Abstract .edu



A Comprehensive Analysis of COVID-19 Transmission and Mortality Rates at the County level in the United States considering Socio-Demographics, Health Indicators, Mobility Trends and Health Care Infrastructure AttributesTanmoy Bhowmik1*, Sudipta Dey Tirtha1, Naveen Chandra Iraganaboina1, Naveen Eluru11Department of Civil, Environmental & Construction Engineering, University of Central Florida*Corresponding author (TB)Email: tanmoy78@knights.ucf.eduAbstractBackground: Several research efforts have evaluated the impact of various factors including a) socio-demographics, (b) health indicators, (c) mobility trends, and (d) health care infrastructure attributes on COVID-19 transmission and mortality rate. However, earlier research focused only on a subset of variable groups (predominantly one or two) that can contribute to the COVID-19 transmission/mortality rate. The current study effort is designed to remedy this by analyzing COVID-19 transmission/mortality rates considering a comprehensive set of factors in a unified framework. Methods and findings: We study two per capita dependent variables: (1) daily COVID-19 transmission rates and (2) total COVID-19 mortality rates. The first variable is modeled using a linear mixed model while the later dimension is analyzed using a linear regression approach. The model results are augmented with a sensitivity analysis to predict the impact of mobility restrictions at a county level. Several county level factors including proportion of African-Americans, income inequality, health indicators associated with Asthma, Cancer, HIV and heart disease, percentage of stay at home individuals, testing infrastructure and Intensive Care Unit capacity impact transmission and/or mortality rates. From the policy analysis, we find that enforcing a stay at home order that can ensure a 50% stay at home rate can result in a potential reduction of about 33% in daily cases. Conclusions: The model framework developed can be employed by government agencies to evaluate the influence of reduced mobility on transmission rates at a county level while accommodating for various county specific factors. Based on our policy analysis, the study findings support a county level stay at home order for regions currently experiencing a surge in transmission. The model framework can also be employed to identify vulnerable counties that need to be prioritized based on health indicators for current support and/or preferential vaccination plans (when available).Keywords: COVID-19, transmission rate, mortality rate, linear mixed model, policy analysis, vulnerable countiesIntroductionCoronavirus disease 2019 (COVID-19) pandemic, as of August 20th, has spread to 188 countries with a reported 23.1 million cases and 802 thousand fatalities ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1101/2020.01.23.20018549V2","URL":"?","accessed":{"date-parts":[["2020","7","12"]]},"author":[{"dropping-particle":"","family":"Worldometer","given":"","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Worldometer","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"page":"1-22","title":"Coronavirus Cases & Mortality","type":"webpage"},"uris":[""]}],"mendeley":{"formattedCitation":"(1)","plainTextFormattedCitation":"(1)","previouslyFormattedCitation":"(1)"},"properties":{"noteIndex":0},"schema":""}(1). The pandemic has affected the mental and physical health of people across the world significantly taxing the social, health and economic systems ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","accessed":{"date-parts":[["2020","7","12"]]},"author":[{"dropping-particle":"","family":"Azizi","given":"Mo","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"the World Bank","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"page":"5-9","title":"The Global Economic Outlook During the COVID-19 Pandemic: A Changed World","type":"webpage","volume":"12"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1101/2021.02.19.21252117","author":[{"dropping-particle":"","family":"Bhowmik","given":"Tanmoy","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Eluru","given":"Naveen","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-2","issued":{"date-parts":[["0"]]},"title":"A Comprehensive County Level Framework to Identify Factors Affecting Hospital Capacity and Predict Future Hospital Demand","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(2,3)","plainTextFormattedCitation":"(2,3)","previouslyFormattedCitation":"(2,3)"},"properties":{"noteIndex":0},"schema":""}(2,3). Among the various countries affected, United States has reported the highest number of confirmed cases (5.5 million) and deaths (173 thousand) in the world ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","abstract":"TOTAL CASES 1,122,486 29,671 New Cases* TOTAL DEATHS 65,735 1,452 New Deaths*","accessed":{"date-parts":[["2020","7","12"]]},"author":[{"dropping-particle":"","family":"Centers for Disease Control and Prevention (CDC)","given":"","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Coronavirus Disease 2019 (COVID-19)","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"page":"1-4","title":"Cases in the U.S.","type":"webpage","volume":"2019"},"uris":[""]}],"mendeley":{"formattedCitation":"(4)","plainTextFormattedCitation":"(4)","previouslyFormattedCitation":"(4)"},"properties":{"noteIndex":0},"schema":""}(4). In this context, it is important that we clearly understand the factors affecting COVID-19 transmission and mortality rate to prescribe policy actions grounded in empirical evidence to slow the spread of the transmission and/or prepare action plans for potential vaccination programs in the near future. Towards contributing to these objectives, the current study develops a comprehensive framework for examining COVID-19 transmission and mortality rates in the United States using COVID-19 data at a county level encompassing about 93% of the US population. The