Infant vaccination coverage trajectories and immunisation system archetypes in the WHO African region, 2019-2024: a longitudinal analysis of 47 countries.

Abstract

Introduction: COVID-19 disrupted childhood vaccination, but country-specific heterogeneity in diphtheria-tetanus-pertussis-containing vaccines (DTP3) coverage across the WHO African region remains poorly understood. This study addressed this gap by quantifying temporal trajectories and identifying latent structural patterns in DTP3 coverage across all 47 countries in the WHO African region from 2019 to 2024, spanning the onset and recovery phases of the COVID-19 pandemic.

Methods: Country-level DTP3 coverage data from 2019 to 2024 were extracted from the WHO/United Nations Children's Fund Estimates of National Immunisation Coverage, published in July 2025. Country-specific ordinary least squares regression models estimated annual rates of change (slopes) in DTP3 coverage, while linear mixed-effects modelling decomposed within-country and between-country variance components. Principal component (PC) analysis (PCA) was then applied to the 6-year coverage matrix to derive two dominant, uncorrelated components capturing coverage level (PC1) and temporal trajectory (PC2). K-means clustering of PC scores grouped countries into empirically distinct immunisation system archetypes.

Results: Between 2019 and 2024, the estimated annual rate of change in DTP3 coverage was positive in 21 countries and negative in 24; Eritrea and Burkina Faso showed no evidence of a linear trend. The fixed-effect slope was 0.07 percentage points per year (95% CI -0.34 to 0.47; p=0.75), indicating no statistically significant net regional change, while between-country heterogeneity was substantial (random-intercept variance, 188.18; random-slope variance, 1.18). PCA showed that PC1 dominated, explaining 93% of variance (capturing sustained coverage level), with PC2 contributing an additional 3% (capturing temporal trajectory). K-means clustering of PC scores yielded four archetypes: cluster 1 (high coverage, consolidating; n=15), cluster 2 (chronically low coverage, divergent trajectories; n=9), cluster 3 (mid-low coverage, predominantly improving; n=11) and cluster 4 (mid-high coverage, erosion risk; n=12).

Conclusion: This study generated an empirically derived, model-driven typology of immunisation system performance across the WHO African region, characterised by four distinct DTP3 coverage archetypes. The typology provides decision-makers with a granular basis for differentiating programmatic strategies, prioritising resource deployment and designing cluster-specific implementation research rather than applying uniform regional solutions.