Bayesian analysis of space-time variation in disease risk

Towards development of functional climate-driven early warning systems for climate-sensitive infectious diseases: Statistical models and recommendations

Climate, weather and environmental change have significantly influenced patterns of infectious disease transmission, necessitating the development of early warning systems to anticipate potential impacts and respond in a timely and effective way. Statistical modelling plays a pivotal role in understanding the intricate relationships between climatic factors and infectious disease transmission. For example, time series regression modelling and spatial cluster analysis have been employed to identify risk factors and predict spatial and temporal patterns of infectious diseases. Recently advanced spatio-temporal models and machine learning offer an increasingly robust framework for modelling uncertainty, which is essential in climate-driven disease surveillance due to the dynamic and multifaceted nature of the data. Moreover, Artificial Intelligence (AI) techniques, including deep learning and neural networks, excel in capturing intricate patterns and hidden relationships within climate and environmental data sets. Web-based data has emerged as a powerful complement to other datasets encompassing climate variables and disease occurrences. However, given the complexity and non-linearity of climate-disease interactions, advanced techniques are required to integrate and analyse these diverse data to obtain more accurate predictions of impending outbreaks, epidemics or pandemics. This article presents an overview of an approach to creating climate-driven early warning systems with a focus on statistical model suitability and selection, along with recommendations for utilizing spatio-temporal and machine learning techniques. By addressing the limitations and embracing the recommendations for future research, we could enhance preparedness and response strategies, ultimately contributing to the safeguarding of public health in the face of evolving climate challenges.

Effects of the Sacramento Neighborhood Alcohol Prevention Project on rates of child abuse and neglect 7 years post-implementation (1999-2010)

INTRODUCTION

Evaluations of alcohol environmental prevention efforts examine short-term effects of these interventions on alcohol-related problems. We examine whether the effects of the Sacramento Neighborhood Alcohol Prevention Project (SNAPP), an alcohol environmental intervention aimed to reduce alcohol-related problems in two neighbourhoods, on child abuse and neglect remained 7 years post-implementation.

METHODS

SNAPP used a quasi-experimental non-equivalent control group design, where intervention activities occurred in the South area, followed by those in the North area 2 years later. Our sample size is 3912 space-time units (326 census block groups × 12 years [1999-2010]). Outcomes were measured at the household level and included: (i) all foster care entries total; and (ii) the subset of foster care entries that were alcohol related. Data were analysed using Bayesian conditionally autoregressive space-time models.

RESULTS

We find that the decreases in total (relative rate [RR] = 0.882, 95% credible interval [CrI] 0.795, 0.980) and alcohol-related (RR = 0.888, 95% CrI 0.791, 0.997) foster care entries remain in the North intervention area although the magnitude of those changes are smaller than immediately post-intervention. Increases found in alcohol-related foster care entries in the South area immediately post-intervention were not significant 7 years later (RR = 1.128, 95% CrI 0.975, 1.307).

DISCUSSION AND CONCLUSIONS

Reductions in child abuse and neglect due to an alcohol environmental intervention can be maintained. Environmental interventions that provide community-level primary prevention strategies could be more easily sustained and more cost effective than individual-level interventions, although more research is needed to identify why interventions may be successful in specific contexts and not others.

Relative risk of childhood and adolescence cancer in Iran: spatiotemporal analysis from 1999 to 2016

OBJECTIVE

Cancer is the third leading cause of death in the world with increasing trends in Iran. The study of epidemiology, trend, and geospatial distribution of pediatric cancers provides important information for screening as well as early detection of cancer and policy making. We aimed to assess the spatio-temporal disparity of childhood and adolescence cancer risk among provinces of Iran.

METHODS

In this retrospective study, we estimated geospatial relative risk (RR) of childhood cancer in provinces of Iran using data from 29198 cases. We used BYM and its extended spatiotemporal model in Bayesian setting. This hierarchical model takes spatial and temporal effects into account in the incidence rate estimation simultaneously.

RESULTS

The relative risk of cancer was > 1 for 45% of the provinces, where 27% of provinces had significantly ascending trend. North Khorasan, Yazd and Qazvin provinces had the highest risk rates while Sistan-Baluchistan province showed the lowest risk of cancer. However, the differential trends was highest in Sistan-Baluchistan, Bushehr, Hormozgan, and Kohgilouyeh-Boyerahmad. Both the point estimate and the trend of risk was high in Tehran.

CONCLUSION

The geographic pattern and trend of cancer in children seems to be different from that in adults that urges further studies. This could lead to increased health system capacity and facilitate the access to effective detection, research, care and treatment of childhood cancer.

The Association between the Percentage of Female Law Enforcement Officers and Rape Report, Clearance, and Arrest Rates: A Spatiotemporal Analysis of California

Rape is an underreported violent crime that frequently remains uncleared (open) in the legal system. Rape disproportionately affects women, with 91% of rape victim-survivors estimated to be female. However, law enforcement agencies, the entry point into the criminal justice system, are predominantly comprised of male officers. According to the theory of representative bureaucracy, groups with greater representation in a bureaucratic system are more likely to have their interests protected. This study aims to determine if California law enforcement agencies with a higher percentage of female officers are more likely to have higher rates of rape reporting, clearances, and arrests. No previous study has examined this relationship using statewide data. Crimes and Clearances, Monthly Arrest and Citation Register, and Uniform Crime Reporting data for California (2013-2016) were aggregated into 499 Law Enforcement Reporting Areas (LERA). Bayesian space-time Poisson regressions controlling for LERA demographics and crime produced scaled relative rates for three outcomes: (a) rape report rate: number of reports relative to population ages 18+; (b) rape clearance rate: number of clearances relative to reports; and (c) rape arrest rate: number of arrests for rape relative to reports. A 5% increase in the percentage of female officers within an agency was associated with a 6.2% increase in the rape report rate (ARR: 1.062, 95% credible interval (CI) [1.048, 1.077]), a 2.9% decrease in the clearance rate (ARR: 0.971 95% CI [0.950, 0.993]), and no change in the rape arrest rates (ARR: 1.010; 95% CI [0.981, 1.039]) across all LERA. Thus, increased female officer representation was associated with an increase in rape reporting rates but associated with a decrease in rape clearance rates. The theory of representative bureaucracy was only partially supported, and these relationships may not be causal. The quantity of rape reports received by an agency, employment and promotion practices of agencies, and victim-survivor's attitudes toward officer's gender should also be considered.

A scalable approach for short-term disease forecasting in high spatial resolution areal data

Short-term disease forecasting at specific discrete spatial resolutions has become a high-impact decision-support tool in health planning. However, when the number of areas is very large obtaining predictions can be computationally intensive or even unfeasible using standard spatiotemporal models. The purpose of this paper is to provide a method for short-term predictions in high-dimensional areal data based on a newly proposed "divide-and-conquer" approach. We assess the predictive performance of this method and other classical spatiotemporal models in a validation study that uses cancer mortality data for the 7907 municipalities of continental Spain. The new proposal outperforms traditional models in terms of mean absolute error, root mean square error, and interval score when forecasting cancer mortality 1, 2, and 3 years ahead. Models are implemented in a fully Bayesian framework using the well-known integrated nested Laplace estimation technique.

Evolution of spatial disease clusters via a Bayesian space-time variability modelling

This study proposes to use exceedance posterior probabilities of a space-time random-effects model to study the temporal dynamics of clusters. The local time trends specified for each area is further smoothed over space. We modelled the common spatial and the space-varying temporal trend using a multivariate Markov Random field to incorporate within-area correlations. We estimate the model parameters within a fully Bayesian framework. The exceedance posterior probabilities are further used to classify the common spatial trend into hot-spots, cold-spots, and neutral-spots. The local time trends are classified into increasing, decreasing, and stable trends. The results is a 3×3 table depicting the time trends within clusters. As a demonstration, we apply the proposed methodology to study the evolution of spatial clustering of intestinal parasite infections in Ghana. We find the methodology presented in this paper applicable and extendable to other or multiple tropical diseases which may have different space-time conceptualizations.

