Estimation of the poliovirus type 2 immunity gap in South Africa

Abstract

In the context of polio eradication efforts, accurate assessment of vaccination programme effectiveness is essential to public health planning and decision making. Such assessments are often based on zero-dose children, estimated using the number of children who did not receive the first dose of the Diphtheria-Tetanus-Pertussis containing vaccine as a proxy. Our study introduces a novel approach to directly estimate the number of children susceptible to poliovirus type 2 (PV2) and uses this approach to provide district-level estimates for South Africa of susceptible children born between 2017 and 2022. We used district-level data on annual doses of inactivated poliovirus vaccine (IPV) administered, live births, and population sizes, from 2017 through 2022. We imputed missing vaccination data, implemented flexible assumptions regarding dose distribution in the eligible population, and used estimated efficacy values for one, two, three, and four doses of IPV, to compute the number of susceptible and immune children by birth year. We validated our approach by comparing an intermediary output with zero-dose children (ZDC) estimated using data reported by WHO/UNICEF Estimates of National Immunization Coverage (WUENIC). Our results indicate high heterogeneity in susceptibility to PV2 across South Africa's 52 districts as of the end of 2022. In children under 5 years, PV2 susceptibility ranged from approximately 30 % in districts including Xhariep (31.9 %), Ekurhuleni (30.1 %), and Central Karoo (29.8 %), to less than 4 % in Sarah Baartman (1.9 %), Buffalo City (2.1 %), and eThekwini (3.2 %). Our susceptibility estimates were consistently higher than ZDC over the timeframe. We estimated that ZDC decreased nationally from 155,168 (152,737-158,523) in 2017 to 108,593 in 2021, and increased to 127,102 in 2022, a trend consistent with ZDC derived from data reported by WUENIC. While our approach provides a more comprehensive profile of PV2 susceptibility, our susceptibility and ZDC estimates generally agree in the ranking of districts according to risk.

Keywords: IPV; Immunity; Poliovirus; South Africa; Vaccination; Zero-dose children.

Fig. 1

National-level comparison of SACEMA and WUENIC-derived zero-dose children (ZDC) estimates, alongside SACEMA estimates of numbers of children fully susceptible to PV2, by birth year. Black vertical bars represent the 2.5th and 97.5th quantiles of 1,000 bootstrap samples. As IPV data was complete for 2020–2022, bootstrapping was not necessary, hence the lack of variation in these years. There is a clear decreasing trend in estimates between 2017 and 2021, except for susceptibility estimates in 2020, and an increase from 2021 to 2022 in all three estimates.

Fig. 2

District-level proportions of fully susceptible children in the 0–4 age group in South Africa as of the end of end 2022. We estimate varying levels of susceptibility across districts, with higher levels of susceptibility in the northern and western regions of the country. In contrast, districts along the coast, particularly in the Kwa-Zulu Natal and Eastern Cape provinces, have lower levels of susceptibility.