Socioexposomics of COVID-19 across New Jersey: a comparison of geostatistical and machine learning approaches

Abstract

Background Disparities in adverse COVID-19 health outcomes have been associated with multiple social and environmental stressors. However, research is needed to evaluate the consistency and efficiency of methods for studying these associations at local scales. Objective To assess socioexposomic associations with COVID-19 outcomes across New Jersey and evaluate consistency of findings from multiple modeling approaches. Methods We retrieved data for COVID-19 cases and deaths for the 565 municipalities of New Jersey up to the end of the first phase of the pandemic, and calculated mortality rates with and without long-term-care (LTC) facility deaths. We considered 84 spatially heterogeneous environmental, demographic and socioeconomic factors from publicly available databases, including air pollution, proximity to industrial sites/facilities, transportation-related noise, occupation and commuting, neighborhood and housing characteristics, age structure, racial/ethnic composition, poverty, etc. Six geostatistical models (Poisson/Negative-Binomial regression, Poison/Negative-Binomial mixed effect model, Poisson/Negative-Binomial Bersag-York-Mollie spatial model) and two Machine Learning (ML) methods (Random Forest, Extreme Gradient Boosting) were implemented to assess association patterns. The Shapley effects plot was established for explainable ML and change of support validation was introduced to compare performances of different approaches. Results We found robust positive associations of COVID-19 mortality with historic exposures to NO2, population density, percentage of minority and below high school education, and other social and environmental factors. Exclusion of LTC deaths does not significantly affect correlations for most factors but findings can be substantially influenced by model structures and assumptions. The best performing geostatistical models involved flexible structures representing data variations. ML methods captured association patterns consistent with the best performing geostatistical models, and furthermore detected consistent nonlinear associations not captured by geostatistical models. Significance The findings of this work improve the understanding of how social and environmental disparities impacted COVID-19 outcomes across New Jersey.

Towards Improved XAI-Based Epidemiological Research into the Next Potential Pandemic

“…The analyzed studies focus on different research objectives. Depending on the focus, they generate different explanations on the significance of different influential factors on the spread of the pandemic, such as compliance with interventions [133], population density [134], population movement and gathering [76,78,92], lock down effects [18,120], labor and unemployment effects [72,81], closure and regulation of schools [129,135], vaccination [130][131][132], spatial effects [74,136], weather conditions [127,137,138], country dietary and cultural effects [51,61], virus variants [98], and health infrastructural impacts [139].…”

“…The feature selection step is performed in most of the studies. It consists, e.g., of statistical analysis in [72,133], and stepwise wrapping methods based on k-fold validation [61,76,92,136,137,139].…”

“…For example, [17] introduces ensemble models of pre-trained CNNs with large changes in CNN weights and applied uncertain-CAM approaches to their model explanation. In the epidemiological context, uncertainty is rarely performed based on Bayesian approaches [137]. Elaborating more on different non-deterministic approaches, including ensemble models (to best synthesize the predictions of multiple basic models) [93] and Bayesian neural networks (to infer distributions over the models' weights and outputs) [155], can enhance future epidemiological pipelines.…”

Spatial statistical methodologies in COVID-19 Studies: A systematic review

“…For instance, based on a geostatistical method, modelling the spatial distribution of the COVID-19 infection risk in mainland Portugal was done using a direct block sequential simulation (48). In the United States, a total of six geostatistical models were used to analyse association patterns when examining socioexposomic connections with COVID-19 outcomes across New Jersey (49). Geostatistical methods were also utilized to detect COVID-19 hotspots in the 2020 North Carolina general election in the US (50).…”

Spatial Clusters of Cancer Mortality in Brazil: A Machine Learning Modeling Approach

“…These include predicting the mortality risk during chemotherapy [4,5], predicting in-hospital mortality [6,7], and making prognostic predictions [8][9][10]. Yet, their application in ecological epidemiology modeling is still growing [11][12][13].…”