Spatial and Spatio-Temporal Epidemiology最新文献

{"title":"BARTSIMP: Flexible spatial covariate modeling and prediction using Bayesian Additive Regression Trees","authors":"Alex Ziyu Jiang ,&nbsp;Jon Wakefield","doi":"10.1016/j.sste.2025.100757","DOIUrl":"10.1016/j.sste.2025.100757","url":null,"abstract":"<div><div>Prediction is a classic challenge in spatial statistics and the inclusion of spatial covariates can greatly improve predictive performance when incorporated into a model with latent spatial effects. It is desirable to develop flexible regression models that allow for nonlinearities and interactions in the covariate specification. Existing machine learning approaches that allow for spatial dependence in the residuals fail to provide reliable uncertainty estimates. In this paper, we investigate the combination of a Gaussian process spatial model with a Bayesian Additive Regression Tree (BART) model. The computational burden of the approach is reduced by combining Markov chain Monte Carlo (MCMC) with the Integrated Nested Laplace Approximation (INLA) technique. We study the performance of the method first via simulation. We then use the model to predict anthropometric responses in Kenya, with the data collected via a complex sampling design. In particular, household survey data are collected via stratified two-stage unequal probability cluster sampling, which requires special care when modeled.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"55 ","pages":"Article 100757"},"PeriodicalIF":1.7,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145325733","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatio-temporal clustering analysis, temporal trends, and inequality in oral and oropharyngeal cancer mortality in Brazil over 44 years (1980–2023)","authors":"José Mário Nunes da Silva ,&nbsp;Fabrício Dos Santos Menezes ,&nbsp;Diego Rodrigues Mendonça e Silva ,&nbsp;Tarsila Guimarães Vieira da Silva ,&nbsp;Luiz Paulo Kowalski","doi":"10.1016/j.sste.2025.100758","DOIUrl":"10.1016/j.sste.2025.100758","url":null,"abstract":"<div><div>This study analyzed the temporal trends, spatial and spatio-temporal patterns of OC and OPC mortality in Brazil between 1980 and 2023, and explored their association with socioeconomic inequality. We conducted an ecological study using age- and sex-standardized mortality rates, smoothed via a local empirical Bayesian method. We assessed temporal trends through joinpoint regression. We evaluated global and local spatial autocorrelation and detected spatio-temporal clusters using a retrospective space–time scan statistic based on a Poisson model. We observed a decrease in OC mortality, particularly among men aged 40–59 years in the Southeast and South regions. In contrast, OPC mortality increased throughout the study period in both sexes, especially among individuals aged 60–79 years, with the largest increases occurring in the North, Northeast, and Central-West regions. Moran’s I revealed significant spatial dependence for both cancers. Spatial analyses identified persistent high-risk clusters in the Southeast and South, which expanded toward the Northeast and Central-West. Spatio-temporal analysis showed a recent shift of major OC clusters from the Southeast and South towards the Northeast, whereas OPC clusters continued to expand into the Central-West. Municipalities within clusters characterized by a low Human Development Index exhibited comparatively stronger increases in mortality trends for both cancers. These results underscore the need for more equitable and regionally tailored public policies to strengthen cancer control efforts in Brazil.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"55 ","pages":"Article 100758"},"PeriodicalIF":1.7,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145269594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Climate-driven potential for tularemia in East Africa: skill testing and ecological consistency of a transferred risk model","authors":"Komi Mensah Agboka ,&nbsp;Allan Muohi Ngángá ,&nbsp;Bonoukpoè Mawuko Sokame ,&nbsp;Steve Soh Bernard Baleba ,&nbsp;Tobias Landmann ,&nbsp;Elfatih M. Abdel-Rahman ,&nbsp;Chrysantus M. Tanga ,&nbsp;Souleymane Diallo","doi":"10.1016/j.sste.2025.100756","DOIUrl":"10.1016/j.sste.2025.100756","url":null,"abstract":"<div><div>Tularemia, a neglected zoonosis, remains underreported in Africa despite growing concern over its climate-driven expansion. This study aims to quantify the specific contribution of climate to tularemia risk using a climate attribution framework. We trained a Least Squares Dummy Variable (LSDV) fixed-effects panel model on United States (U.S.) county-level tularemia incidence data from 2011–2020 (n = 500, R² = 0.90), incorporating only climatic predictors: cumulative temperature, cumulative precipitation, and their respective variabilities. The climate-only model