study effort is designed with the objective of including a universal set of factors affecting COVID-19 in the analysis of transmission and mortality rates. We employ an exhaustive set of county level characteristics including (a) socio-demographics, (b) health indicators, (c) mobility trends, and (d) health care infrastructure attributes. We recognize that analysis of COVID-19 data without including potentially important factors , as has been the case with earlier work, is likely to yield incorrect/biased estimates for the factors considered. The framework proposed for understanding and quantifying the influence of these factors can allow policy makers to (a) evaluate the influence of population behavior factors such as mobility trends on virus transmission (while accounting for other county level factors), (b) identify priority locations for health infrastructure support as the pandemic evolves, and (c) prioritize vulnerable counties across the country for vaccination (when available). In recent months, a number of research efforts have examined COVID-19 data in several countries to identify the factors influencing COVID-19 transmission and mortality. Given the focus of our current study, we restrict our review to studies that explore COVID-19 transmission and mortality rate at an aggregated spatial scale. To elaborate, these studies explored COVID-19 transmission and mortality rates at the national ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.dsx.2020.03.013","ISSN":"18780334","PMID":"32247212","abstract":"Background and aims: Many patients with coronavirus disease 2019 (COVID-19) have underlying cardiovascular (CV) disease or develop acute cardiac injury during the course of the illness. Adequate understanding of the interplay between COVID-19 and CV disease is required for optimum management of these patients. Methods: A literature search was done using PubMed and Google search engines to prepare a narrative review on this topic. Results: Respiratory illness is the dominant clinical manifestation of COVID-19; CV involvement occurs much less commonly. Acute cardiac injury, defined as significant elevation of cardiac troponins, is the most commonly reported cardiac abnormality in COVID-19. It occurs in approximately 8–12% of all patients. Direct myocardial injury due to viral involvement of cardiomyocytes and the effect of systemic inflammation appear to be the most common mechanisms responsible for cardiac injury. The information about other CV manifestations in COVID-19 is very limited at present. Nonetheless, it has been consistently shown that the presence of pre-existing CV disease and/or development of acute cardiac injury are associated with significantly worse outcome in these patients. Conclusions: Most of the current reports on COVID-19 have only briefly described CV manifestations in these patients. Given the enormous burden posed by this illness and the significant adverse prognostic impact of cardiac involvement, further research is required to understand the incidence, mechanisms, clinical presentation and outcomes of various CV manifestations in COVID-19 patients.","author":[{"dropping-particle":"","family":"Bansal","given":"Manish","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Diabetes and Metabolic Syndrome: Clinical Research and Reviews","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2020"]]},"page":"247-250","title":"Cardiovascular disease and COVID-19","type":"article-journal","volume":"14"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.2139/ssrn.3565703","ISSN":"1556-5068","abstract":"We combine GPS data on changes in average distance traveled by individuals at the county level with COVID-19 case data and other demographic information to estimate how individual mobility is affected by local disease prevalence and restriction orders to stay-at-home. We find that a rise of local infection rate from 0% to 0.003% 1 is associated with a reduction in mobility by 2.31%. An official stay-at-home restriction order corresponds to reducing mobility by 7.87%. Counties with larger shares of population over age 65, lower share of votes for the Republican Party in the 2016 Presidential Election, and higher population density are more responsive to disease prevalence and restriction orders.","author":[{"dropping-particle":"","family":"Engle","given":"Samuel","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Stromme","given":"John","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhou","given":"Anson","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"title":"Staying at Home: Mobility Effects of COVID-19","type":"article-journal"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.1371/journal.pone.0236003","ISSN":"19326203","PMID":"32706790","abstract":"The emergence and fast global spread of COVID-19 has presented one of the greatest public health challenges in modern times with no proven cure or vaccine. Africa is still early in this epidemic, therefore the extent of disease severity is not yet clear. We used a mathematical model to fit to the observed cases of COVID-19 in South Africa to estimate the basic reproductive number and critical vaccination coverage to control the disease for different hypothetical vaccine efficacy scenarios. We also estimated the percentage reduction in effective contacts due to the social distancing measures implemented. Early model estimates show that COVID-19 outbreak in South Africa had a basic reproductive number of 2.95 (95% credible interval [CrI] 2.83-3.33). A vaccine with 70% efficacy had the capacity to contain COVID-19 outbreak but at very higher vaccination coverage 94.44% (95% Crl 92.44-99.92%) with a vaccine of 100% efficacy requiring 66.10% (95% Crl 64.72-69.95%) coverage. Social distancing measures put in place have so far reduced the number of social contacts by 80.31% (95% Crl 79.76-80.85%). These findings suggest that a highly efficacious vaccine would have been required to contain COVID-19 in South Africa. Therefore, the current social distancing measures to reduce contacts will remain key in controlling the infection in the absence of vaccines and other therapeutics.","author":[{"dropping-particle":"","family":"Mukandavire","given":"Zindoga","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Nyabadza","given":"Farai","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Malunguza","given":"Noble J.