Bayesian spatiotemporal modelling for disease mapping: an application to preeclampsia and gestational diabetes in Florida, United States

Morbidities generally show patterns of concentration that vary by space and time. Disease mapping models are useful in estimating the spatiotemporal patterns of disease risks and are therefore pivotal for effective disease surveillance, resource allocation, and the development of prevention strategies. This study considers six spatiotemporal Bayesian hierarchical models based on two spatial conditional autoregressive priors. It could serve as a guideline on the development and application of Bayesian hierarchical models to assess the emerging risk trends, risk clustering, and spatial inequality trends, with estimation of covariables' effects on the interested disease risk. The method is applied to the Florida Birth Record data between 2006 and 2015 to study two cardiovascular risk factors: preeclampsia and gestational diabetes. High-risk clusters were detected in North Central Florida for preeclampsia and in Central Florida for gestational diabetes. While the adjusted disease trend was stable, spatial inequality peaked in 2011-2012 for both diseases. Exposure to PM2.5 at first or/and second trimester increased the risk of preeclampsia and gestational diabetes, but the magnitude is less severe compared to previous studies. In conclusion, this study underscores the significance of selecting appropriate disease mapping models in estimating the intricate spatiotemporal patterns of disease risk and suggests the importance of localized interventions to reduce health disparities. The result also identified an opportunity to study potential risk factors of preeclampsia, as the spike of risk in North Central Florida cannot be explained by current covariables.

Changes in Opioid and Benzodiazepine Poisoning Deaths After Cannabis Legalization in the US: A County-level Analysis, 2002-2020

BACKGROUND

Cannabis legalization for medical and recreational purposes has been suggested as an effective strategy to reduce opioid and benzodiazepine use and deaths. We examined the county-level association between medical and recreational cannabis laws and poisoning deaths involving opioids and benzodiazepines in the US from 2002 to 2020.

METHODS

Our ecologic county-level, spatiotemporal study comprised 49 states. Exposures were state-level implementation of medical and recreational cannabis laws and state-level initiation of cannabis dispensary sales. Our main outcomes were poisoning deaths involving any opioid, any benzodiazepine, and opioids with benzodiazepines. Secondary analyses included overdoses involving natural and semi-synthetic opioids, synthetic opioids, and heroin.

RESULTS

Implementation of medical cannabis laws was associated with increased deaths involving opioids (rate ratio [RR] = 1.14; 95% credible interval [CrI] = 1.11, 1.18), benzodiazepines (RR = 1.19; 95% CrI = 1.12, 1.26), and opioids+benzodiazepines (RR = 1.22; 95% CrI = 1.15, 1.30). Medical cannabis legalizations allowing dispensaries was associated with fewer deaths involving opioids (RR = 0.88; 95% CrI = 0.85, 0.91) but not benzodiazepine deaths; results for recreational cannabis implementation and opioid deaths were similar (RR = 0.81; 95% CrI = 0.75, 0.88). Recreational cannabis laws allowing dispensary sales was associated with consistent reductions in opioid- (RR = 0.83; 95% CrI = 0.76, 0.91), benzodiazepine- (RR = 0.79; 95% CrI = 0.68, 0.92), and opioid+benzodiazepine-related poisonings (RR = 0.83; 95% CrI = 0.70, 0.98).

CONCLUSIONS

Implementation of medical cannabis laws was associated with higher rates of opioid- and benzodiazepine-related deaths, whereas laws permitting broader cannabis access, including implementation of recreational cannabis laws and medical and recreational dispensaries, were associated with lower rates. The estimated effects of the expanded availability of cannabis seem dependent on the type of law implemented and its provisions.

Spatiotemporal bayesian modelling of scorpionism and its risk factors in the state of São Paulo, Brazil

BACKGROUND

Scorpion stings in Brazil represent a major public health problem due to their incidence and their potential ability to lead to severe and often fatal clinical outcomes. A better understanding of scorpionism determinants is essential for a precise comprehension of accident dynamics and to guide public policy. Our study is the first to model the spatio-temporal variability of scorpionism across municipalities in São Paulo (SP) and to investigate its relationship with demographic, socioeconomic, environmental, and climatic variables.

METHODOLOGY

This ecological study analyzed secondary data on scorpion envenomation in SP from 2008 to 2021, using the Integrated Nested Laplace Approximation (INLA) to perform Bayesian inference for detection of areas and periods with the most suitable conditions for scorpionism.

PRINCIPAL FINDINGS

From the spring of 2008 to 2021, the relative risk (RR) increased eight times in SP, from 0.47 (95%CI 0.43-0.51) to 3.57 (95%CI 3.36-3.78), although there has been an apparent stabilization since 2019. The western, northern, and northwestern parts of SP showed higher risks; overall, there was a 13% decrease in scorpionism during winters. Among the covariates considered, an increase of one standard deviation in the Gini index, which captures income inequality, was associated with a 11% increase in scorpion envenomation. Maximum temperatures were also associated with scorpionism, with risks doubling for temperatures above 36°C. Relative humidity displayed a nonlinear association, with a 50% increase in risk for 30-32% humidity and reached a minimum of 0.63 RR for 75-76% humidity.

CONCLUSIONS

Higher temperatures, lower humidity, and social inequalities were associated with a higher risk of scorpionism in SP municipalities. By capturing local and temporal relationships across space and time, authorities can design more effective strategies that adhere to local and temporal considerations.

Projecting malaria elimination in Thailand using Bayesian hierarchical spatiotemporal models

Thailand has set a goal of eliminating malaria by 2024 in its national strategic plan. In this study, we used the Thailand malaria surveillance database to develop hierarchical spatiotemporal models to analyze retrospective patterns and predict Plasmodium falciparum and Plasmodium vivax malaria incidences at the provincial level. We first describe the available data, explain the hierarchical spatiotemporal framework underlying the analysis, and then display the results of fitting various space-time formulations to the malaria data with the different model selection metrics. The Bayesian model selection process assessed the sensitivity of different specifications to obtain the optimal models. To assess whether malaria could be eliminated by 2024 per Thailand's National Malaria Elimination Strategy, 2017-2026, we used the best-fitted model to project the estimated cases for 2022-2028. The study results based on the models revealed different predicted estimates between both species. The model for P. falciparum suggested that zero P. falciparum cases might be possible by 2024, in contrast to the model for P. vivax, wherein zero P. vivax cases might not be reached. Innovative approaches in the P. vivax-specific control and elimination plans must be implemented to reach zero P. vivax and consequently declare Thailand as a malaria-free country.

Spatiotemporal analysis of the effect of global development indicators on child mortality

BACKGROUND

Child mortality continue to be a major public health issue in most developing countries; albeit there has been a decline in global under-five deaths. The differences in child mortality can best be explained by socioeconomic and environmental inequalities among countries. In this study, we explore the effect of country-level development indicators on under-five mortality rates. Specifically, we examine potential spatio-temporal heterogeneity in the association between major world development indicators on under-five mortality, as well as, visualize the global differential time trend of under-five mortality rates.

METHODS

The data from 195 countries were curated from the World Bank's World Development Indicators (WDI) spanning from 2000 to 2017 and national estimates for under-five mortality from the UN Inter-agency Group for Child Mortality Estimation (UN IGME).We built parametric and non-parametric Bayesian space-time interaction models to examine the effect of development indicators on under-five mortality rates. We also used employed Bayesian spatio-temporal varying coefficient models to assess the spatial and temporal variations in the effect of development indicators on under-five mortality rates.

RESULTS

In both parametric and non-parametric models, the results show indicators of good socioeconomic development were associated with a reduction in under-five mortality rates while poor indicators were associated with an increase in under-five mortality rates. For instance, the parametric model shows that gross domestic product (GDP) (β = - 1.26, [CI - 1.51; - 1.01]), current healthcare expenditure (β = - 0.40, [CI - 0.55; - 0.26]) and access to basic sanitation (β = - 0.03, [CI - 0.05; - 0.01]) were associated with a reduction under-five mortality. An increase in the proportion practising open defecation (β = 0.14, [CI 0.08; 0.20]) an increase under-five mortality rate. The result of the spatial components spatial variation in the effect of the development indicators on under-five mortality rates. The spatial patterns of the effect also change over time for some indicators, such as PM2.5.

CONCLUSION

The findings show that the burden of under-five mortality rates was considerably higher among sub-Saharan African countries and some southern Asian countries. The findings also reveal the trend in reduction in the sub-Saharan African region has been slower than the global trend.

Spatio-temporal distribution characteristics of the risk of viral hepatitis B incidence based on INLA in 14 prefectures of Xinjiang from 2004 to 2019

This study aimed to explore the spatio-temporal distribution characteristics and risk factors of hepatitis B (HB) in 14 prefectures of Xinjiang, China, and to provide a relevant reference basis for the prevention and treatment of HB. Based on HB incidence data and risk factor indicators in 14 prefectures in Xinjiang from 2004 to 2019, we explored the distribution characteristics of the risk of HB incidence using global trend analysis and spatial autocorrelation analysis and established a Bayesian spatiotemporal model to identify the risk factors of HB and their spatio-temporal distribution to fit and extrapolate the Bayesian spatiotemporal model using the Integrated Nested Laplace Approximation (INLA) method. There was spatial autocorrelation in the risk of HB and an overall increasing trend from west to east and north to south. The natural growth rate, per capita GDP, number of students, and number of hospital beds per 10, 000 people were all significantly associated with the risk of HB incidence. From 2004 to 2019, the risk of HB increased annually in 14 prefectures in Xinjiang, with Changji Hui Autonomous Prefecture, Urumqi City, Karamay City, and Bayangol Mongol Autonomous Prefecture having the highest rates.