explained 86% of variance in the training data, demonstrating strong climate influence on tularemia disease dynamics. We then applied the model to East Africa, using environmental similarity analysis to assess transferability. Results show moderate-to-high climatic analogues in northern Kenya, eastern Uganda, and South Sudan. Between 2017 and 2020, predicted tularemia suitability increased by a median of +0.18 compared to the 2012–2015 baseline, particularly in arid and semi-arid zones. Low interannual variability suggests persistent climatic suitability. A thermal plausibility test showed strong agreement (r = 0.82) between predicted risk and the Gaussian thermal profile of <em>Francisella tularensis</em>. Our findings suggest that climate alone can spatially explain tularemia risk across Africa’s drylands. This method provides a transferable framework for early warning in data-poor regions and supports anticipatory surveillance in the context of climate change.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"55 ","pages":"Article 100756"},"PeriodicalIF":1.7,"publicationDate":"2025-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145120052","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Providing spatial support during a major cholera outbreak in Port-au-Prince, Haiti: Creative mapping solutions in a challenging data poor environment","authors":"Andrew Curtis ,&nbsp;Jayakrishnan Ajayakumar ,&nbsp;Rigan Louis ,&nbsp;Vanessa Rouzier ,&nbsp;J Glenn Morris Jr","doi":"10.1016/j.sste.2025.100753","DOIUrl":"10.1016/j.sste.2025.100753","url":null,"abstract":"<div><div>In this paper we describe the spatial data challenges faced in terms of providing accurate and timely analysis for a clinic during a cholera epidemic that spread through Port au Prince, Haiti in late 2022. This “triage” spatial epidemiology involved developing a bespoke geocoder that allowed for weekly maps of spread to be created in near real time. Resulting case data were also analyzed using a novel grid heatmapping approach which considers the epidemiological curve for each neighborhood. Adding further complexity during this period to both the data generation, and explaining cholera amplification and spread patterns, was a rising gang presence in the Port au Prince neighborhoods. Results identify a coastal pattern of amplification, which is expected given the informal settlement style living environments found in many of these neighborhoods. A second pattern then emerges of spread along a western and southern axis, which is far better captured in the grid heat mapping approach because of the lower numbers of patients seeking care at the clinic. The combination of traditional cartography and grid heat mapping help reveal the overall pattern of the epidemic, while also identifying key neighborhoods that require additional epidemiological investigation. Knowing why these neighborhoods played such an important role, possibly due to specific gang activity, is important in terms of understanding future disease spread in and around Port au Prince. Indeed, results presented can help contextualize official cholera reporting in 2025 where data availability is still hampered by ongoing gang rule.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"55 ","pages":"Article 100753"},"PeriodicalIF":1.7,"publicationDate":"2025-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145050553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatial patterns of testicular cancer diagnosis in Australia, 2010-2019","authors":"Charlotte K․ Bainomugisa ,&nbsp;Paramita Dasgupta ,&nbsp;Jessica K. Cameron ,&nbsp;Ben Tran ,&nbsp;Susanna M. Cramb ,&nbsp;Peter D. Baade","doi":"10.1016/j.sste.2025.100745","DOIUrl":"10.1016/j.sste.2025.100745","url":null,"abstract":"<div><h3>Aim</h3><div>To investigate the spatial patterns of the incidence rates of testicular cancer, and broad regional differences in survival, between 2010 and 2019 in Australia using national population-based cancer registry data.</div></div><div><h3>Methods</h3><div>Incidence data including residential location at diagnosis were obtained from the Australian Cancer Database, with mortality followed-up until end of 2019. Incidence spatial patterns were modelled using Bayesian spatial Leroux models and spatial heterogeneity tested using the maximised excess events test. Relative survival rates by broad region were modelled using flexible parametric relative survival models.</div></div><div><h3>Results</h3><div>From all the notifications of testicular cancer (<em>n</em> = 8217), the age-standardized incidence rate was 8.9 cases per 100,000 males each year. We found evidence of significant spatial variation in the incidence of testicular cancer across small geographical areas, with some areas including those in Tasmania showing standardised incidence ratios above the national average. The 5-year relative survival estimate was 97.5 % [95 % CI: 97.1–97.9].