Assessing COVID-19 risk with temporal indices and geographically weighted ordinal logistic regression in US counties.

PURPOSE: Research on the novel coronavirus diseases 2019 (COVID-19) mainly relies on cross-sectional data, but this approach fails to consider the temporal dimension of the pandemic. This study assesses three temporal dimensions of the COVID-19 infection risk in US counties, namely probability of occurrence, duration of the pandemic, and intensity of transmission, and investigate local patterns of the factors associated with these risks. METHODS: Analyzing daily data between January 22 and September 11, 2020, we categorize the contiguous US counties into four risk groups-High-Risk, Moderate-Risk, Mild-Risk, and Low-Risk-and then apply both conventional (i.e., non-spatial) and geographically weighted (i.e., spatial) ordinal logistic regression model to understand the county-level factors raising the COVID-19 infection risk. The comparisons of various model fit diagnostics indicate that the spatial models better capture the associations between COVID-19 risk and other factors. RESULTS: The key findings include (1) High- and Moderate-Risk counties are clustered in the Black Belt, the coastal areas, and Great Lakes regions. (2) Fragile labor markets (e.g., high percentages of unemployed and essential workers) and high housing inequality are associated with higher risks. (3) The Monte Carlo tests suggest that the associations between covariates and COVID-19 risk are spatially non-stationary. For example, counties in the northeastern region and Mississippi Valley experience a stronger impact of essential workers on COVID-19 risk than those in other regions, whereas the association between income ratio and COVID-19 risk is stronger in Texas and Louisiana. CONCLUSIONS: The COVID-19 infection risk levels differ greatly across the US and their associations with structural inequality and sociodemographic composition are spatially non-stationary, suggesting that the same stimulus may not lead to the same change in COVID-19 risk. Potential interventions to lower COVID-19 risk should adopt a place-based perspective.

Exploring heterogeneities with geographically weighted quantile regression: An enhancement based on the bootstrap approach.

Geographically weighted quantile regression (GWQR) has been proposed as a spatial analytical technique to simultaneously explore two heterogeneities, one of spatial heterogeneity with respect to data relationships over space and one of response heterogeneity across different locations of the outcome distribution. However, one limitation of GWQR framework is that the existing inference procedures are established based on asymptotic approximation, which may suffer computation difficulties or yield incorrect estimates with finite samples. In this paper, we suggest a bootstrap approach to address this limitation. Our bootstrap enhancement is first validated by a simulation experiment and then illustrated with an empirical US mortality data. The results show that the bootstrap provides a practical alternative for inference in GWQR and enhances the utilization of GWQR.

When homogeneity meets heterogeneity: the geographically weighted regression with spatial lag approach to prenatal care utilization.

Using geographically weighted regression (GWR), a recent study by Shoff and colleagues (2012) investigated the place-specific risk factors for prenatal care utilisation in the United States of America (USA) and found that most of the relationships between late or no prenatal care and its determinants are spatially heterogeneous. However, the GWR approach may be subject to the confounding effect of spatial homogeneity. The goal of this study was to address this concern by including both spatial homogeneity and heterogeneity into the analysis. Specifically, we employed an analytic framework where a spatially lagged (SL) effect of the dependent variable is incorporated into the GWR model, which is called GWR-SL. Using this framework, we found evidence to argue that spatial homogeneity is neglected in the study by Shoff et al. (2012) and that the results change after considering the SL effect of prenatal care utilisation. The GWR-SL approach allowed us to gain a placespecific understanding of prenatal care utilisation in USA counties. In addition, we compared the GWR-SL results with the results of conventional approaches (i.e., ordinary least squares and spatial lag models) and found that GWR-SL is the preferred modelling approach. The new findings help us to better estimate how the predictors are associated with prenatal care utilisation across space, and determine whether and how the level of prenatal care utilisation in neighbouring counties matters.

Using quantile regression to examine the effects of inequality across the mortality distribution in the U.S. counties.