Hierarchical spatiotemporal modeling of human visceral leishmaniasis in Rio Grande do Norte, Brazil

Visceral leishmaniasis (VL) is a neglected tropical disease that is globally distributed and has the potential to cause very serious illness. Prior literature highlights the emergence and spread of VL is influenced by multiple factors, such as socioeconomic status, sanitation levels or animal and human reservoirs. The study aimed to retrospectively investigate the presence and infectiousness of VL in Rio Grande do Norte (RN), Brazil between 2007 and 2020. We applied a hierarchical Bayesian approach to estimate municipality-specific relative risk of VL across space and time. The results show evidence that lower socioeconomic status is connected to higher municipality-specific VL risk. Overall, estimates reveal spatially heterogeneous VL risks in RN, with a high probability that VL risk for municipalities within the West Potiguar mesoregion are more than double the expected VL risk. Additionally, given the data available, results indicate there is a high probability of increasing VL risk in the municipalities of Natal, Patu and Pau dos Ferros. These findings demonstrate opportunities for municipality-specific public health policy interventions and warrant future research on identifying epidemiological drivers in at-risk regions.

A Spatio-temporal Bayesian model to estimate risk and influencing factors related to tuberculosis in Chongqing, China, 2014-2020

BACKGROUND

Tuberculosis (TB) is a serious infectious disease that is one of the leading causes of death worldwide. This study aimed to investigate the spatial and temporal distribution patterns and potential influencing factors of TB incidence risk, and to provide a scientific basis for the prevention and control of TB.

METHODS

We collected reported cases of TB in 38 districts and counties in Chongqing from 2014 to 2020 and data on environment, population characteristics and economic factors during the same period. By constructing a Bayesian spatio-temporal model, we explored the spatio-temporal distribution pattern of TB incidence risk and potential influencing factors, identified key areas and key populations affected by TB, compared the spatio-temporal distribution characteristics of TB in populations with different characteristics, and explored the differences in the influence of various social and environmental factors.

RESULTS

The high-risk areas for TB incidence in Chongqing from 2014 to 2020 were mainly concentrated in southeastern and northeastern regions of Chongqing, and the overall relative risk (RR) of TB showed a decreasing trend during the study period, while RR of TB in main urban area and southeast of Chongqing showed an increasing trend. The RR of TB was relatively high in the main urban area for the female population and the population aged 0-29 years, and the RR of TB for the population aged 30-44 years in the main urban area and the population aged 60 years or older in southeast of Chongqing had an increasing trend, respectively. For each 1 μg/m³ increase in SO₂ and 1% increase in the number of low-income per 1000 non-agricultural households (LINA per 1000 persons), the RR of TB increased by 0.35% (95% CI: 0.08-0.61%) and 0.07% (95% CI: 0.05-0.10%), respectively. And LINA per 1000 persons had the greatest impact on the female population and the over 60 years old age group. Although each 1% increase in urbanization rate (UR) was associated with 0.15% (95% CI: 0.11-0.17%) reduction in the RR of TB in the whole population, the RR increased by 0.18% (95% CI: 0.16-0.21%) in the female population and 0.37% (95% CI: 0.34-0.45%) in the 0-29 age group.

CONCLUSION

This study showed that high-risk areas for TB were concentrated in the southeastern and northeastern regions of Chongqing, and that the elderly population was a key population for TB incidence. There were spatial and temporal differences in the incidence of TB in populations with different characteristics, and various socio-environmental factors had different effects on different populations. Local governments should focus on areas and populations at high risk of TB and develop targeted prevention interventions based on the characteristics of different populations.

Risk Factors and Spatiotemporal Analysis of Classical Swine Fever in Ecuador

Classical swine fever (CSF) is one of the most important re-emergent swine diseases worldwide. Despite concerted control efforts in the Andean countries, the disease remains endemic in several areas, limiting production and trade opportunities. In this study, we aimed to determine the risk factors and spatiotemporal implications associated with CSF in Ecuador. We analysed passive surveillance and vaccination campaign datasets from 2014 to 2020; Then, we structured a herd-level case-control study using a logistic and spatiotemporal Bayesian model. The results showed that the risk factors that increased the odds of CSF occurrence were the following: swill feeding (OR 8.53), time until notification (OR 2.44), introduction of new pigs during last month (OR 2.01) and lack of vaccination against CSF (OR 1.82). The spatiotemporal model showed that vaccination reduces the risk by 33%. According to the priority index, the intervention should focus on Morona Santiago and Los Rios provinces. In conclusion, the results highlight the complexity of the CSF control programs, the importance to improve the overall surveillance system and the need to inform decision-makers and stakeholders.

Spatial disparity and factors associated with dementia mortality: A cross-sectional study in Zhejiang Province, China

Objective

Evidence of spatial disparity in dementia mortality in China has been found to have higher dementia mortality in eastern and rural China. Regional factors of physical and social features may be influencing this spatial disparity. However, the extent of spatial difference in dementia mortality across small regional localities is unclear. This study aims to investigate the geographic variations in mortality and risk of all dementia subtypes and identify the effect of the associated environmental risk factors.

Methods

We used surveillance data on death reports from Alzheimer's disease and other forms of dementia in Zhejiang province from 2015 to 2019. We estimated the relative risk of dementia mortality using a Bayesian spatial model. We mapped predicted relative risk to visualize the risk of death from different types of dementia and to identify risk factors associated with dementia.

Results

Thirty thousand three hundred and ninety-eight deaths attributable to dementia as the underlying or related cause (multiple causes) were reported during 2015-2019. Counties and districts in the southeast and west of Zhejiang province had significantly higher standardized mortality ratios than others. Counties and districts with a smaller proportion of residents aged 60 years or older, poorer economic status, insufficient health resources, and worse pollution had a higher risk of deaths due to dementia.

Conclusion

Higher risks of dementia mortality were found in counties and districts with poorer economic status, insufficient health resources, and worse pollution in Zhejiang. Our study adds new evidence on the association between socioeconomic and environmental factors and the mortality risk due to dementia.

Spatiotemporal trends and ecological determinants of cardiovascular mortality among 2844 counties in mainland China, 2006-2020: a Bayesian modeling study of national mortality registries

BACKGROUND

Cardiovascular disease (CVD) is the leading cause of death in China. No previous study has reported CVD mortality at county-level, and little was known about the nonmedical ecological factors of CVD mortality at such small scale in mainland China. Understanding the spatiotemporal variations of CVD mortality and examining its nonmedical ecological factors would be of great importance to tailor local public health policies.

METHODS

By using national mortality registration data in China, this study used hierarchical spatiotemporal Bayesian model to demonstrate spatiotemporal distribution of CVD mortality in 2844 counties during 2006 to 2020 and investigate how nonmedical ecological determinants have affected CVD mortality inequities from the spatial perspectives.

RESULTS

During 2006-2020, the age-standardized mortality rate (ASMR) of CVD decreased from 284.77 per 100,000 in 2006 to 241.34 per 100,000 in 2020. Among 2844 counties, 1144 (40.22%) were hot spots counties with a higher CVD mortality risk compared to the national average and located mostly in northeast, north central, and westernmost regions; on the contrary, 1551 (54.53%) were cold spots counties and located mostly in south and southeast coastal counties. CVD mortality risk decreased from 2006 to 2020 was larger in counties where CVD mortality rate had been higher in 2006 in most of the counties, vice versa. Nationwide, nighttime light intensity (NTL) was the major influencing factor of CVD mortality, a higher NTL appeared to be negatively associated with a lower CVD mortality, with one unit increase in NTL, and the CVD mortality risk will decrease 11% (relative risk of NTL was estimated as 0.89 with 95% confidence interval of 0.83-0.94).

CONCLUSIONS

Substantial between-county discrepancies of CVD mortality distribution were observed during past 15 years in mainland China. Nonmedical ecological determinants were estimated to significantly explain the overall and local spatiotemporal patterns of this CVD mortality risk. Targeted considerations are needed to integrate primary care with clinical care through intensifying further strategies to narrow unequally distribution of CVD mortality at local scale. The approach to county-level analysis with small area models has the potential to provide novel insights into Chinese disease-specific mortality burden.