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Cuadros","given":"Diego F.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Shiri","given":"Tinevimbo","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Musuka","given":"Godfrey","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"PLoS ONE","id":"ITEM-3","issue":"7 July","issued":{"date-parts":[["2020"]]},"page":"e0236003","title":"Quantifying early COVID-19 outbreak transmission in South Africa and exploring vaccine efficacy scenarios","type":"article-journal","volume":"15"},"uris":[""]},{"id":"ITEM-4","itemData":{"DOI":"10.14218/erhm.2020.00023","abstract":"BACKGROUND AND OBJECTIVES The coronavirus disease 2019 (COVID-19) infected 586,000 patients in the U.S. However, the COVDI-19 daily incidence and deaths in the U.S. are poorly understood. Internet search interest was found highly correlated with COVID-19 daily incidence in China, but not yet applied to the U.S. Therefore, we examined the association of search interest with COVID-19 daily incidence and deaths in the U.S. METHODS We extracted the COVDI-19 daily incidence and death data in the U.S. from two population-based datasets. The search interest of COVID-19 related terms was obtained using Google Trends. Pearson correlation test and general linear model were used to examine correlations and predict future trends, respectively. RESULTS There were 555,245 new cases and 22,019 deaths of COVID-19 reported in the U.S. from March 1 to April 12, 2020. The search interest of COVID, COVID pneumonia, and COVID heart were correlated with COVDI-19 daily incidence with about 12-day of delay (Pearson r=0.978, 0.978 and 0.979, respectively) and deaths with 19-day of delay (Pearson r=0.963, 0.958 and 0.970, respectively). The COVID-19 daily incidence and deaths appeared to both peak on April 10. The 4-day follow-up with prospectively collected data showed moderate to good accuracies for predicting new cases (Pearson r=-0.641 to -0.833) and poor to good acuracies for daily new deaths (Pearson r=0.365 to 0.935). CONCLUSIONS Search terms related to COVID-19 are highly correlated with the trends in COVID-19 daily incidence and deaths in the U.S. The prediction-models based on the search interest trend reached moderate to good accuracies.","author":[{"dropping-particle":"","family":"Yuan","given":"Xiaoling","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Xu","given":"Jie","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Hussain","given":"Sabiha","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wang","given":"He","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Gao","given":"Nan","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhang","given":"Lanjing","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Exploratory Research and Hypothesis in Medicine","id":"ITEM-4","issue":"000","issued":{"date-parts":[["2020"]]},"page":"1-6","title":"Trends and Prediction in Daily New Cases and Deaths of COVID-19 in the United States: An Internet Search-Interest Based Model","type":"article-journal","volume":"000"},"uris":[""]}],"mendeley":{"formattedCitation":"(5–8)","plainTextFormattedCitation":"(5–8)","previouslyFormattedCitation":"(5–8)"},"properties":{"noteIndex":0},"schema":""}(5–8), regional ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"abstract":"In the absence of a vaccine or more effective treatment options, containing the spread of novel coronavirus disease 2019 (COVID-19) must rely on non-pharmaceutical interventions. All U.S. states adopted social-distancing measures in March and April of 2020, though they varied in both timing and scope. Kentucky began by closing public schools and restaurant dining rooms on March 16th before progressing to closing other non-essential businesses and eventually issuing a “Healthy at Home” order with restrictions similar to the shelter-in-place (SIPO) orders adopted by other states. We aim to quantify the impact of these measures on COVID-19 case growth in the state. An event-study model allows us to link adoption of social distancing measures across the Midwest and South to the growth rate of cases, allowing for effects to emerge gradually to account for the lag between infection and positive test result. We then use the results to predict how the number of cases would have evolved in Kentucky in the absence of these policy measures – in other words, if the state had relied on voluntary social distancing alone. We estimate that, by April 25, Kentucky would have had 44,482 confirmed COVID-19 cases without social distancing restrictions, as opposed to the 3,857 actually observed.","author":[{"dropping-particle":"","family":"Courtemanche","given":"Charles J","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Garuccio","given":"Joseph","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Le","given":"Anh","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Pinkston","given":"Joshua C","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Yelowitz","given":"Aaron","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Institute for the study of free enterprise working papers","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"page":"1-23","title":"Did Social-Distancing Measures in Kentucky Help to Flatten the COVID-19 Curve?","type":"article-journal","volume":"4"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1371/journal.pone.0234763","ISSN":"19326203","PMID":"32628673","abstract":"This work applies a data-driven coding method