</div></div><div><h3>Conclusion</h3><div>There is a need to raise awareness of testicular cancer in high-risk geographical areas and age groups, and to conduct further research into drivers of localised spatial patterns.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"55 ","pages":"Article 100745"},"PeriodicalIF":1.7,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145050554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatio-temporal methods to handle missing data in syndromic surveillance with applications to health management information system data","authors":"Nicholas B. Link ,&nbsp;Anuraag Gopaluni ,&nbsp;Isabel Fulcher ,&nbsp;Emma Jean Boley ,&nbsp;Rachel C. Nethery ,&nbsp;Bethany Hedt-Gauthier","doi":"10.1016/j.sste.2025.100736","DOIUrl":"10.1016/j.sste.2025.100736","url":null,"abstract":"<div><div>Syndromic surveillance monitors infectious diseases, especially in situations where direct disease monitoring is unavailable. However, conventional syndromic surveillance methods face challenges in handling missing data, particularly when the missing completely at random (MCAR) assumption is violated. Additionally, these methods often do not leverage spatio-temporal techniques that can reduce bias and improve their performance. This study addresses both of these limitations by comparing a baseline syndromic surveillance model with a frequentist spatio-temporal model used in infectious diseases and a Bayesian spatio-temporal conditional autoregressive (CAR) model.</div><div>Drawing inspiration from COVID-19 symptom data collected via routine health systems in Liberia, we conduct simulations with various data generating processes, spatio-temporal correlation structures, and missing data mechanisms. Across the diverse simulations for outbreak detection, the baseline model and the Bayesian CAR model had high specificity, thus limiting outbreak false alarms. The findings underscore the importance of considering spatio-temporal models for syndromic surveillance.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"54 ","pages":"Article 100736"},"PeriodicalIF":1.7,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144779656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geostatistical and machine learning approaches for high-resolution mapping of vaccination coverage","authors":"C. Edson Utazi ,&nbsp;Ortis Yankey ,&nbsp;Somnath Chaudhuri ,&nbsp;Iyanuloluwa D. Olowe ,&nbsp;M. Carolina Danovaro-Holliday ,&nbsp;Attila N. Lazar ,&nbsp;Andrew J. Tatem","doi":"10.1016/j.sste.2025.100744","DOIUrl":"10.1016/j.sste.2025.100744","url":null,"abstract":"<div><div>Recently, there has been a growing interest in the production of high-resolution maps of vaccination coverage. These maps have been useful for uncovering geographic inequities in coverage and improving targeting of interventions to reach marginalized populations. Different methodological approaches have been developed for producing these maps using mostly geolocated household survey data and geospatial covariate information. However, it remains unclear how much the predicted coverage maps produced by the various methods differ, and which methods yield more reliable estimates. Here, we explore the predictive performance of these methods and resulting implications for spatial prioritization to fill this gap. Using Nigeria Demographic and Health Survey as a case study, we generate 1 × 1 km and district level maps of indicators of vaccination coverage using geostatistical, machine learning (ML) and hybrid methods and evaluate predictive performance via cross-validation. Our results show similar predictive performance for five of the seven methods investigated, although two geostatistical approaches are the best performing methods. The worst-performing methods are two ML approaches. We find marked differences in spatial prioritization using these methods, which could potentially result in missing important underserved populations, although broad similarities exist. Our study can help guide map production for other health and development metrics.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"54 ","pages":"Article 100744"},"PeriodicalIF":1.7,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Variable Screening Methods in Conditional Logistic Individual Level Models of Disease Spread","authors":"Tahmina Akter ,&nbsp;Rob Deardon","doi":"10.1016/j.sste.2025.100742","DOIUrl":"10.1016/j.sste.2025.100742","url":null,"abstract":"<div><div>The conditional logistic individual-level model is a recently developed infectious disease model, particularly suited for modeling spatial-based infection risk. It is designed to reduce computational complexity and expand the range of available statistical software for data analysis (Akter &amp; Deardon, 2025). This study aims to apply and evaluate different variable selection techniques for the newly introduced conditional logistic individual-level models (CL-ILMs). These variable selection methods include forward and backward stepwise Akaike information criterion (AIC), least absolute shrinkage and selection operator (Lasso), spike-and-slab prior (SS prior), and two-stage screening methods. The ultimate goal is to boost model performance and interpretability, and to reduce the risk of overfitting ultimately leading to more robust and effective models. We examine and compare the performance of these methods using simulated data and real-life data from the outbreak of foot-and-mouth disease in the UK in 2001.