Bayesian spatio-temporal modelling and mapping of malaria and anaemia among children between 0 and 59 months in Nigeria

BACKGROUND/M&M

A vital aspect of disease management and policy making lies in the understanding of the universal distribution of diseases. Nevertheless, due to differences all-over host groups and space-time outbreak activities, data are subject to intricacies. Herein, Bayesian spatio-temporal models were proposed to model and map malaria and anaemia risk ratio in space and time as well as to ascertain risk factors related to these diseases and the most endemic states in Nigeria. Parameter estimation was performed by employing the R-integrated nested Laplace approximation (INLA) package and Deviance Information Criteria were applied to select the best model.

RESULTS

In malaria, model 7 which basically suggests that previous trend of an event cannot account for future trend i.e., Interaction with one random time effect (random walk) has the least deviance. On the other hand, model 6 assumes that previous event can be used to predict future event i.e., (Interaction with one random time effect (ar1)) gave the least deviance in anaemia.

DISCUSSION

For malaria and anaemia, models 7 and 6 were selected to model and map these diseases in Nigeria, because these models have the capacity to receive strength from adjacent states, in a manner that neighbouring states have the same risk. Changes in risk and clustering with a high record of these diseases among states in Nigeria was observed. However, despite these changes, the total risk of malaria and anaemia for 2010 and 2015 was unaffected.

CONCLUSION

Notwithstanding the methods applied, this study will be valuable to the advancement of a spatio-temporal approach for analyzing malaria and anaemia risk in Nigeria.

Small area disease mapping of cancer incidence in British Columbia using Bayesian spatial models and the smallareamapp R Package

INTRODUCTION

There is an increasing interest in small area analyses in cancer surveillance; however, technical capacity is limited and accessible analytical approaches remain to be determined. This study demonstrates an accessible approach for small area cancer risk estimation using Bayesian hierarchical models and data visualization through the smallareamapp R package.

MATERIALS AND METHODS

Incident lung (N = 26,448), female breast (N = 28,466), cervical (N = 1,478), and colorectal (N = 25,457) cancers diagnosed among British Columbia (BC) residents between 2011 and 2018 were obtained from the BC Cancer Registry. Indirect age-standardization was used to derive age-adjusted expected counts and standardized incidence ratios (SIRs) relative to provincial rates. Moran's I was used to assess the strength and direction of spatial autocorrelation. A modified Besag, York and Mollie model (BYM2) was used for model incidence counts to calculate posterior median relative risks (RR) by Community Health Service Areas (CHSA; N = 218), adjusting for spatial dependencies. Integrated Nested Laplace Approximation (INLA) was used for Bayesian model implementation. Areas with exceedance probabilities (above a threshold RR = 1.1) greater or equal to 80% were considered to have an elevated risk. The posterior median and 95% credible intervals (CrI) for the spatially structured effect were reported. Predictive posterior checks were conducted through predictive integral transformation values and observed versus fitted values.

RESULTS

The proportion of variance in the RR explained by a spatial effect ranged from 4.4% (male colorectal) to 19.2% (female breast). Lung cancer showed the greatest number of CHSAs with elevated risk (Nwomen = 50/218, Nmen = 44/218), representing 2357 total excess cases. The largest lung cancer RRs were 1.67 (95% CrI = 1.06-2.50; exceedance probability = 96%; cases = 13) among women and 2.49 (95% CrI = 2.14-2.88; exceedance probability = 100%; cases = 174) among men. Areas with small population sizes and extreme SIRs were generally smoothed towards the null (RR = 1.0).

DISCUSSION

We present a ready-to-use approach for small area cancer risk estimation and disease mapping using BYM2 and exceedance probabilities. We developed the smallareamapp R package, which provides a user-friendly interface through an R-Shiny application, for epidemiologists and surveillance experts to examine geographic variation in risk. These methods and tools can be used to estimate risk, generate hypotheses, and examine ecologic associations while adjusting for spatial dependency.

Spatiotemporal Distribution of HIV Self-testing Kits Purchased on the Web and Implications for HIV Prevention in China: Population-Based Study

BACKGROUND

HIV self-testing (HIVST) holds great promise for expanding HIV testing. Nonetheless, large-scale data on HIVST behavior are scant. Millions of HIVST kits are sold through e-commerce platforms each year.

OBJECTIVE

This study aims to analyze the spatiotemporal distribution of the HIVST kit-purchasing population (HIVSTKPP) in China.

METHODS

Deidentified transaction data were retrieved from a leading e-commerce platform in China. A joinpoint regression model was used to examine annual trends of the HIVSTKPP rates by calculating average annual percentage change. Bayesian spatiotemporal analysis was performed to locate hot spots with HIVSTKPP rates. Spatial autocorrelation analysis and space-time cluster analysis were conducted to identify clusters of HIVSTKPP. High-high clusters of HIVSTKPP can be identified by spatial autocorrelation analysis, and high-high clusters indicate that a region and its surrounding region jointly had a higher-than-average HIVSTKPP rate. Spatial regression analysis was used to elucidate the association between the number of HIV testing facilities, urbanization ratio (the proportion of urban population in the total population), and gross domestic product per capita and the HIVSTKPP.

RESULTS

Between January 1, 2016, and December 31, 2019, a total of 2.18 million anonymous persons in China placed 4.15 million orders and purchased 4.51 million HIVST kits on the web. In each of these 4 years, the observed monthly size of the HIVSTKPP peaked in December, the month of World AIDS Day. HIVSTKPP rates per 100,000 population significantly increased from 20.62 in 2016 to 64.82 in 2019 (average annual percentage change=48.2%; P<.001). Hot spots were mainly located in municipalities, provincial capitals, and large cities, whereas high-high clusters and high-demand clusters were predominantly detected in cities along the southeast coast. We found positive correlations between a region's number of HIV testing facilities, urbanization ratio, and gross domestic product per capita and the HIVSTKPP.

CONCLUSIONS

Our study identified key areas with larger demand for HIVST kits for public health policy makers to reallocate resources and optimize the HIV care continuum. Further research combining spatiotemporal patterns of HIVST with HIV surveillance data is urgently needed to identify potential gaps in current HIV-monitoring practices.

Mapping Geographic Trends in Early Childhood Social, Emotional, and Behavioural Difficulties in Glasgow: 2010-2017

Measuring variation in childhood mental health supports the development of local early intervention strategies. The methodological approach used to investigate mental health trends (often determined by the availability of individual level data) can affect decision making. We apply two approaches to identify geographic trends in childhood social, emotional, and behavioural difficulties using the Strengths and Difficulties Questionnaire (SDQ). SDQ forms were analysed for 35,171 children aged 4-6 years old across 180 preschools in Glasgow, UK, between 2010 and 2017 as part of routine monitoring. The number of children in each electoral ward and year with a high SDQ total difficulties score (≥15), indicating a high risk of psychopathology, was modelled using a disease mapping model. The total difficulties score for an individual child nested in their preschool and electoral ward was modelled using a multilevel model. For each approach, linear time trends and unstructured spatial random effects were estimated. The disease mapping model estimated a yearly rise in the relative rate (RR) of high scores of 1.5-5.0%. The multilevel model estimated an RR increase of 0.3-1.2% in average total scores across the years, with higher variation between preschools than between electoral wards. Rising temporal trends may indicate worsening social, emotional, and behavioural difficulties over time, with a faster rate for the proportion with high scores than for the average total scores. Preschool and ward variation, although minimal, highlight potential priority areas for local service provision. Both methodological approaches have utility in estimating and predicting children's difficulties and local areas requiring greater intervention.

Environmental Drivers of Monkeypox Transmission in the Democratic Republic of the Congo

Monkeypox (MPX) is an emergent severe zoonotic disease resembling that of smallpox. To date, most cases of human MPX have been reported in the Democratic Republic of the Congo (DRC). While the number of cases has increased steadily in the DRC over the last 30 years, the environmental risk factors that drive the spatiotemporal dynamics of MPX transmission remain poorly understood. This study aimed to investigate the spatiotemporal associations between environmental risk factors and annual MPX incidence in the DRC. All MPX cases reported weekly at the health zone level over a 16-year period (2000-2015) were analyzed. A Bayesian hierarchical generalized linear mixed model was conducted to identify the spatiotemporal associations between annual MPX incidence and three types of environmental risk factors illustrating environment as a system resulting from physical, social and cultural interactions Primary forest (IRR 1.034 [1.029-1.040]), economic well-being (IRR 1.038 [1.031-1.047]), and temperature (IRR 1.143 [1.028-1.261]) were positively associated with annual MPX incidence. Our study shows that physical environmental risk factors alone cannot explain the emergence of MPX outbreaks in the DRC. Economic level and cultural practices participate from environment as a whole and thus, must be considered to understand exposure to MPX risk Future studies should examine the impact of these factors in greater detail.