for prediction of the COVID-19 spreading profile in any given population that shows an initial phase of epidemic progression. Based on the historical data collected for COVID-19 spreading in 367 cities in China and the set of parameters of the augmented Susceptible-Exposed-Infected-Removed (SEIR) model obtained for each city, a set of profile codes representing a variety of transmission mechanisms and contact topologies is formed. By comparing the data of an early outbreak of a given population with the complete set of historical profiles, the best fit profiles are selected and the corresponding sets of profile codes are used for prediction of the future progression of the epidemic in that population. Application of the method to the data collected for South Korea, Italy and Iran shows that peaks of infection cases are expected to occur before mid April, the end of March and the end of May 2020, and that the percentage of population infected in each city or region will be less than 0.01%, 0.5% and 0.5%, for South Korea, Italy and Iran, respectively.","author":[{"dropping-particle":"","family":"Zhan","given":"Choujun","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Tse","given":"Chi K.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Lai","given":"Zhikang","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Hao","given":"Tianyong","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Su","given":"Jingjing","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"PLoS ONE","id":"ITEM-2","issue":"7 July","issued":{"date-parts":[["2020"]]},"title":"Prediction of COVID-19 spreading profiles in South Korea, Italy and Iran by data-driven coding","type":"article-journal","volume":"15"},"uris":[""]}],"mendeley":{"formattedCitation":"(9,10)","plainTextFormattedCitation":"(9,10)","previouslyFormattedCitation":"(9,10)"},"properties":{"noteIndex":0},"schema":""}(9,10), state ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1109/ICNC.2010.5583921","ISBN":"9781424459612","abstract":"This paper considers scheduling a multiple-load vehicle of a material handling system (MHS) in a manufacturing system. The orders of handling materials are generated dynamically by the manufacturing system. This paper proposes a support vector machine (SVM) based scheduling approach to tackle the scheduling problem. The approach generates training samples and uses the samples to train an SVM offline. The trained SVM is then used online. Online information at decision points are structured as an input to the SVM. The output of the SVM is used to make decisions such as \"wait for next order\" or \"deliver immediately\". Simulation results show that the proposed approach outperforms other approaches. ?2010 IEEE.","author":[{"dropping-particle":"","family":"Ci","given":"Chen","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhou","given":"Binghai","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Xi","given":"Lifeng","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Proceedings - 2010 6th International Conference on Natural Computation, ICNC 2010","id":"ITEM-1","issued":{"date-parts":[["2010"]]},"page":"3768-3772","title":"A support vector machine based scheduling approach for a material handling system","type":"paper-conference","volume":"7"},"uris":[""]}],"mendeley":{"formattedCitation":"(11)","plainTextFormattedCitation":"(11)","previouslyFormattedCitation":"(11)"},"properties":{"noteIndex":0},"schema":""}(11), county ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3565703","ISSN":"1556-5068","abstract":"We combine GPS data on changes in average distance traveled by individuals at the county level with COVID-19 case data and other demographic information to estimate how individual mobility is affected by local disease prevalence and restriction orders to stay-at-home. We find that a rise of local infection rate from 0% to 0.003% 1 is associated with a reduction in mobility by 2.31%. An official stay-at-home restriction order corresponds to reducing mobility by 7.87%. Counties with larger shares of population over age 65, lower share of votes for the Republican Party in the 2016 Presidential Election, and higher population density are more responsive to disease prevalence and restriction orders.","author":[{"dropping-particle":"","family":"Engle","given":"Samuel","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Stromme","given":"John","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhou","given":"Anson","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Staying at Home: Mobility Effects of COVID-19","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.2139/ssrn.3585921","ISSN":"1556-5068","abstract":"This paper provides insights for policymakers to evaluate the impact of staying at home and lockdown policies by investigating possible links between individual mobility and the spread of the COVID-19 virus in Italy. By relying on the daily data, the empirical evidence suggests that an increase in the number of visits to public spaces such as workspaces, parks, retail areas, and the use of public transportation is associated with an increase in the positive COVID-19 cases in a subsequent week. On the contrary, the increased intensity of staying in residential spaces is related to a decrease in the confirmed cases of COVID-19 significantly. Results are robust after controlling for the lockdown period. Empirical evidence underlines the importance of the lockdown decision. Further, there is substantial regional variation among the twenty regions of Italy. Individual presence in public vs. residential spaces has a more significant e?ect on the number of COVID-19 cases in the Lombardy region.","author":[{"dropping-particle":"","family":"Bilgin","given":"Nuriye Melisa","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"title":"Tracing COVID-19 Spread in Italy with Mobility