Evaluation of the Entomological Adaptive Surveillance Framework for malaria vector monitoring: a comparative field trial with routine surveillance in Ghana and Mozambique.

Introduction: Accurate mosquito surveillance is essential for guiding targeted interventions. To capture the spatial and temporal heterogeneity of mosquito populations, surveillance designs would ideally be flexible and evidence-based, that is, informed by prior data. This study compares a newly developed entomological adaptive surveillance framework (EASF) and routine entomological surveillance in Ghana and Mozambique. Routine surveillance refers to standard practices by National Malaria Control Programmes, while EASF enables the strategic selection of sampling locations based on modelled risk or environmental criteria. The study evaluates both approaches based on Methods: A Bayesian framework for exceedance probability is employed in this adaptive design to select the number and location of the adaptive sites. Estimates are based on a Bayesian spatiotemporal model with nugget effects to analyse mosquito abundance, using log-transformed counts to address heavy-tailed distributions. Results: The EASF outperformed routine surveillance in most metrics, including mosquito catch rates, model robustness and reduction in uncertainty. Notably, when standardised, EASF sites yielded higher mosquito catches and more stable predictions, as indicated by lower coefficients of variation. While EASF generally improved model inference and predictive accuracy, performance varied by country. Nevertheless, EASF consistently identified proportionally fewer obsolete locations than routine surveillance, demonstrating its efficiency in targeting informative sites. Conclusions: EASF offers an effective, evidence-based approach to improving surveillance precision, enabling surveillance programmes to dynamic transmission systems and emerging needs while maintaining operational feasibility. Integrating adaptive and routine designs can enhance surveillance efficiency, either by improving accuracy, reducing site numbers or accelerating detection of ecological changes. The key to effective entomological surveillance is not rigidly achieving a target, but continuously adapting towards it.