Forecasting of overall and aggressive prostate cancer incident counts at the small area level

OBJECTIVES

This study aims to forecast overall and aggressive prostate cancer counts at the local government area (LGA) level over 10 years (2019-2028) in Victoria, Australia, using Victorian Cancer Registry (2001-2018) data.

METHODS

We used the Age-Period-Cohort approach to estimate the annual age-specific incidence and used Bayesian spatiotemporal models that account for non-linear temporal trends and area-level risk factors. We evaluated the models' performance by withholding and comparing forecasts with the 2014-2018 data.

RESULTS

There were 80,449 prostate cancer cases between 2001 and 2018, with an overall increasing trend. Compared to 2001, prostate cancer incidence increased by 69%, from 3049 to 5167 cases in 2018. Prostate cancer counts are expected to reach 7631 cases in 2028, a further 48% increase. Unexplained area-level spatial variation was substantially reduced after adjusting for the area-level elderly population. Aggressive prostate cancer cases increased by 107% between 2001 and 2018 and are expected to rise by 123% increase in 2028. The proportion of aggressive prostate cancer cases will increase to 31% in 2028 from 20% in 2018. By 2028, overall and aggressive prostate cancer cases are projected to be increasing in 66% and 61% of LGAs.

CONCLUSION

Prostate cancer cases are projected to rise at the state level and most LGAs in the next 10 years, with much steeper increases in aggressive cases. Population growth and an ageing population have primarily contributed to this rise besides prostate-specific antigen testing. These prediction estimates help inform prostate cancer burden and facilitate efficient healthcare delivery.

Trends and correlates of spatially aggregated alcohol-involved crashes among Whites and Hispanics in California

AIMS

This paper examines trends and correlates of alcohol-involved motor vehicle crashes (AMVCs) in California between 2005 and 2016 among Hispanic and non-Hispanic Whites (Whites hereafter). Together these two groups comprise 76% of the state population. The paper also examines whether alcohol outlet density, percentage of Hispanics in census tract populations, and distance to the U.S./Mexico border are related to greater risks for AMVCs. The border is of interest given the greater availability of alcohol in the area.

METHODS

Crash data come from Statewide Integrated Traffic Records System maintained by the California Highway Patrol. Sociodemographic and community characteristics data from the U.S. Census and alcohol outlet density were aggregated to census tracts. Total motor vehicle crashes and AMVCs were related to these characteristics using hierarchical Bayesian Poisson space-time models.

RESULTS

There were over two million injury and fatality crashes during the period of analysis, of which 11% were AMVCs. About 1.7% of these crashes had fatalities. The rate of AMVCs increased among both Whites and Hispanics until 2008. After 2008, the rate among Whites declined through 2016 while the rate among Hispanics declined for 2 years (2009 and 2010) and increased thereafter. Crash distance from the border (RR = 1.016, 95% CI = 1.010 to 1.022) and percent Hispanic population (RR = 1.006; 95% CI = 1.003 to 1.009) were well-supported results with 95% credible intervals that did not include 1. The percentages of the following: bars/pubs, males, individuals aged 18 to 29 and 40 to 49 years, U.S. born population, individuals below the 150% poverty level, unemployed, housing vacant, and housing owner-occupied were all positively associated with AMVCs and well supported.

CONCLUSIONS

Between 2005 and 2016 the rate of AMVCs in California declined among Whites but not among Hispanics. Population-level indicators of percent Hispanic population, distance to the U.S. Mexico border, gender, age distribution, and socioeconomic stability were positively associated with crash rates, indicating that important contextual characteristics help determine the level of AMVC rates in communities.

Where Is Air Quality Improving, and Who Benefits? A Study of PM2.5 and Ozone Over 15 Years

In the United States, concentrations of criteria air pollutants have declined in recent decades. Questions remain regarding whether improvements in air quality are equitably distributed across subpopulations. We assessed spatial variability and temporal trends in concentrations of particulate matter with an aerodynamic diameter ≤2.5 μm (PM2.5) and ozone (O3) across North Carolina from 2002-2016, and associations with community characteristics. Estimated daily PM2.5 and O3 concentrations at 2010 Census tracts were obtained from the Fused Air Quality Surface Using Downscaling archive and averaged to create tract-level annual PM2.5 and O3 estimates. We calculated tract-level measures of: racial isolation of non-Hispanic Black individuals, educational isolation of non-college educated individuals, the neighborhood deprivation index (NDI), and percentage of the population in urban areas. We fitted hierarchical Bayesian space-time models to estimate baseline concentrations of and time trends in PM2.5 and O3 for each tract, accounting for spatial between-tract correlation. Concentrations of PM2.5 and O3 declined by 6.4 μg/m3 and 13.5 ppb, respectively. Tracts with lower educational isolation and higher urbanicity had higher PM2.5 and more pronounced declines in PM2.5. Racial isolation was associated with higher PM2.5 but not with the rate of decline in PM2.5. Despite declines in pollutant concentrations, over time, disparities in exposure increased for racially and educationally isolated communities.

Empirical dynamic modeling of the association between ambient PM2.5 and under-five mortality across 2851 counties in Mainland China, 1999-2012

BACKGROUND

Ambient fine particulate matter (PM_2.5) pollution has been associated with mortality from various diseases, however, its association with under-five mortality rate (U5MR) has remained largely unknown.

METHODS

Based on the U5MR data across 2851 counties in Mainland China from 1999 to 2012, we employed approximate Bayesian latent Gaussian models to assess the association between ambient PM_2.5 and U5MR at the county level for the whole nation and sub-regions. GDP growth rate, normalized difference vegetation index (NDVI), temperature, and night-time light were included as covariates using a smoothing function. We further implemented an empirical dynamic model (EDM) to explore the potential causal relationship between PM_2.5 and U5MR.

RESULTS

We observed a declining trend in U5MR in most counties throughout the study period. Spatial heterogeneity in U5MR was observed. Nationwide analysis suggested that each 10 µg/m³ increase in annual concentration of PM_2.5 was associated with an increase of 1.2 (95% CI: 1.0 - 1.3) per 1000 live births in U5MR. Regional analyses showed that the strongest positive association was located in the Northeastern part of China [1.8 (95% CI: 1.4 - 2.1)]. The EDM showed a significant causal association between PM_2.5 and U5MR, with an embedding dimension of 5 and 7, and nonlinear values θ of 4 and 6, respectively.

CONCLUSION

China exhibited a downward trend in U5MR from 1999 to 2012, with spatial heterogeneity observed across the country. Our analysis reveals a positive association between PM_2.5 and U5MR, which may support a causal relationship.

Environmental risk factors of airborne viral transmission: Humidity, Influenza and SARS-CoV-2 in the Netherlands

OBJECTIVE

The relationship between specific humidity and influenza/SARS-CoV-2 in the Netherlands is evaluated over time and at regional level.

DESIGN

Parametric and non-parametric correlation coefficients are calculated to quantify the relationship between humidity and influenza, using five years of weekly data. Bayesian spatio-temporal models-with a Poisson and a Gaussian likelihood-are estimated to find the relationship between regional humidity and the daily cases of SARS-CoV-2 in the municipalities and provinces of the Netherlands.

RESULTS

An inverse (negative) relationship is observed between specific humidity and the incidence of influenza between 2015 and 2019. The space-time analysis indicates that an increase of specific humidity of one gram of water vapor per kilogram of air (1 g/kg) is related to a reduction of approximately 5% in the risk of COVID-19 infections.

CONCLUSIONS

The increase in humidity during the outbreak of the SARS-CoV-2 in the Netherlands may have helped to reduce the risk of regional COVID-19 infections. Policies that lead to an increase in household specific humidity to over 6g/Kg will help reduce the spread of respiratory viruses such as influenza and SARS-CoV-2.

