RESUMO

This paper aims to extend the Besag model, a widely used Bayesian spatial model in disease mapping, to a non-stationary spatial model for irregular lattice-type data. The goal is to improve the model's ability to capture complex spatial dependence patterns and increase interpretability. The proposed model uses multiple precision parameters, accounting for different intensities of spatial dependence in different sub-regions. We derive a joint penalized complexity prior to the flexible local precision parameters to prevent overfitting and ensure contraction to the stationary model at a user-defined rate. The proposed methodology can be used as a basis for the development of various other non-stationary effects over other domains such as time. An accompanying R package fbesag equips the reader with the necessary tools for immediate use and application. We illustrate the novelty of the proposal by modeling the risk of dengue in Brazil, where the stationary spatial assumption fails and interesting risk profiles are estimated when accounting for spatial non-stationary. Additionally, we model different causes of death in Brazil, where we use the new model to investigate the spatial stationarity of these causes.

Assuntos

RESUMO

OBJECTIVE: To estimate trends in the prevalence of child stunting in the population of children under 5 years of age covered by public health programmes, between 2009 and 2014 in Misiones, Argentina. METHODS: Using Bayesian model-based geostatistics, we evaluated 724 872 anthropometric measurements corresponding to 110 633 children. In order to identify disparities at local scale, we evaluated the hypotheses of a differential reduction of stunting according to the geographical location (at two-level spatial resolution) and to the socioeconomic level in a rural or urban environment. RESULTS: The prevalence of stunting had fallen significantly in the province overall. Sex and age defined gender disparities at individual level, and there were regional disparities with higher prevalence values in the north and northeast regions. In these areas, stunting decreased to a greater degree during the studied period, although the spatial pattern remained smoother. Stunting increased in peripheral urban and dispersed rural areas that are socioeconomically vulnerable. CONCLUSIONS: The spatial multi-level geostatistical estimates of child undernutrition provide a precision public health tool to target public policies to those populations with the greatest need, in order to reduce health disparities.

OBJECTIF: Estimer les tendances dans la prévalence du retard de croissance dans la population des enfants de moins de 5 ans couverts par les programmes de santé publique, entre 2009 et 2014 à Misiones, en Argentine. MÉTHODES: En utilisant la géostatistique basée sur un modèle bayésien, nous avons évalué 724.872 mesures anthropométriques correspondant à 110.633 enfants. Afin d'identifier les disparités à l'échelle locale, nous avons évalué les hypothèses d'une réduction différentielle du retard de croissance en fonction de la situation géographique (à une résolution spatiale à deux niveaux) et du niveau socioéconomique en milieu rural ou urbain. RÉSULTATS: La prévalence du retard de croissance avait considérablement diminué dans l'ensemble de la province. Le sexe et l'âge définissaient des disparités entre les sexes au niveau individuel, et il y avait des disparités régionales avec des valeurs de prévalence plus élevées dans les régions du nord et du nord-est. Dans ces régions, le retard de croissance a diminué plus fortement au cours de la période étudiée, bien que le modèle spatial soit resté plus lisse. Le retard de croissance a augmenté dans les zones urbaines périphériques et les zones rurales dispersées qui sont socioéconomiquement vulnérables. CONCLUSIONS: Les estimations géostatistiques spatiales à plusieurs niveaux de la sous-nutrition infantile fournissent un outil de santé publique de précision pour cibler les politiques publiques sur les populations qui en ont le plus besoin, afin de réduire les disparités en matière de santé.

Assuntos

RESUMO

BACKGROUND: Hantavirus disease in humans is rare but frequently lethal in the Neotropics. Several abundant and widely distributed Sigmodontinae rodents are the primary hosts of Orthohantavirus and, in combination with other factors, these rodents can shape hantavirus disease. Here, we assessed the influence of host diversity, climate, social vulnerability and land use change on the risk of hantavirus disease in Brazil over 24 years. METHODS: Landscape variables (native forest, forestry, sugarcane, maize and pasture), climate (temperature and precipitation), and host biodiversity (derived through niche models) were used in spatiotemporal models, using the 5570 Brazilian municipalities as units of analysis. RESULTS: Amounts of native forest and sugarcane, combined with temperature, were the most important factors influencing the increase of disease risk. Population at risk (rural workers) and rodent host diversity also had a positive effect on disease risk. CONCLUSIONS: Land use change-especially the conversion of native areas to sugarcane fields-can have a significant impact on hantavirus disease risk, likely by promoting the interaction between the people and the infected rodents. Our results demonstrate the importance of understanding the interactions between landscape change, rodent diversity, and hantavirus disease incidence, and suggest that land use policy should consider disease risk. Meanwhile, our risk map can be used to help allocate preventive measures to avoid disease.

Assuntos