Accurate mapping and disaggregation of key health and demographic risk factors have become increasingly important for disease surveillance, which can reveal geographical social inequalities for improved health interventions and for monitoring progress on relevant Sustainable Development Goals (SDGs). Household surveys like the Demographic and Health Surveys have been widely used as a proxy for mapping SDG-related household characteristics. However, there is no consensus on the workflow to be used, and different methods have been implemented with varying complexities. This study aims to compare multiple modelling frameworks to model indicators of human vulnerability to malaria (SDG Target 3.3) in Senegal. These indicators were categorised into socioeconomic (e.g., stunting prevalence, wealth index) and malaria prevention indicators (e.g., indoor residual spraying, insecticide-treated net ownership). We compared three categories of the commonly used methods: (1) spatial interpolation methods (i.e., inverse distance weighting, thin plate splines, kriging), (2) ensemble methods (i.e., random forest), and (3) Bayesian geostatistical models. Most indicators could be modelled with medium to high predictive accuracy, with R² values ranging from 0.40 to 0.86. No method or method category emerged as the best, but performance varied widely. Overall, socioeconomic indicators were generally better predicted by covariate-based models (e.g., random forest and Bayesian models), while methods using spatial autocorrelation alone (e.g., thin plate splines) performed better for variables with heterogeneous spatial structure, such as ethnicity and malaria prevention indicators. Increasing the complexity of the models did not always improve predictive performance, e.g., thin plate splines sometimes outperformed random forest or Bayesian geostatistical models. Beyond performance, we compared the different methods using other criteria (e.g., the ability to constrain the prediction range or to quantify prediction uncertainty) and discussed their implications for selecting a modelling approach tailored to the needs of the end user.

Knowing where people are is crucial for policymakers, particularly for the efficient allocation of resources in their country and the development of effective, people-centred policies. However, rural population distribution maps suffer from biases related to the type of dataset used to predict population density, such as the use of nighttime lights datasets in areas without electricity. This renders widely used datasets irrelevant in rural areas and biases nationwide models towards urban areas. To compensate for such biases, we aim at understanding the importance and relationship between water-related covariates and population densities in a random forest model across the urban-rural gradient. By extending a recursive feature elimination framework, we show that commonly used covariates are only selected when modelling the whole country. However, once the highest density areas are removed, water-related characteristics (especially distance to boreholes) become important covariates of population density outside of densely populated areas. This has important implications for modelling population in rural areas, including for a better estimation of the size of remote communities. When seeking to produce country-level population maps, we encourage further studies to explicitly account for rural areas by considering the urban-rural gradient and encourage the use of water-related datasets.

Background: Achieving malaria elimination requires identifying hotspots and key risk factors to guide targeted interventions. In this context, spatial models are crucial to support decision making. Yet, in elimination settings, national-level models suffer from some limitations. They often rely on prevalence survey data, which may be less reliable in low-transmission settings due to excess zeroes and under-representation of high-risk age groups. In addition, they often assume stationary risk factors, while pre-elimination strategies may lead to the emergence of (temporarily) distinct endemic profiles within countries. To address these biases, we model malaria incidence in Senegal, targeting malaria elimination for 2030, using routine malaria data and stratifying models by transmission level. Methods: Malaria incidence rates by health facility were calculated using reported cases and estimated catchment population, allowing for overlap between health facility catchment areas. Malaria incidence was then modelled using fine-resolution environmental and socio-economic covariates within a Bayesian hierarchical framework implemented using the INLA-SPDE approach. Models were stratified by season and endemicity level and compared with a national benchmark model to assess the heterogeneity of malaria risk factors. Finally, dasymetric disaggregation was applied to generate high-resolution (1 km) maps of malaria incidence. Results: The modelled incidence of malaria decreased by three cases per 1,000 people between 2017 (47 per 1,000 people) and 2019 (44 per 1,000 people), before increasing by 15 cases per 1,000 people in 2021 (59 per 1,000 people). At the sub-national level, malaria incidence followed an increasing gradient from the urbanised north-west to the south-eastern regions. Despite comparatively low seasonal variation in malaria risk factors, the importance of specific risk factors varied across areas with different levels of endemicity. Some factors (e.g. precipitation, ethnicity and urban agriculture) were particularly associated with either low, moderate, or high transmission settings. Fine-scale maps of the posterior estimates show higher malaria incidence for areas closer to a health facility than those further away, thus highlighting potential underdiagnosis of malaria in remote areas. Conclusion: Our findings underscore the importance of taking both endemicity and risk factors into account while evaluating malaria incidence. This approach allows the identification of risk factors specific to each endemicity level, essential for the design and implementation of more effective targeted interventions. In the near future, as prevalence data become less reliable and the quality of health system data continues to improve, fine-scale incidence models and maps will become critical tools to support decision-making.