Application of Bayesian Space-Time interaction models for Deer-Vehicle crash hotspot identification

The objective of this research was to identify and prioritize deer-vehicle crash (DVC) hotspots using five years of crash data. This study applied Bayesian spatiotemporal models for the identification of the DVC hotspots. The Bayesian spatiotemporal model allows to observe area-specific trends in the DVC data and highlights specific locations where DVC occurrence is deteriorating or improving over time. Census Tracts (CTs) were used as the geographic units to aggregate DVC, land use, and transportation infrastructure related data of Minnesota (MN) for the year 2015 to 2019. Several tests were conducted to evaluate the performance of the hotspot identification methods. The result showed that Type-I spatiotemporal interaction model (Model-2) outperforms other four space-time models in terms of predicting DVC frequency in CTs and hotspot identification performance test measures. Results showed that forest area, vegetation, and wetland percentages were positively associated with DVC frequency, whereas the percentage of developed land use was negatively associated with DVC frequency. The findings of this study suggest that the deer population plays an important role in DVCs, which indicates that deer population management is necessary to minimize the DVC risks. Using the final Type-I spatiotemporal interaction model, 65 "High-High" CTs were identified, where both the posterior mean of the decision parameter (potential for safety improvement) and the area-specific trend were higher. The distribution of the identified hotspots showed that the risk of DVCs was more in suburban areas with mixed land use conditions. These CTs represent high-risk zones, which need immediate safety improvement measures to reduce the DVC risks. As DVC can occur at any roadway segment location, DVC hotspots information is important for safety engineers and policymakers to implement area specific countermeasures.

Variance partitioning in spatio-temporal disease mapping models

Bayesian disease mapping, yet if undeniably useful to describe variation in risk over time and space, comes with the hurdle of prior elicitation on hard-to-interpret random effect precision parameters. We introduce a reparametrized version of the popular spatio-temporal interaction models, based on Kronecker product intrinsic Gaussian Markov random fields, that we name the variance partitioning model. The variance partitioning model includes a mixing parameter that balances the contribution of the main and interaction effects to the total (generalized) variance and enhances interpretability. The use of a penalized complexity prior on the mixing parameter aids in coding prior information in an intuitive way. We illustrate the advantages of the variance partitioning model using two case studies.

Spatiotemporal Trends and Distributions of Malaria Incidence in the Northwest Ethiopia

Understanding and extracting noticeable patterns of malaria surveillance data at the district level are crucial for malaria prevention, control, and elimination progress. This study aimed to analyze spatiotemporal trends and nonparametric dynamics of malaria incidences in northwest Ethiopia, considering spatial and temporal correlations. The data were analyzed using count regression spatiotemporal models under the Bayesian setups, and parameters were estimated using integrated nested Laplace approximations (INLA). The region had a declining linear trend, and the average annual malaria incidence rate was 24.8 per 1,000 persons between 2012 and 2020. The malaria incidence rate was decreased by 0.984 (95% CI: 0.983, 0.986) per unit increase in months between July 2012 and June 2020. Districts found in the western and northwestern parts of the region had a steeper trend, while districts in the eastern and southern parts had a less steep trend than the average trend of the region. Compared to the regional level trend, the decreasing rate of malaria incidence trends was lower in most town administrations. The nonparametric dynamics showed that the monthly malaria incidence had a sinusoidal wave shape that varied throughout study periods. Malaria incidence had a decreasing linear trend changed across districts of the study region, and the steepness of trends of districts might not depend on incidences. Thus, an intervention and controlling mechanism that considers malaria incidences and district-specific differential trends would be indispensable to mitigate malaria transmission in the region.

Spatio-temporal disease risk estimation using clustering-based adjacency modelling

Conditional autoregressive models are typically used to capture the spatial autocorrelation present in areal unit disease count data when estimating the spatial pattern in disease risk. This correlation is represented by a binary neighbourhood matrix based on a border sharing specification, which enforces spatial correlation between geographically neighbouring areas. However, enforcing such correlation will mask any discontinuities in the disease risk surface, thus impeding the detection of clusters of areas that exhibit higher or lower risks compared to their neighbours. Here we propose novel methodology to account for these clusters and discontinuities in disease risk via a two-stage modelling approach, which either forces the clusters/discontinuities to be the same for all time periods or allows them to evolve dynamically over time. Stage one constructs a set of candidate neighbourhood matrices to represent a range of possible cluster/discontinuity structures in the data, and stage two estimates an appropriate structure(s) by treating the neighbourhood matrix as an additional parameter to estimate within a Bayesian spatio-temporal disease mapping model. The effectiveness of our novel methodology is evidenced by simulation, before being applied to a new study of respiratory disease risk in Greater Glasgow, Scotland from 2011 to 2017.

Geographical inequalities in the decreasing 28-day mortality following incident acute myocardial infarction: a Danish register-based cohort study, 1987-2016

Background

Mortality following acute myocardial infarction (AMI) has decreased in western countries for decades; however, it remains unknown whether the decrease is distributed equally across the population independently of residential location. This study investigated whether the observed decreasing 28-day mortality following an incident AMI in Denmark from 1987 to 2016 varied geographically at municipality level after accounting for sociodemographic characteristics.

Methods

A register-based cohort study design was used to investigate 28-day mortality among individuals with an incident AMI. Global spatial autocorrelation (within sub-periods) was analysed at municipality level using Moran's I. Analysis of spatio-temporal autocorrelation before and after adjusting for sociodemographic characteristics was performed using logistic regression and conditional autoregressive models with inference in a Bayesian setting.

Results

In total, 368,839 individuals with incident AMI were registered between 1987 and 2016 in Denmark; 128,957 incident AMIs were fatal. The 28-day mortality decreased over time at national level with an odds ratio of 0.788 (95% credible interval (0.784, 0.792)) per 5-year period after adjusting for sociodemographic characteristics. The decrease in the 28-day mortality was geographically unequally distributed across the country and in a geographical region in northern Jutland, the 28-day mortality decreased significantly slower (4–12%) than at national level.

Conclusions

During the period from 1987 to 2016, the 28-day mortality following an incident AMI decreased substantially in Denmark. However, in a local geographical region, the 28-day mortality decreased significantly slower than in the rest of the country both before and after adjusting for sociodemographic differences. Efforts should be made to keep geographical trend inequalities in the 28-day mortality to a minimum.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12872-022-02519-7.

Space-time interactions in Bayesian disease mapping with recent tools: Making things easier for practitioners

Spatio-temporal disease mapping studies the distribution of mortality or incidence risks in space and its evolution in time, and it usually relies on fitting hierarchical Poisson mixed models. These models are complex for practitioners as they generally require adding constraints to correctly identify and interpret the different model terms. However, including constraints may not be straightforward in some recent software packages. This paper focuses on NIMBLE, a library of algorithms that contains among others a configurable system for Markov chain Monte Carlo (MCMC) algorithms. In particular, we show how to fit different spatio-temporal disease mapping models with NIMBLE making emphasis on how to include sum-to-zero constraints to solve identifiability issues when including spatio-temporal interactions. Breast cancer mortality data in Spain during the period 1990-2010 is used for illustration purposes. A simulation study is also conducted to compare NIMBLE with R-INLA in terms of parameter estimates and relative risk estimation. The results are very similar but differences are observed in terms of computing time.

Spatio-temporal modeling for malnutrition in tribal population among states of India a Bayesian approach

Exploring Bayesian spatio-temporal methods to analyze spatial dependence in malnutrition at the state level for tribal children (less than 3 years) population of India and change over time (three rounds of NFHS-2(1998-99),3(2005-06) and 4(2015-16)). The Bayesian model, fitted by Markov chain Monte Carlo simulation using OpenBUGS, for spatial autocorrelation (through spatial random effects modeling). The model estimated (1) mean time trend and (2) spatial random effects. Results of spatio-temporal modeling for stunting, wasting and underweight exhibited a declining mean trend across the study region from NFHS-2 to NFHS-4. Spatial random effects exhibited spatial dependence for various states in stunting, wasting and underweight tribal children. Future research should analyze spatio-temporal distribution for malnutrition at district level which will require NFHS-5 data. Also, analysis can be done capturing spatio-temporal interaction and identifying hot spots and cold spots at district level.

Spatial Pattern and Spillover of Abatement Effect of Chinese Environmental Protection Tax Law on PM2.5 Pollution

Particulate matter (PM2.5) pollution is a threat to public health, and environmental taxation is an important regulatory mode controlling PM2.5 pollution. In 2018, China implemented the Environmental Protection Tax Law (EPTL) targeting PM2.5 pollution. Based on in-situ monitoring and emission inventory data, a Bayesian hierarchical spatiotemporal model combining a two-period trends difference method was employed to measure the abatement effects of China’s EPTL on PM2.5 pollution (AEEPTLPM). On this basis, a spatial spillover index (SSI) of the AEEPTLPM is proposed. Applying this index, we calculated the spatial spillover characteristics of the AEEPTLPM in mainland China at a provincial scale in 2018–2019. The results show that the EPTL has had significant abatement effects on both in-situ-monitored PM2.5 concentrations and local total industrial PM2.5 emissions. Additionally, the two types of AEEPTLPM display distinct spatial heterogeneity. A correlation between the AEEPTLPM and the degree of PM2.5 pollution was observed; areas with serious PM2.5 pollution have higher AEEPTLPM levels, and vice versa. The SSI indicates that the AEEPTLPM exhibits significant spatial spillover characteristics, and spatial heterogeneity is also present.

