The Malaria Vaccine Implementation Programme Reduced Clinical Malaria in Kenya, 2020 to 2022 Impact evaluation with predictive time-series models of routine surveillance data

Background : Quantifying the impact of malaria vaccines on outpatient malaria burden in children is of interest to many countries introducing either of the World Health Organization (WHO)-recommended vaccines (RTS,S/AS01 [RTS,S] or R21/Matrix-M [R21[). The cluster randomized implementation of RTS,S by the Kenya Ministry of Health during the Malaria Vaccine Implementation Programme (MVIP) provided a unique opportunity to measure the impact of RTS,S on uncomplicated malaria cases reported through routine malaria case surveillance data in western Kenya. Methods : From 23 implementing and 23 comparison subcounties in western Kenya, monthly numbers of confirmed uncomplicated malaria cases among individuals under 5 years of age (<5y) and 5 years of age and older (≥5y) were extracted from the Kenya Health Management Information System, stored in a DHIS2 instance. Facilities that reported data ≥11 months per year from January 2015 through December 2022 (139 implementing, 141 comparison) were included in the analysis. Prevaccination data (January 2015–December 2019) were used to predict counterfactual case counts during the RTS,S vaccination period (January 2020 through December 2022). Models of <5y malaria cases included a covariate for ≥5y malaria cases to control for non-RTS,S malaria control measures. The impact was estimated as the difference between predicted and observed cases across implementing and comparison areas. Results : The proportion of vaccine eligible children (RTS,S doses at 6, 7, 9, and 24 months) among all children <5y was small at the start of MVIP but was estimated to reach 60% by December 2022. During the 3-years after introduction of RTS,S, <5y malaria cases malaria cases were 2.3% lower than predicted in non-implementing facilities and 13.8% lower in implementing facilities, a reduction of 11.6%. In the first year of following introduction of RTS,S, the reduction was only 5.6%, increasing to 8.2% after 2 years. Conclusion : This analysis revealed that routine health data, although age-aggregated to both vaccine eligible and noneligible children <5 years, demonstrated a reduction in malaria cases from one year after vaccine introduction. A further reduction in cases was observed as more vaccine-eligible children comprised the under 5 population. High quality routine data can be used to monitor vaccine impact on clinical malaria.