS incidence ratio less than one, the colour white represents a CRS incidence ratio of one, and shades of red represent a CRS incidence ratio of greater than one. At R₀ = 11, we estimated six areas may experience a long-term increase in CRS post-RCV introduction (i.e. rural areas in Rajasthan, Uttar Pradesh, Madhya Pradesh, Bihar, Meghalaya and Nagaland); and at R₀ = 9, rural Rajasthan was estimated to have a long-term increase in CRS. As R₀ increases, so does the number of simulated areas estimated to have a long-term CRS incidence ratio greater than one in the ‘60% catch-up + routine vaccine’ scenario compared with the ‘private-sector vaccine’ scenario.