Spatiotemporal high-resolution prediction and mapping: methodology and application to dengue disease

Dengue disease has become a major public health problem. Accurate and precise identification, prediction and mapping of high-risk areas are crucial elements of an effective and efficient early warning system in countering the spread of dengue disease. In this paper, we present the fusion area-cell spatiotemporal generalized geoadditive-Gaussian Markov random field (FGG-GMRF) framework for joint estimation of an area-cell model, involving temporally varying coefficients, spatially and temporally structured and unstructured random effects, and spatiotemporal interaction of the random effects. The spatiotemporal Gaussian field is applied to determine the unobserved relative risk at cell level. It is transformed to a Gaussian Markov random field using the finite element method and the linear stochastic partial differential equation approach to solve the "big n" problem. Sub-area relative risk estimates are obtained as block averages of the cell outcomes within each sub-area boundary. The FGG-GMRF model is estimated by applying Bayesian Integrated Nested Laplace Approximation. In the application to Bandung city, Indonesia, we combine low-resolution area level (district) spatiotemporal data on population at risk and incidence and high-resolution cell level data on weather variables to obtain predictions of relative risk at subdistrict level. The predicted dengue relative risk at subdistrict level suggests significant fine-scale heterogeneities which are not apparent when examining the area level. The relative risk varies considerably across subdistricts and time, with the latter showing an increase in the period January-July and a decrease in the period August-December.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s10109-021-00368-0.

Spatiotemporal trends in life expectancy and impacts of economic growth and air pollution in 134 countries: A Bayesian modeling study

Life expectancy (LE) varies across countries in space and time, and economic growth and air pollution are two important influence factors to LE. This study mainly aims to investigate spatiotemporal trends in LE in 134 countries from 1960 to 2016 by using Bayesian spatiotemporal modeling. Further, the relations between per capita gross domestic product (GDP_pc) and population-weighted fine particulate matter (pwPM_2.5) and LE are investigated from a global perspective from 1998 to 2016 by using the Bayesian regression model. The results illustrated the heterogeneity of spatiotemporal trends in LE globally. Specifically, Africa and South-East Asia show much lower LE levels, and the Americas, European, and Western Pacific exhibit a relatively higher LE level compared to the overall level. The countries with low overall levels of LE show a relatively stronger upward trend than the overall upward trend and vice versa. In addition, this study demonstrates that the spatial differences in effects of influence factors on LE in the six WHO regions in the 134 countries. Africa shows the highest positive regression coefficient of GDP_pc and lowest negative regression coefficient of pwPM_2.5 on LE than other regions in the world. Furthermore, it shows the complexity of the interaction between economic growth and air pollution on LE across six WHO regions. Our findings suggest the public policies to reduce the health damage caused by air pollution, especially in Africa, Eastern Mediterranean, and Europe where the pwPM_2.5 negatively affect the LE benefits from economic growth.

Spatiotemporal Analysis of the Association Between Pain Management Clinic Laws and Opioid Prescribing and Overdose Deaths

Pain management clinic (PMC) laws were enacted by 12 states to promote appropriate opioid prescribing, but their impact is inadequately understood. We analyzed county-level opioid overdose deaths (National Vital Statistics System) and patients filling long-duration (≥30 day) or high-dose (≥90 morphine milligram equivalents per day) opioid prescriptions (IQVIA, Inc.) in the United States in 2010-2018. We fitted Besag-York-Mollié spatiotemporal models to estimate annual relative rates (RRs) of overdose and prevalence ratios (PRs) of high-risk prescribing associated with any PMC law and 3 provisions: payment restrictions, site inspections, and criminal penalties. Laws with criminal penalties were significantly associated with reduced PRs of long-duration and high-dose opioid prescriptions (adjusted PR = 0.82, 95% credible interval (CrI): 0.82, 0.82, and adjusted PR = 0.73, 95% CI: 0.73, 0.74 respectively) and reduced RRs of total and natural/semisynthetic opioid overdoses (adjusted RR = 0.86, 95% CrI: 0.80, 0.92, and adjusted RR = 0.84, and 95% CrI: 0.77, 0.92, respectively). Conversely, PMC laws were associated with increased relative rates of synthetic opioid and heroin overdose deaths, especially criminal penalties (adjusted RR = 1.83, 95% CrI: 1.59, 2.11, and adjusted RR = 2.59, 95% CrI: 2.22, 3.02, respectively). Findings suggest that laws with criminal penalties were associated with intended reductions in high-risk opioid prescribing and some opioid overdoses but raise concerns regarding unintended consequences on heroin/synthetic overdoses.

Spatio-temporal variation in tuberculosis incidence and risk factors for the disease in a region of unbalanced socio-economic development

BACKGROUND

Previous research pointed to a close relationship between the incidence of tuberculosis (TB) in aging populations and socio-economic conditions, however there has been lack of studies focused on a region of unbalanced socio-economic development. The aim of this paper is to explore the spatio-temporal variation in TB incidence and examine risk determinants of the disease among aging populations in a typical region.

METHODS

Data on TB-registered cases between 2009 and 2014, in addition to social-economic factors, were collected for each district/county in Beijing, Tianjin and Hebei, a region characterized by an aging population and disparities in social-economic development. A Bayesian space-time hierarchy model (BSTHM) was used to reveal spatio-temporal variation in the incidence of TB among the elderly in this region between 2009 to 2014. GeoDetector was applied to measure the determinant power (q statistic) of risk factors for TB among the elderly.

RESULTS

The incidence of TB among the elderly exhibited geographical spatial heterogeneity, with a higher incidence in underdeveloped rural areas compared with that in urban areas. Hotspots of TB incidence risk among the elderly were mostly located in north-eastern and southern areas in the study region, far from metropolitan areas. Areas with low risk were distributed mainly in the Beijing-Tianjin metropolitan areas. Social-economic factors had a non-linear influence on elderly TB incidence, with the dominant factors among rural populations being income (q = 0.20) and medical conditions (q = 0.17). These factors had a non-linear interactive effect on the incidence of TB among the elderly, with medical conditions and the level of economic development having the strongest effect (q = 0.54).

CONCLUSIONS

The findings explain spatio-temporal variation in TB incidence and risk determinants of elderly TB in the presence of disparities in social-economic development. High-risk zones were located mainly in rural areas, far from metropolitan centres. Medical conditions and the economic development level were significantly associated with elderly TB incidence, and these factors had a non-linear interactive effect on elderly TB incidence. The findings can help to optimize the allocation of health resources and to control TB transmission in the aging population in this region.

Spatio-Temporal Spread Pattern of COVID-19 in Italy

This paper investigates the spatio-temporal spread pattern of COVID-19 in Italy, during the first wave of infections, from February to October 2020. Disease mappings of the virus infections by using the Besag–York–Mollié model and some spatio-temporal extensions are provided. This modeling framework, which includes a temporal component, allows the studying of the time evolution of the spread pattern among the 107 Italian provinces. The focus is on the effect of citizens’ mobility patterns, represented here by the three distinct phases of the Italian virus first wave, identified by the Italian government, also characterized by the lockdown period. Results show the effectiveness of the lockdown action and an inhomogeneous spatial trend that characterizes the virus spread during the first wave. Furthermore, the results suggest that the temporal evolution of each province’s cases is independent of the temporal evolution of the other ones, meaning that the contagions and temporal trend may be caused by some province-specific aspects rather than by the subjects’ spatial movements.

Multiple sclerosis incidence rate in southern Iran: a Bayesian epidemiological study

Background

Multiple Sclerosis (MS) remains to be a public health challenge, due to its unknown biological mechanisms and clinical impacts on young people. The prevalence of this disease in Iran is reported to be 5.30 to 74.28 per 100,000-person. Because of high prevalence of this disease in Fars province, the purpose of this study was to assess the spatial pattern of MS incidence rate by modeling both the associations s of spatial dependence between neighboring regions and risk factors in a Bayesian Poisson model, which can lead to the improvement of health resource allocation decisions.

Method

Data from 5468 patients diagnosed with MS were collected, according to the McDonald’s criteria. New cases of MS were reported by the MS Society of Fars province from 1991 until 2016. The association between the percentage of people with low vitamin D intake, smoking, abnormal BMI and alcohol consumption in addition to spatial structure in a Bayesian spatio-temporal hierarchical model were used to determine the relative risk and trend of MS incidence rate in 29 counties of Fars province.

