Geospatial approaches for mapping zero-dose prevalence and routine childhood immunisation coverage in low- and...

Zero-dose (ZD) children remain a critical public health concern, particularly in low- and lower-middle-income countries (LLMICs), where over 80% of the global ZD population resides, disproportionately concentrated among the most marginalised. Geospatial approaches have emerged as effective tools for identifying and targeting immunisation gaps. However, no review has systematically documented the spatial approaches used to predict or map ZD prevalence and routine immunisation (RI) coverage. This scoping review addresses this gap across LLMICs.

We searched six databases for peer-reviewed articles published up to 2025, on spatial modelling of childhood vaccination coverage in LLMICs. We extracted details on study characteristics, covariate types and sources, modelling methods, and limitations; stratifying findings between studies mapping RI coverage and those estimating ZD prevalence. Articles were thematically summarised focusing on geospatial data, modelling approaches, and corresponding gaps.

We included 102 articles from 68 LLMICs, with 70% published between 2021 and 2024, and 87% concentrated in Ethiopia, Nigeria, and India. Most studies mapped RI coverage indicators, only 19.6% assessed ZD prevalence, based on two distinct definitions. Studies relied predominantly on survey-based vaccination data. Covariate data were dominated by demographic factors (49%) with limited representation of hard-to-reach contexts such as conflict areas and nomadic populations. Methods included clustering and autocorrelation analysis (54%), spatial interpolation (45%), small-area estimation (13%), and machine learning (8%). Key gaps included inconsistent ZD definitions, data inaccuracies and scarcity, and limited use of routine vaccination data.

Geospatial mapping of RI coverage is expanding across LLMICs but relies predominantly on exploratory approaches insufficient for precise targeting, while dedicated ZD mapping remains sparse. Geospatial modelling is constrained by overreliance on survey data, limited application of routine data, and underrepresentation of marginalised populations. Addressing these gaps requires integrated data systems, and reproducible modelling approaches underpinned by sustained investment in local analytical capacity.