Background: Female genital mutilation (FGM) is a human rights violation that still affects more than 3 million girls aged 0–14 years each year. To achieve the Sustainable Development Goal 2030 agenda, efforts have been made at the local, national and international levels to end the practice by the year 2030. However, the recent COVID-19 pandemic may have reversed the progress made due to increased rates of early marriage of girls, violence against children and school closures during lockdowns. Although some surveys have examined changes in FGM prevalence over the COVID-19 period, changes at the national and sub-national levels among 0–14 years old girls have not been quantified. Objectives: This study aimed to understand the potential impacts of the COVID-19 pandemic on the likelihood of FGM among girls aged 0–14 years, and whether it affected progress towards the elimination of FGM. Design: We used Bayesian hierarchical regression models implemented within the integrated nested Laplace approximations frameworks. Methods: We modelled the likelihood and prevalence of FGM among girls aged 0–14 years before and after the COVID-19 pandemic in Nigeria, with respect to individual- and community-level characteristics, using Bayesian hierarchical models. We used the 2018 Demographic and Health Survey as the pre-COVID-19 period and the 2021 Multiple Indicator Cluster Survey as the post-COVID-19 period. Results: At the state level, FGM prevalence varied geographically and increased by 23% and 27% in the northwestern states of Katsina and Kana, respectively. There were 11% increase in Kwara and 14% increase in Oyo. However, at the national level, the prevalence of FGM was found to decrease from 19.5% to 12.3% between 2018 and 2021. Cultural factors were identified as the key drivers of FGM among 0–14 years old girls in Nigeria. The changes in the likelihood of girls undergoing FGM across the two time periods also varied across ethnic and religious groups following COVID-19 pandemic. Conclusion: Our findings highlight that FGM is still a social norm in some states/regions and groups in Nigeria, thereby highlighting the need for a continued but accelerated FGM interventions throughout the country.

Background: Malaria distribution in sub-Saharan Africa is becoming increasingly spatially heterogeneous. Pre-elimination areas, high-transmission settings and urban areas with increasing transmission may all coexist within the same country. Effective intervention planning requires identifying high-risk areas and their drivers—an area in which geospatial models can provide valuable insights. Yet these models often assume stationary risk factors across transmission levels and frequently rely on prevalence survey data, which are unreliable in low-transmission settings. This study uses routine public-sector health facility data (ie, DHIS2) to model malaria incidence in Senegal from 2017 to 2021, accounting for transmission levels, and introduces novel fine-scale estimates of reported malaria incidence derived from dasymetric disaggregation. Methods: Health facility catchment populations were estimated using a probabilistic approach, allowing individuals to attend multiple facilities based on relative travel time. Malaria incidence per catchment was modelled using high-resolution environmental and socioeconomic covariates within a Bayesian hierarchical modelling framework. Separate models were fitted by season and endemicity-urbanisation level to assess the heterogeneity of risk factors. We adapted dasymetric disaggregation methods from population mapping to redistribute facility-level cases to 1 km pixels. Results: Estimated malaria incidence increased from 60 cases per 1000 people in 2017 to 71 cases in 2021. Key risk factors varied across endemicity-urbanisation levels; some factors were uniquely associated with low (eg, distance to roads), moderate (eg, cropland) and high-transmission settings (eg, flooded vegetation). Fine-scale maps represent malaria cases that the public-sector health system routinely reports, rather than the total malaria burden, and revealed potential underdiagnosis in remote areas. Conclusions: Our findings emphasise the importance of considering endemicity levels when evaluating malaria risk factors in heterogeneous transmission contexts. As health system data quality continues to improve, this approach offers a scalable alternative for mapping reported malaria incidence, supporting better targeted interventions and the identification of underdiagnosed areas.

Background: Female genital mutilation (FGM) is a human rights violation that still affects more than 3 million girls aged 0-14 years each year. To achieve the Sustainable Development Goal 2030 agenda, efforts have been made at the local, national and international levels to end the practice by the year 2030. However, the recent COVID-19 pandemic may have reversed the progress made due to increased rates of early marriage of girls, violence against children and school closures during lockdowns. Although some surveys have examined changes in FGM prevalence over the COVID-19 period, changes at the national and sub-national levels among 0-14 years old girls have not been quantified. ObjectivesThis study aimed to understand the potential impacts of the COVID-19 pandemic on the likelihood of FGM among girls aged 0-14 years, and whether it affected progress towards the elimination of FGM. Design We used Bayesian hierarchical regression models implemented within the integrated nested Laplace Approximations (INLA) frameworks.MethodsWe modelled the likelihood and prevalence of FGM among girls aged 0-14 years before and after the COVID-19 pandemic in Nigeria, with respect to individual and community-level characteristics, using Bayesian hierarchical models. We used the 2018 Demographic and Health Survey as the pre-COVID-19 period and the 2021 Multiple Indicator Cluster Survey as the post-COVID-19 period. ResultsAt the state level, FGM prevalence varied geographically and increased by 23% and 27% in the northwestern states of Katsina and Kana respectively. There were 11% increase in Kwara and 14% increase in Oyo. However, at the national level the prevalence of FGM was found to decrease from 19.5% to 12.3% between 2018 and 2021. Cultural factors were identified as the key drivers of FGM among 0-14 years old girls in Nigeria. The changes in the likelihood of girls undergoing FGM across the two time periods also varied across ethnic and religious groups following COVID-19 pandemic.ConclusionOur findings highlight that FGM is still a social norm in some states/regions and groups in Nigeria, thereby highlighting the need for a continued but accelerated FGM interventions throughout the country.