Data","type":"article-journal"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.1016/j.scitotenv.2020.138884","ISSN":"18791026","PMID":"32335404","abstract":"During the first 90 days of the COVID-19 outbreak in the United States, over 675,000 confirmed cases of the disease have been reported, posing unprecedented socioeconomic burden to the country. Due to inadequate research on geographic modeling of COVID-19, we investigated county-level variations of disease incidence across the continental United States. We compiled a geodatabase of 35 environmental, socioeconomic, topographic, and demographic variables that could explain the spatial variability of disease incidence. Further, we employed spatial lag and spatial error models to investigate spatial dependence and geographically weighted regression (GWR) and multiscale GWR (MGWR) models to locally examine spatial non-stationarity. The results suggested that even though incorporating spatial autocorrelation could significantly improve the performance of the global ordinary least square model, these models still represent a significantly poor performance compared to the local models. Moreover, MGWR could explain the highest variations (adj. R2: 68.1%) with the lowest AICc compared to the others. Mapping the effects of significant explanatory variables (i.e., income inequality, median household income, the proportion of black females, and the proportion of nurse practitioners) on spatial variability of COVID-19 incidence rates using MGWR could provide useful insights to policymakers for targeted interventions.","author":[{"dropping-particle":"","family":"Mollalo","given":"Abolfazl","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Vahedi","given":"Behzad","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rivera","given":"Kiara M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-3","issued":{"date-parts":[["2020"]]},"title":"GIS-based spatial modeling of COVID-19 incidence rate in the continental United States","type":"article-journal","volume":"728"},"uris":[""]},{"id":"ITEM-4","itemData":{"DOI":"10.31235/osf.io/hzj7a","abstract":"To what degree does social distancing have a causal effect on the spread of SARS-CoV-2? To generate causal evidence, we show that week to week changes in weather conditions provided a natural experiment that altered daily travel and movement outside the home, and thus affected social distancing in the first several weeks when Covid-19 began to spread in many U.S. counties. Using aggregated mobile phone location data and leveraging changes in social distancing driven by weekly weather conditions, we provide the first causal evidence on the effect of social distancing on the spread of SARS-CoV-2. Results show that a 1 percent increase in distance traveled leads to an 8.1 percent increase in new cases per capita in the following week, and a 1 percent increase in non-essential visits leads to a 6.9 percent increase in new cases per capita in the following week. Results are stronger in densely populated counties and close to zero in less densely populated counties.","author":[{"dropping-particle":"","family":"Sharkey","given":"Patrick","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wood","given":"George","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-4","issued":{"date-parts":[["2020"]]},"publisher":"SocArXiv","title":"The Causal Effect of Social Distancing on the Spread of SARS-CoV-2","type":"article-journal"},"uris":[""]},{"id":"ITEM-5","itemData":{"DOI":"10.1101/2020.04.05.20054502","abstract":"Objectives: United States government scientists estimate that COVID-19 may kill tens of thousands of Americans. Many of the pre-existing conditions that increase the risk of death in those with COVID-19 are the same diseases that are affected by long-term exposure to air pollution. We investigated whether long-term average exposure to fine particulate matter (PM2.5) is associated with an increased risk of COVID-19 death in the United States. Design: A nationwide, cross-sectional study using county-level data. Data sources: COVID-19 death counts were collected for more than 3,000 counties in the United States (representing 98% of the population) up to April 22, 2020 from Johns Hopkins University, Center for Systems Science and Engineering Coronavirus Resource Center. Main outcome measures: We fit negative binomial mixed models using county-level COVID-19 deaths as the outcome and county-level long-term average of PM2.5 as the exposure. In the main analysis, we adjusted by 20 potential confounding factors including population size, age distribution, population density, time since the beginning of the outbreak, time since state’s issuance of stay-at-home order, hospital beds, number of individuals tested, weather, and socioeconomic and behavioral variables such as obesity and smoking. We included a random intercept by state to account for potential correlation in counties within the same state. We conducted more than 68 additional sensitivity analyses. Results: We found that an increase of only 1 ??g/m3 in PM2.5 is associated with an 8% increase in the the COVID-19 death rate (95% confidence interval [CI]: 2%, 15%). The results were statistically significant and robust to secondary and sensitivity analyses. Conclusions: A small increase in long-term exposure to PM2.5 leads to a large increase in the COVID-19 death rate. Despite the inherent limitations of the ecological study design, our results underscore the importance of continuing to enforce existing air pollution regulations to protect human health both during and after the COVID-19 crisis. The data and code are publicly available so our analyses can be updated routinely. 