Evaluating the impact of a small number of areas on spatial estimation

A Systematic Review of Areal Units and Adjacency Used in Bayesian Spatial and Spatio-Temporal Conditional Autoregressive Models in Health Research

Advancements in Bayesian spatial and spatio-temporal modelling have been observed in recent years. Despite this, there are unresolved issues about the choice of appropriate spatial unit and adjacency matrix in disease mapping. There is limited systematic review evidence on this topic. This review aimed to address these problems. We searched seven databases to find published articles on this topic. A modified quality assessment tool was used to assess the quality of studies. A total of 52 studies were included, of which 26 (50.0%) were on infectious diseases, 10 (19.2%) on chronic diseases, 8 (15.5%) on maternal and child health, and 8 (15.5%) on other health-related outcomes. Only 6 studies reported the reasons for using the specified spatial unit, 8 (15.3%) studies conducted sensitivity analysis for prior selection, and 39 (75%) of the studies used Queen contiguity adjacency. This review highlights existing variation and limitations in the specification of Bayesian spatial and spatio-temporal models used in health research. We found that majority of the studies failed to report the rationale for the choice of spatial units, perform sensitivity analyses on the priors, or evaluate the choice of neighbourhood adjacency, all of which can potentially affect findings in their studies.

The Potential of Surveillance Data for Dengue Risk Mapping: An Evaluation of Different Approaches in Cuba

To better guide dengue prevention and control efforts, the use of routinely collected data to develop risk maps is proposed. For this purpose, dengue experts identified indicators representative of entomological, epidemiological and demographic risks, hereafter called components, by using surveillance data aggregated at the level of Consejos Populares (CPs) in two municipalities of Cuba (Santiago de Cuba and Cienfuegos) in the period of 2010-2015. Two vulnerability models (one with equally weighted components and one with data-derived weights using Principal Component Analysis), and three incidence-based risk models were built to construct risk maps. The correlation between the two vulnerability models was high (tau > 0.89). The single-component and multicomponent incidence-based models were also highly correlated (tau ≥ 0.9). However, the agreement between the vulnerability- and the incidence-based risk maps was below 0.6 in the setting with a prolonged history of dengue transmission. This may suggest that an incidence-based approach does not fully reflect the complexity of vulnerability for future transmission. The small difference between single- and multicomponent incidence maps indicates that in a setting with a narrow availability of data, simpler models can be used. Nevertheless, the generalized linear mixed multicomponent model provides information of covariate-adjusted and spatially smoothed relative risks of disease transmission, which can be important for the prospective evaluation of an intervention strategy. In conclusion, caution is needed when interpreting risk maps, as the results vary depending on the importance given to the components involved in disease transmission. The multicomponent vulnerability mapping needs to be prospectively validated based on an intervention trial targeting high-risk areas.

Spatio-Temporal Pattern and Meteo-Climatic Determinants of Visceral Leishmaniasis in Italy

Historically, visceral leishmaniasis (VL) in Italy was constrained to Mediterranean areas. However, in the last 20 years, sand fly vectors and human cases of VL have been detected in northern Italy, traditionally classified as a cold area unsuitable for sand fly survival. We aim to study the spatio-temporal pattern and climatic determinants of VL incidence in Italy. National Hospital Discharge Register records were used to identify incident cases of VL between 2009 and 2016. Incident rates were computed for each year (N = 8) and for each province (N = 110). Data on mean temperature and cumulative precipitation were obtained from the ERA5-Land re-analysis. Age- and sex-standardized incidence rates were modeled with Bayesian spatial and spatio-temporal conditional autoregressive Poisson models in relation to the meteo-climatic parameters. Statistical inference was based on Monte Carlo−Markov chains. We identified 1123 VL cases (incidence rate: 2.4 cases/1,000,000 person-years). The highest incidence rates were observed in southern Italy, even though some areas of northern Italy experienced high incidence rates. Overall, in the spatial analysis, VL incidence rates were positively associated with average air temperatures (β for 1 °C increase in average mean average temperature: 0.14; 95% credible intervals (CrI): 0.01, 0.27) and inversely associated with average precipitation (β for 20 mm increase in average summer cumulative precipitation: −0.28, 95% CrI: −0.42, −0.13). In the spatio-temporal analysis, no association between VL cases and season-year specific temperature and precipitation anomalies was detected. Our findings indicate that VL is endemic in the whole Italian peninsula and that climatic factors, such as air temperature and precipitation, might play a relevant role in shaping the geographical distribution of VL cases. These results support that climate change might affect leishmaniasis distribution in the future.

