Estimating the global impact of rotavirus vaccines on child mortality

Abstract

Objectives: We estimated the global impact of rotavirus vaccines on deaths among children under five years old by year.

Methods: We used a proportionate outcomes model with a finely disaggregated age structure to estimate rotavirus deaths prevented by vaccination over the period 2006-2019 in 186 countries. We ran deterministic and probabilistic uncertainty analyses and compared our estimates to surveillance-based estimates in 20 countries.

Results: We estimate that rotavirus vaccines prevented 139,000 under-five rotavirus deaths (95% uncertainty interval 98,000-201,000) in the period 2006-2019. In 2019 alone, rotavirus vaccines prevented 15% (95% uncertainty interval 11-21%) of under-five rotavirus deaths (0.5% of child mortality). Assuming global use of rotavirus vaccines and coverage equivalent to other co-administered vaccines could prevent 37% of under-five rotavirus deaths (1.2% of child mortality). Our estimates were sensitive to the choice of rotavirus mortality burden data and several vaccine impact modeling assumptions. The World Health Organization's recommendation to remove age restrictions in 2012 could have prevented up to 17,000 rotavirus deaths in the period 2013-2019. Our modeled estimates of rotavirus vaccine impact were broadly consistent with estimates from post-vaccination surveillance sites.

Conclusion: Rotavirus vaccines have made a valuable contribution to global public health. Enhanced rotavirus mortality prevention strategies are needed in countries with high mortality in under-5-year-old children.

Keywords: Childhood diarrhea; Diarrhea mortality; Rotavirus; Rotavirus vaccination.

Figure 1

Number of countries (panel a) and infants (panel b) using rotavirus vaccines by year (2006-2019) and 2022-23 World Bank income group. HIC, high-income country; LIC, low-income country; MIC, middle-income country.

Figure 2

Number of rotavirus deaths prevented (and not prevented) by rotavirus vaccination in 186 countries by year (2006-2019) and 2022-23 World Bank income group. RVGE, rotavirus gastroenteritis; HIC, high-income country; LIC, low-income country; MIC, middle-income country. Caption: The overall decrease in RVGE deaths each year is due to the decreasing trend in RVGE mortality estimated by all 3 RVGE mortality datasets (Global Burden of Disease Study, Maternal and Child Epidemiology Estimation Group and World Health Organization and Centers for Disease Control and Prevention). Nine countries (Albania, Benin, Chile, Kyrgyzstan, North Macedonia, Russia, Sweden, Timor-Leste, Thailand) reported partial/restricted introduction before 2022 and two countries (China and Vietnam) both had domestic vaccines available on the private market. However, World Health Organization-UNICEF Estimates of National Immunization Coverage rotavirus vaccination coverage was 0% for the period 2006-2019 in all 11 countries, so they were excluded from the figure and analysis. The Philippines and Venezuela suspended their programs in 2016 and 2018 respectively but all other countries used rotavirus vaccine until the end of 2019. World Bank income groups for 2022-23 were applied to all historical years 2006-2019 and thus do not account for movement between income groups over time.

Figure 3

Number of RVGE deaths averted by rotavirus vaccination in 186 countries by year (2006-2019): base case (95% uncertainty interval) and alternative scenarios. GBD, Global Burden of Disease Study; MCEE, Maternal and Child Epidemiology Estimation Group; RVGE, rotavirus gastroenteritis; WHO/CDC, World Health Organization and Centers for Disease Control and Prevention. Caption: 95% uncertainty intervals are based on 100 probabilistic runs per country and include uncertainty in the population aged <5 years, RVGE mortality rates aged <5 years, RVGE age distributions, Rotavirus vaccine coverage, timeliness, efficacy, and waning. Deterministic scenarios are also shown for different mortality datasets, an age-restricted scenario and a scenario with higher impact assuming products with 2-dose schedules are assigned the same impact as products with 3-dose schedules.

Figure 4

Number of rotavirus deaths in 186 countries by year (2006-2019): base case and alternative scenarios. GBD, Global Burden of Disease Study; MCEE, Maternal and Child Epidemiology Estimation Group; WHO/CDC, World Health Organization and Centers for Disease Control and Prevention; WUENIC, World Health Organization-UNICEF Estimates of National Immunization Coverage. Caption: all scenarios are adjusted for rotavirus vaccine use. The dashed black line shows the historical impact that could have been achieved had rotavirus vaccines been introduced globally in 2006 at existing Diphtheria, Tetanus, and Pertussis coverage levels. Our modeled estimates vary slightly from the official estimates produced by each mortality estimation group because we used a standard demographic dataset (United Nations World Population Prospects 2022 revision) and a standard set of 186 countries. In addition, our model tracks the experience of each birth cohort from birth to age 5.0 years then allocates estimated age-specific deaths into the appropriate calendar years post hoc.