3","author":[{"dropping-particle":"","family":"Wu","given":"Xiao","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Nethery","given":"Rachel C","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Sabath","given":"Benjamin M","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Braun","given":"Danielle","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Dominici","given":"Francesca","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"medRxiv : the preprint server for health sciences","id":"ITEM-5","issued":{"date-parts":[["2020"]]},"title":"Exposure to air pollution and COVID-19 mortality in the United States: A nationwide cross-sectional study.","type":"article-journal"},"uris":[""]},{"id":"ITEM-6","itemData":{"DOI":"10.1101/2020.05.14.20101667","abstract":"Since the emerging of the &quot;novel coronavirus&quot; SARS-CoV-2 and the corresponding respiratory disease COVID-19, the virus has spread all over the world. In Europe, Germany is currently one of the most affected countries. In March 2020, a &quot;lockdown&quot; was established to contain the virus spread, including the closure of schools and child day care facilities as well as forced social distancing and bans of any public gathering. The present study attempts to analyze whether these governmental interventions had an impact on the declared aim of &quot;flattening the curve&quot;, referring to the epidemic curve of new infections. This analysis is conducted from a regional perspective. On the level of the 412 German counties, logistic growth models were estimated based on reported cases of infections, aiming at determining the regional growth rate of infections and the point of inflection where infection rates begin to decrease and the curve flattens. All German counties exceeded the peak of new infections between the beginning of March and the middle of April. In a large majority of German counties, the epidemic curve has flattened before the social ban was established (March 23). In a minority of counties, the peak was already exceeded before school closures. The growth rates of infections vary spatially depending on the time the virus emerged. Counties belonging to states which established an additional curfew show no significant improvement with respect to growth rates and mortality. On the contrary, growth rates and mortality are significantly higher in Bavaria compared to whole Germany. The results raise the question whether social ban measures and curfews really contributed to the curve flattening. Furthermore, mortality varies strongly across German counties, which can be attributed to infections of people belonging to the “risk group”, especially residents of retirement peting Interest StatementThe authors have declared no competing interest.Funding StatementNo external funding was received.Author DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with …","author":[{"dropping-particle":"","family":"Wieland","given":"Thomas","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"medRxiv","id":"ITEM-6","issued":{"date-parts":[["2020"]]},"page":"2020.05.14.20101667","publisher":"Cold Spring Harbor Laboratory Press","title":"Flatten the Curve! Modeling SARS-CoV-2/COVID-19 Growth in Germany on the County Level","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(6,12–16)","plainTextFormattedCitation":"(6,12–16)","previouslyFormattedCitation":"(6,12–16)"},"properties":{"noteIndex":0},"schema":""}(6,12–16), city ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.scitotenv.2020.138201","ISSN":"18791026","PMID":"32408450","abstract":"Background: Coronavirus disease 2019 (COVID-19) has become a severe public health problem globally. Both epidemiological and laboratory studies have shown that ambient temperature could affect the transmission and survival of coronaviruses. This study aimed to determine whether the temperature is an essential factor in the infection caused by this novel coronavirus. Methods: Daily confirmed cases and meteorological factors in 122 cities were collected between January 23, 2020, to February 29, 2020. A generalized additive model (GAM) was applied to explore the nonlinear relationship between mean temperature and COVID-19 confirmed cases. We also used a piecewise linear regression to determine the relationship in detail. Results: The exposure-response curves suggested that the relationship between mean temperature and COVID-19 confirmed cases was approximately linear in the range of <3 °C and became flat above 3 °C. When mean temperature (lag0–14) was below 3 °C, each 1 °C rise was associated with a 4.861% (95% CI: 3.209–6.513) increase in the daily number of COVID-19 confirmed cases. These findings were robust in our sensitivity analyses. Conclusions: Our results indicate that mean temperature has a positive linear relationship with the number of COVID-19 cases with a threshold of 3 °C. There is no evidence supporting that case counts of COVID-19 could decline when the weather becomes warmer, which provides useful implications for policymakers and the public.","author":[{"dropping-particle":"","family":"Xie","given":"Jingui","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhu","given":"Yongjian","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Association between ambient temperature and COVID-19 infection in 122 cities from China","type":"article-journal","volume":"724"},"uris":[""]}],"mendeley":{"formattedCitation":"(17)","plainTextFormattedCitation":"(17)","previouslyFormattedCitation":"(17)"},"properties":{"noteIndex":0},"schema":""}(17) and zip code levels ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3572329","ISSN":"1556-5068","abstract":"New York City is the hot spot of the Covid-19 pandemic in the United States. This paper merges information on the number of tests and the number of infections at the New York City zip code level with demographic and socioeconomic information from the decennial census and the American Community Surveys. People residing in poor or immigrant neighbourhoods were less likely to be tested; but the likelihood that a test was positive was larger in those neighbourhoods, as well as in neighbourhoods with larger households or predominantly black populations. The rate of infection in the population depends on both the frequency of tests and on the fraction of positive tests among those tested. The non-randomness in testing across New York City neighbourhoods indicates that the observed correlation between the rate of infection and the socioeconomic characteristics of a community tells an incomplete story of how the pandemic evolved in a congested urban setting.","author":[{"dropping-particle":"","family":"Borjas","given":"George J.