Results

County-level crude incidence rates ranged from 0.22 to 11.31 cases per 100,000-person population. The highest relative risk was estimated at 1.80 in the county of Shiraz, the capital of Fars province, while the lowest relative risk was estimated at 0.11 in Zarindasht county in southern of Fars. The percentages of vitamin D supplementation intake and smoking were significantly associated with the incidence rate of MS. The results showed that 1% increase in vitamin D supplementation intake is associated with 2% decrease in the risk of MS and 1% increase in smoking is associated with 16% increase in the risk of MS.

Conclusion

Bayesian spatio-temporal analysis of MS incidence rate revealed that the trend in the south and south east of Fars province is less steep than the mean trend of this disease. The lower incidence rate was associated with a higher percentage of vitamin D supplementation intake and a lower percentage of smoking. Previous studies have also shown that smoking and low vitamin D, among all covariates or risk factors, might be associated with high incidence of MS.

Spatiotemporal Associations Between Social Vulnerability, Environmental Measurements, and COVID-19 in the Conterminous United States

This study summarizes the results from fitting a Bayesian hierarchical spatiotemporal model to coronavirus disease 2019 (COVID-19) cases and deaths at the county level in the United States for the year 2020. Two models were created, one for cases and one for deaths, utilizing a scaled Besag, York, Mollié model with Type I spatial-temporal interaction. Each model accounts for 16 social vulnerability and 7 environmental variables as fixed effects. The spatial pattern between COVID-19 cases and deaths is significantly different in many ways. The spatiotemporal trend of the pandemic in the United States illustrates a shift out of many of the major metropolitan areas into the United States Southeast and Southwest during the summer months and into the upper Midwest beginning in autumn. Analysis of the major social vulnerability predictors of COVID-19 infection and death found that counties with higher percentages of those not having a high school diploma, having non-White status and being Age 65 and over to be significant. Among the environmental variables, above ground level temperature had the strongest effect on relative risk to both cases and deaths. Hot and cold spots, areas of statistically significant high and low COVID-19 cases and deaths respectively, derived from the convolutional spatial effect show that areas with a high probability of above average relative risk have significantly higher Social Vulnerability Index composite scores. The same analysis utilizing the spatiotemporal interaction term exemplifies a more complex relationship between social vulnerability, environmental measurements, COVID-19 cases, and COVID-19 deaths.

Chronic high risk of intimate partner violence against women in disadvantaged neighborhoods: An eight-year space-time analysis

We conducted a small-area ecological longitudinal study to analyze neighborhood contextual influences on the spatio-temporal variations in intimate partner violence against women (IPVAW) risk in a southern European city over an eight-year period. We used geocoded data of IPVAW cases with associated protection orders (n = 5867) in the city of Valencia, Spain (2011-2018). The city's 552 census block groups were used as the neighborhood units. Neighborhood-level covariates were: income, education, immigrant concentration, residential instability, alcohol outlet density, and criminality. We used a Bayesian autoregressive approach to spatio-temporal disease mapping. Neighborhoods with low levels of income and education and high levels of residential mobility and criminality had higher relative risk of IPVAW. Spatial patterns of high risk of IPVAW persisted over time during the eight-year period analyzed. Areas of stable low risk and with increasing or decreasing risk were also identified. Our findings link neighborhood disadvantage to the existence and persistence over time of spatial inequalities in IPVAW risk, showing that high risk of IPVAW can become chronic in disadvantaged neighborhoods. Our analytic approach provides specific risk estimates at the small-area level that are informative for intervention purposes, and can be useful to assess the effectiveness of prevention efforts in reducing IPVAW.

A Review of Spatiotemporal Models for Count Data in R Packages. A Case Study of COVID-19 Data

Spatiotemporal models for count data are required in a wide range of scientific fields, and they have become particularly crucial today because of their ability to analyze COVID-19-related data. The main objective of this paper is to present a review describing the most important approaches, and we monitor their performance under the same dataset. For this review, we focus on the three R-packages that can be used for this purpose, and the different models assessed are representative of the two most widespread methodologies used to analyze spatiotemporal count data: the classical approach and the Bayesian point of view. A COVID-19-related case study is analyzed as an illustration of these different methodologies. Because of the current urgent need for monitoring and predicting data in the COVID-19 pandemic, this case study is, in itself, of particular importance and can be considered the secondary objective of this work. Satisfactory and promising results have been obtained in this second goal. With respect to the main objective, it has been seen that, although the three models provide similar results in our case study, their different properties and flexibility allow us to choose the model depending on the application at hand.

Good Samaritan laws and overdose mortality in the United States in the fentanyl era

BACKGROUND

As of July 2018, 45 United States (US) states and the District of Columbia have enacted an overdose Good Samaritan law (GSL). These laws, which provide limited criminal immunity to individuals who request assistance during an overdose, may be of importance in the current wave of the overdose epidemic, which is driven primarily by illicit opioids including heroin and fentanyl. There are substantial differences in the structures of states' GSL laws which may impact their effectiveness. This study compared GSLs which have legal provisions protecting from arrest and laws which have more limited protections.

METHODS

Using national county-level overdose mortality data from 3109 US counties, we examined the association of enactment of GSLs with protection from arrest and GSLs with more limited protections with subsequent overdose mortality between 2013 and 2018. Since GSLs are often enacted in conjunction with Naloxone Access Laws (NAL), we examined the effect of GSLs separately and in conjunction with NAL. We conducted these analyses using hierarchical Bayesian spatiotemporal Poisson models.

RESULTS

GSLs with protections against arrest enactment in conjunction with a NAL were associated with 7% lower rates of all overdose deaths (rate ratio (RR): 0.93% Credible Interval (CI): 0.89-0.97), 10% lower rates in opioid overdose deaths (RR: 0.90; CI: 0.85-0.95) and 11% lower rates of heroin/synthetic overdose mortality (RR: 0.89; CI: 0.82-0.96) two years after enactment, compared to rates in states without these laws. Significant reductions in overdose mortality were not seen for GSLs with protections for charge or prosecution.

CONCLUSION

GSLs with more expansive legal protections combined with a NAL, were associated with lower rates of overdose deaths, although these risk reductions take time to manifest. Policy makers should consider enacting and implementing more expansive GSLs with arrest protections to increase the likelihood people will contact emergency services in the event of an overdose.

The Spatial Overlap of Police Calls Reporting Street-Level and Behind-Closed-Doors Crime: A Bayesian Modeling Approach

Traditionally, intimate-partner violence has been considered a special type of crime that occurs behind closed doors, with different characteristics from street-level crime. The aim of this study is to analyze the spatial overlap of police calls reporting street-level and behind-closed-doors crime. We analyzed geocoded police calls in the 552 census-block groups of the city of Valencia, Spain, related to street-level crime (N = 26,624) and to intimate-partner violence against women (N = 11,673). A Bayesian joint model was run to analyze the spatial overlap. In addition, two Bayesian hierarchical models controlled for different neighborhood characteristics to analyze the relative risks. Results showed that 66.5% of the total between-area variation in risk of reporting street-level crime was captured by a shared spatial component, while for reporting IPVAW the shared component was 91.1%. The log relative risks showed a correlation of 0.53, with 73.6% of the census-block groups having either low or high values in both outcomes, and 26.4% of the areas with mismatched risks. Maps of the shared component and the relative risks are shown to detect spatial differences. These results suggest that although there are some spatial differences between police calls reporting street-level and behind-closed-doors crime, there is also a shared distribution that should be considered to inform better-targeted police interventions.

Forecasting of Lung Cancer Incident Cases at the Small-Area Level in Victoria, Australia

Predicting lung cancer cases at the small-area level is helpful to quantify the lung cancer burden for health planning purposes at the local geographic level. Using Victorian Cancer Registry (2001-2018) data, this study aims to forecast lung cancer counts at the local government area (LGA) level over the next ten years (2019-2028) in Victoria, Australia. We used the Age-Period-Cohort approach to estimate the annual age-specific incidence and utilised Bayesian spatio-temporal models that account for non-linear temporal trends and area-level risk factors. Compared to 2001, lung cancer incidence increased by 28.82% from 1353 to 1743 cases for men and 78.79% from 759 to 1357 cases for women in 2018. Lung cancer counts are expected to reach 2515 cases for men and 1909 cases for women in 2028, with a corresponding 44% and 41% increase. The majority of LGAs are projected to have an increasing trend for both men and women by 2028. Unexplained area-level spatial variation substantially reduced after adjusting for the elderly population in the model. Male and female lung cancer cases are projected to rise at the state level and in each LGA in the next ten years. Population growth and an ageing population largely contributed to this rise.