Journal of Data Science

The new coronavirus disease (COVID-19), as a new infectious disease, has relatively strong ability to spread from person to person. This paper studies several meteorological factors and air quality indicators between Shenzhen and Wenzhou, China, and conducts modelling analysis on whether the transmission of COVID-19 is affected by atmosphere. A comparative assessment is made on the characteristics of meteorological factors and air quality in these two typical cities in China and their impacts on the spread of COVID-19. The article uses meteorological data and air quality data, including 7 variables: daily average temperature, daily average relative humidity, daily average wind speed, nitrogen dioxide (NO2), atmospheric fine particulate matter (PM2.5), carbon monoxide (CO) and ozone (O3), a distributed lag non-linear model (DLNM) is constructed to explore the correlation between atmospheric conditions and non-imported confirmed cases of COVID-19, and the relative risk is introduced to measure the lagging effects of meteorological factors and air pollution on the number of non-imported confirmed cases. Our finding indicates that there is significant differences in the relationship between 7 predictors and the transmission of COVID-19 in Shenzhen and Wenzhou. However, all predictors between the two cities have a non-linear relationship with the number of non-imported confirmed cases. The lower daily average temperature has increased the risk of epidemic transmission in the two cities. As the temperature rises, the risk of epidemic transmission in both cities will significantly decrease. The average daily relative humidity has no significant effects on the epidemic situation in Shenzhen, but the lower relative humidity reduces the risk of epidemic spread in Wenzhou. In contrast, meteorological data have significant impacts on the spread of COVID-19 in Wenzhou. The four predictors (NO2, PM2.5, CO, and O3) have significant effects on the number of nonimported confirmed cases. Among them, PM2.5 has a significant positive correlation with the number of non-imported confirmed cases. Daily average wind speed, NO2 and O3 have different effects on the number of non-imported confirmed cases in different cities.

Abstract: Particulate matter smaller than 2.5 microns (PM2.5) is a com monly measured parameter in ground-based sampling networks designed to assess short and long-term air quality. The measurement techniques for ground based PM2.5 are relatively accurate and precise, but monitoring lo cations are spatially too sparse for many applications. Aerosol Optical Depth (AOD) is a satellite based air quality measurement that can be computed for more spatial locations, but measures light attenuation by particulates throughout in entire air column, not just near the ground. The goal of this paper is to better characterize the spatio-temporal relationship between the two measurements. An informative relationship will aid in imputing PM2.5 values for health studies in a way that accounts for the variability in both sets of measurements, something physics based models cannot do. We use a data set of Chicago air quality measurements taken during 2007 and 2008 to construct a weekly hierarchical model. We also demonstrate that AOD measurements and a latent spatio-temporal process aggregated weekly can be used to aid in the prediction of PM2.5measurements.

Abstract: Graphical procedures can be useful for illustrating and evaluating the process of inverse regression. We first review some simple and well-known graphical approaches for univariate linear and nonlinear models. We then propose a new graphical tool applicable to situations where the response is bivariate and repeated measures data are available. The proposed method is illustrated with an example of the age determination of tern chicks using measurements on body weight and wing length.

Statistical models for clinical risk prediction are often derived using data from primary care databases; however, they are frequently used outside of clinical settings. The use of prediction models in epidemiological studies without external validation may lead to inaccurate results. We use the example of applying the QRISK3 model to data from the United Kingdom (UK) Biobank study to illustrate the challenges and provide suggestions for future authors. The QRISK3 model is recommended by the National Institute for Health and Care Excellence (NICE) as a tool to aid cardiovascular risk prediction in English and Welsh primary care patients aged between 40 and 74. QRISK3 has not been externally validated for use in studies where data is collected for more general scientific purposes, including the UK Biobank study. This lack of external validation is important as the QRISK3 scores of participants in UK Biobank have been used and reported in several publications. This paper outlines: (i) how various publications have used QRISK3 on UK Biobank data and (ii) the ways that the lack of external validation may affect the conclusions from these publications. We then propose potential solutions for addressing these challenges; for example, model recalibration and considering alternative models, for the application of traditional statistical models such as QRISK3, in cohorts without external validation.