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Demographic Determinants of Testing Incidence and COVID-19 Infections in New York City Neighborhoods","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(18)","plainTextFormattedCitation":"(18)","previouslyFormattedCitation":"(18)"},"properties":{"noteIndex":0},"schema":""}(18). A majority of these studies considered transmission rate as the response variable (transmission rate per capita). The main approach employed to identify the factors affecting the response variables is the linear regression approach. In their analysis, researchers employed a host of independent variables from four variable categories: socio-demographics, health indicators, mobility trends and health care infrastructure attributes. For socio- demographics, studies found income, race and age distribution have a positive association with the COVID-19 transmission ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3572329","ISSN":"1556-5068","abstract":"New York City is the hot spot of the Covid-19 pandemic in the United States. This paper merges information on the number of tests and the number of infections at the New York City zip code level with demographic and socioeconomic information from the decennial census and the American Community Surveys. People residing in poor or immigrant neighbourhoods were less likely to be tested; but the likelihood that a test was positive was larger in those neighbourhoods, as well as in neighbourhoods with larger households or predominantly black populations. The rate of infection in the population depends on both the frequency of tests and on the fraction of positive tests among those tested. The non-randomness in testing across New York City neighbourhoods indicates that the observed correlation between the rate of infection and the socioeconomic characteristics of a community tells an incomplete story of how the pandemic evolved in a congested urban setting.","author":[{"dropping-particle":"","family":"Borjas","given":"George J.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Demographic Determinants of Testing Incidence and COVID-19 Infections in New York City Neighborhoods","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.2139/ssrn.3571699","ISSN":"1556-5068","abstract":"COVID-19 is killing African Americans at a rate 7% to 193% higher than the general population. Understanding why, as well as the reasons behind the wide variation, is paramount to saving lives. Here, we test two potential explanations for this effect. On the one hand, African Americans might be dying more because they have a lower average income (‘the Socioeconomic Hypothesis’). On the other hand, they might be dying more because their skin is more resistant to UV radiation, as we previously showed that COVID-19 infections and deaths decrease with higher irradiance (‘the Irradiance Hypothesis’). The two hypotheses are not mutually exclusive. We show that the overrepresentation of African Americans among COVID-19 deaths shows a significant negative correlation with mean solar irradiance, with a 20% decrease in Global Horizontal Irradiance leading to a 76% increase in the overrepresentation of African Americans amongst COVID-19 deaths. We then show that in Michigan, one of the US states with the lowest irradiance in early April, the % of each county’s population that is black, more than its median income, median age or % of the population above 65 years old, predicts COVID-19 morbidity and mortality rates. These results suggest a susceptibility linked to low irradiance may play a large role in African American vulnerability to COVID-19, and that black populations in (darker) locations with lower irradiance may benefit from sunlight exposure during the COVID-19 pandemic.","author":[{"dropping-particle":"","family":"Backer","given":"Alex","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"title":"Why COVID-19 May Be Disproportionately Killing African Americans: Black Overrepresentation among COVID-19 Mortality Increases with Lower Irradiance, Where Ethnicity Is More Predictive of COVID-19 Infection and Mortality Than Median Income","type":"article-journal"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.1016/j.scitotenv.2020.138884","ISSN":"18791026","PMID":"32335404","abstract":"During the first 90 days of the COVID-19 outbreak in the United States, over 675,000 confirmed cases of the disease have been reported, posing unprecedented socioeconomic burden to the country. Due to inadequate research on geographic modeling of COVID-19, we investigated county-level variations of disease incidence across the continental United States. We compiled a geodatabase of 35 environmental, socioeconomic, topographic, and demographic variables that could explain the spatial variability of disease incidence. Further, we employed spatial lag and spatial error models to investigate spatial dependence and geographically weighted regression (GWR) and multiscale GWR (MGWR) models to locally examine spatial non-stationarity. The results suggested that even though incorporating spatial autocorrelation could significantly improve the performance of the global ordinary least square model, these models still represent a significantly poor performance compared to the local models. Moreover, MGWR could explain the highest variations (adj. R2: 68.1%) with the lowest AICc compared to the others. Mapping the effects of significant explanatory variables (i.e., income inequality, median household income, the proportion of