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"54 ","pages":"Article 100742"},"PeriodicalIF":1.7,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144893311","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The causal impact of urbanicity on neighbourhood psychosis prevalence","authors":"Peter Congdon","doi":"10.1016/j.sste.2025.100739","DOIUrl":"10.1016/j.sste.2025.100739","url":null,"abstract":"<div><div>There is considerable evidence of elevated psychosis rates in more urban settings. However, the urbanicity effect is confounded with other neighbourhood contextual effects, such as from deprivation and crime. To assess the nature of the underlying urbanicity effect, removing distorting effects of confounders, we consider a novel method to assessing causality in spatial applications: a propensity weight approach, with weights obtained by entropy optimization, and adjusting for the spatial overlap in the urbanicity effect via a bivariate exposure approach. The application is to the effect of urbanicity on psychosis prevalence in 6856 English neighbourhoods. We use a measure of urbanicity adapted to represent aspects of urban form, rather than simply population density or a binary indicator. The overlap effect in the psychosis outcome model is shown to outweigh the local effect, and we find a clear urbanicity gradient with a relative risk of 1.91 comparing the most and least urban areas, after adjustment for confounding through propensity weighting.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"54 ","pages":"Article 100739"},"PeriodicalIF":1.7,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatial clustering and sociodemographic factors impacting obesity and hypertension in Nepal: Analysis of a national demographic and health survey, 2022","authors":"Biraj Neupane ,&nbsp;Bikram Adhikari ,&nbsp;Niharika Jha ,&nbsp;Ian Brooks ,&nbsp;Csaba Varga","doi":"10.1016/j.sste.2025.100743","DOIUrl":"10.1016/j.sste.2025.100743","url":null,"abstract":"<div><h3>Background</h3><div>Overweight/obesity and hypertension pose significant global health challenges. This study examines the spatial distribution, sociodemographic determinants, and clustering patterns of these conditions in Nepal.</div></div><div><h3>Methods</h3><div>We conducted a comprehensive spatial-epidemiological analysis of 136,235 participants from the 2022 Nepal Demographic and Health Survey. Outcome variables in this study were overweight/obesity (present/absent) and hypertension (present/absent). A weighted descriptive and inferential analysis addressed the complex survey design and non-response rate. We used spatial scan statistics to identify areas with higher or lower-than-expected cases, and geospatial mapping to illustrate the distribution of cases and the significant spatial clusters. Multivariable logistic regression models determine the association between the outcome variables and respondents’ age, gender, marital status, education level, and wealth.</div></div><div><h3>Findings</h3><div>Overall, 42.5 % of respondents were obese, and 38.5 % had hypertension. Respondents who were women, middle-aged, married, educated, wealthy, and living in cities had higher odds of being overweight. Similarly, respondents who were male, older, single, poor, uneducated, and lived in cities had higher odds of having hypertension. A spatial scan statistic using the Bernoulli model identified twelve (seven low and five high rate) significant clusters for obesity and eleven (five low and six high rate) for hypertension.</div></div><div><h3>Conclusion</h3><div>This study showed the utility of health risk mapping across Nepal, emphasizing the complex interaction between sociodemographic and geographic factors impacting the prevalence of obesity and hypertension. The findings highlighted the need for targeted interventions in the high-risk regions of Nepal based on the identified risk factors to mitigate the impact.</div></div>","PeriodicalId":46645,"journal":{"name":"Spatial and Spatio-Temporal Epidemiology","volume":"54 ","pages":"Article 100743"},"PeriodicalIF":1.7,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144886714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}