The U.S. has experienced a resurgence of income inequality in the past decades. The evidence regarding the mortality implications of this phenomenon has been mixed. This study employs a rarely used method in mortality research, quantile regression (QR), to provide insight into the ongoing debate of whether income inequality is a determinant of mortality and to investigate the varying relationship between inequality and mortality throughout the mortality distribution. Analyzing a U.S. dataset where the five-year (1998-2002) average mortality rates were combined with other county-level covariates, we found that the association between inequality and mortality was not constant throughout the mortality distribution and the impact of inequality on mortality steadily increased until the 80th percentile. When accounting for all potential confounders, inequality was significantly and positively related to mortality; however, this inequality-mortality relationship did not hold across the mortality distribution. A series of Wald tests confirmed this varying inequality-mortality relationship, especially between the lower and upper tails. The large variation in the estimated coefficients of the Gini index suggested that inequality had the greatest influence on those counties with a mortality rate of roughly 9.95 deaths per 1000 population (80th percentile) compared to any other counties. Furthermore, our results suggest that the traditional analytic methods that focus on mean or median value of the dependent variable can be, at most, applied to a narrow 20 percent of observations. This study demonstrates the value of QR. Our findings provide some insight as to why the existing evidence for the inequality-mortality relationship is mixed and suggest that analytical issues may play a role in clarifying whether inequality is a robust determinant of population health.

Geographically Weighted Quantile Regression (GWQR): An Application to U.S. Mortality Data.

In recent years, techniques have been developed to explore spatial non-stationarity and to model the entire distribution of a regressand. The former is mainly addressed by geographically weighted regression (GWR), and the latter by quantile regression (QR). However, little attention has been paid to combining these analytical techniques. The goal of this article is to fill this gap by introducing geographically weighted quantile regression (GWQR). This study briefly reviews GWR and QR, respectively, and then outlines their synergy and a new approach, GWQR. The estimations of GWQR parameters and their standard errors, the cross-validation bandwidth selection criterion, and the non-stationarity test are discussed. We apply GWQR to U.S. county data as an example, with mortality as the dependent variable and five social determinants as explanatory covariates. Maps summarize analytic results at the 5, 25, 50, 75, and 95 percentiles. We found that the associations between mortality and determinants vary not only spatially, but also simultaneously across the distribution of mortality. These new findings provide insights into the mortality literature, and are relevant to public policy and health promotion. Our GWQR approach bridges two important statistical approaches, and facilitates spatial quantile-based statistical analyses.

SAS macro programs for geographically weighted generalized linear modeling with spatial point data: applications to health research.

An increasing interest in exploring spatial non-stationarity has generated several specialized analytic software programs; however, few of these programs can be integrated natively into a well-developed statistical environment such as SAS. We not only developed a set of SAS macro programs to fill this gap, but also expanded the geographically weighted generalized linear modeling (GWGLM) by integrating the strengths of SAS into the GWGLM framework. Three features distinguish our work. First, the macro programs of this study provide more kernel weighting functions than the existing programs. Second, with our codes the users are able to better specify the bandwidth selection process compared to the capabilities of existing programs. Third, the development of the macro programs is fully embedded in the SAS environment, providing great potential for future exploration of complicated spatially varying coefficient models in other disciplines. We provided three empirical examples to illustrate the use of the SAS macro programs and demonstrated the advantages explained above.

Examining non-stationary effects of social determinants on cardiovascular mortality after cold surges in Taiwan.

Details about the impact of extreme cold on cardiovascular mortality are little known in the countries with warm winters like Taiwan. This study aimed to examine the ecological associations between various social determinants and cardiovascular mortality after cold surges in Taiwan with a spatial perspective focusing on spatial non-stationarity. The mortality rates at township level in Taiwan were observed from 1997 to 2003. Five social determinants (social disadvantage, lack of economic opportunity, stability, sensitive group, and rurality) were created with the 2000 Taiwan Census data. We analyzed the data using Geographically Weighted Poisson Regression. On average, an immediate increase in cardiovascular mortality was found right after cold surges. All of the five determinants were found to have spatial non-stationary effects on the cardiovascular mortality rates after cold surges. This finding provided an empirical basis for developing public health programs with local emphases on the impacts of extreme cold.

Cold surge: a sudden and spatially varying threat to health?

While cold surge is one of the most conspicuous features of the winter monsoon in East Asia, its impact on human health remains underexplored. Based on the definition by the Central Weather Bureau in Taiwan, we identified four cold surges between 2000 and 2003 and collected the cardiovascular disease mortality data 2 weeks before and 2 weeks after these events. We attempted to answer the following research questions: 1) whether the cold surges impose an adverse and immediate effect on cardiovascular mortality; 2) whether the people living in temperate zones have a higher tolerance of extreme temperature drop than those in the subtropics. With geographic weighting techniques, we not only found that the cardiovascular disease mortality rates increased significantly after the cold surges, but also discovered a spatially varying pattern of tolerance to cold surges. Even within a small study area such as Taiwan, human reaction to severe weather drop differs across space. Needless to say, in the U.S., these findings should be considered in redirecting policy to address populations living in warm places when extreme temperature drops occur.