Evaluation of spatial Bayesian Empirical Likelihood models in analysis of small area data

Bayesian empirical likelihood (BEL) models are becoming increasingly popular as an attractive alternative to fully parametric models. However, they have only recently been applied to spatial data analysis for small area estimation. This study considers the development of spatial BEL models using two popular conditional autoregressive (CAR) priors, namely BYM and Leroux priors. The performance of the proposed models is compared with their parametric counterparts and with existing spatial BEL models using independent Gaussian priors and generalised Moran basis priors. The models are applied to two benchmark spatial datasets, simulation study and COVID-19 data. The results indicate promising opportunities for these models to capture new insights into spatial data. Specifically, the spatial BEL models outperform the parametric spatial models when the underlying distributional assumptions of data appear to be violated.

Cross-sectional study to predict subnational levels of health workers' knowledge about severe malaria treatment in Kenya

OBJECTIVES

This study applied a Bayesian hierarchical ecological spatial model beyond predictor analysis to test for the best fitting spatial effects model to predict subnational levels of health workers' knowledge of severe malaria treatment policy, artesunate dosing, and preparation.

SETTING

County referral government and major faith-based hospitals across 47 counties in Kenya in 2019.

DESIGN AND PARTICIPANTS

A secondary analysis of cross-sectional survey data from 345 health workers across 89 hospitals with inpatient departments who were randomly selected and interviewed.

OUTCOME MEASURES

Three ordinal outcome variables for severe malaria treatment policy, artesunate dose and preparation were considered, while 12 individual and contextual predictors were included in the spatial models.

RESULTS

A third of the health workers had high knowledge levels on artesunate treatment policy; almost three-quarters had high knowledge levels on artesunate dosing and preparation. The likelihood of having high knowledge on severe malaria treatment policy was lower among nurses relative to clinicians (adjusted OR (aOR)=0.48, 95% CI 0.25 to 0.87), health workers older than 30 years were 61% less likely to have high knowledge about dosing compared with younger health workers (aOR=0.39, 95% CI 0.22 to 0.67), while health workers exposed to artesunate posters had 2.4-fold higher odds of higher knowledge about dosing compared with non-exposed health workers (aOR=2.38, 95% CI 1.22 to 4.74). The best model fitted with spatially structured random effects and spatial variations of the knowledge level across the 47 counties exhibited neighbourhood influence.

CONCLUSIONS

Knowledge of severe malaria treatment policies is not adequately and optimally available among health workers across Kenya. The factors associated with the health workers' level of knowledge were cadre, age and exposure to artesunate posters. The spatial maps provided subnational estimates of knowledge levels for focused interventions.

Bayesian spatiotemporal forecasting and mapping of COVID-19 risk with application to West Java Province, Indonesia

The coronavirus disease (COVID-19) has spread rapidly to multiple countries including Indonesia. Mapping its spatiotemporal pattern and forecasting (small area) outbreaks are crucial for containment and mitigation strategies. Hence, we introduce a parsimonious space-time model of new infections that yields accurate forecasts but only requires information regarding the number of incidences and population size per geographical unit and time period. Model parsimony is important because of limited knowledge regarding the causes of COVID-19 and the need for rapid action to control outbreaks. We outline the basics of Bayesian estimation, forecasting, and mapping, in particular for the identification of hotspots. The methodology is applied to county-level data of West Java Province, Indonesia.

Spatial connectivity in mosquito-borne disease models: a systematic review of methods and assumptions

Spatial connectivity plays an important role in mosquito-borne disease transmission. Connectivity can arise for many reasons, including shared environments, vector ecology and human movement. This systematic review synthesizes the spatial methods used to model mosquito-borne diseases, their spatial connectivity assumptions and the data used to inform spatial model components. We identified 248 papers eligible for inclusion. Most used statistical models (84.2%), although mechanistic are increasingly used. We identified 17 spatial models which used one of four methods (spatial covariates, local regression, random effects/fields and movement matrices). Over 80% of studies assumed that connectivity was distance-based despite this approach ignoring distant connections and potentially oversimplifying the process of transmission. Studies were more likely to assume connectivity was driven by human movement if the disease was transmitted by an Aedes mosquito. Connectivity arising from human movement was more commonly assumed in studies using a mechanistic model, likely influenced by a lack of statistical models able to account for these connections. Although models have been increasing in complexity, it is important to select the most appropriate, parsimonious model available based on the research question, disease transmission process, the spatial scale and availability of data, and the way spatial connectivity is assumed to occur.