black females, and the proportion of nurse practitioners) on spatial variability of COVID-19 incidence rates using MGWR could provide useful insights to policymakers for targeted interventions.","author":[{"dropping-particle":"","family":"Mollalo","given":"Abolfazl","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Vahedi","given":"Behzad","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rivera","given":"Kiara M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-3","issued":{"date-parts":[["2020"]]},"title":"GIS-based spatial modeling of COVID-19 incidence rate in the continental United States","type":"article-journal","volume":"728"},"uris":[""]},{"id":"ITEM-4","itemData":{"DOI":"10.1101/2020.05.06.20093195","abstract":"Introduction With the pandemic of COVID-19, the number of confirmed cases and related deaths are increasing in the US. We aimed to understand the potential impact of health and demographic factors on the infection and mortality rates of COVID-19 at the population level. Methods We collected total number of confirmed cases and deaths related to COVID-19 at the county level in the US from January 21, 2020 to April 23, 2020. We extracted health and demographic measures for each US county. Multivariable linear mixed effects models were used to investigate potential correlations of health and demographic characteristics with the infection and mortality rates of COVID-19 in US counties. Results Our models showed that several health and demographic factors were positively correlated with the infection rate of COVID-19, such as low education level and percentage of Black. In contrast, several factors, including percentage of smokers and percentage of food insecure, were negatively correlated with the infection rate of COVID-19. While the number of days since first confirmed case and the infection rate of COVID-19 were negatively correlated with the mortality rate of COVID-19, percentage of elders (65 and above) and percentage of rural were positively correlated with the mortality rate of COVID-19. Conclusions At the population level, health and demographic factors could impact the infection and mortality rates of COVID-19 in US counties.","author":[{"dropping-particle":"","family":"Xie","given":"Zidian","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Li","given":"Dongmei","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Medrxiv","id":"ITEM-4","issued":{"date-parts":[["2020"]]},"number-of-pages":"1-14","title":"Health and Demographic Impact on COVID-19 Infection and Mortality in US Counties","type":"report"},"uris":[""]}],"mendeley":{"formattedCitation":"(13,18–20)","plainTextFormattedCitation":"(13,18–20)","previouslyFormattedCitation":"(13,18–20)"},"properties":{"noteIndex":0},"schema":""}(13,18–20). Regarding health indicators, earlier research found that smokers, obese and individuals with existing health conditions are more likely to be severely affected by COVID-19 ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.scitotenv.2020.138884","ISSN":"18791026","PMID":"32335404","abstract":"During the first 90 days of the COVID-19 outbreak in the United States, over 675,000 confirmed cases of the disease have been reported, posing unprecedented socioeconomic burden to the country. Due to inadequate research on geographic modeling of COVID-19, we investigated county-level variations of disease incidence across the continental United States. We compiled a geodatabase of 35 environmental, socioeconomic, topographic, and demographic variables that could explain the spatial variability of disease incidence. Further, we employed spatial lag and spatial error models to investigate spatial dependence and geographically weighted regression (GWR) and multiscale GWR (MGWR) models to locally examine spatial non-stationarity. The results suggested that even though incorporating spatial autocorrelation could significantly improve the performance of the global ordinary least square model, these models still represent a significantly poor performance compared to the local models. Moreover, MGWR could explain the highest variations (adj. R2: 68.1%) with the lowest AICc compared to the others. Mapping the effects of significant explanatory variables (i.e., income inequality, median household income, the proportion of black females, and the proportion of nurse practitioners) on spatial variability of COVID-19 incidence rates using MGWR could provide useful insights to policymakers for targeted interventions.","author":[{"dropping-particle":"","family":"Mollalo","given":"Abolfazl","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Vahedi","given":"Behzad","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rivera","given":"Kiara M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"GIS-based spatial modeling of COVID-19 incidence rate in the continental United States","type":"article-journal","volume":"728"},"uris":[""]}],"mendeley":{"formattedCitation":"(13)","plainTextFormattedCitation":"(13)","previouslyFormattedCitation":"(13)"},"properties":{"noteIndex":0},"schema":""}(13). In terms of mobility trends, studies showed that staying at home and effective mobility restriction measures significantly lower the COVID-19 transmission rate ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3561142","ISSN":"1556-5068","abstract":"Predicting the binding mode of flexible polypeptides to proteins is an important task that falls outside the domain of applicability of most small molecule and protein?protein docking tools. Here, we test the small molecule flexible ligand docking program Glide on a set of 19 non-α-helical peptides and systematically improve pose prediction accuracy by enhancing Glide sampling for flexible polypeptides. In addition, scoring of the poses was improved by post-processing with physics-based implicit solvent MM- GBSA calculations. Using the best RMSD among the top 10 scoring poses as a metric, the success rate (RMSD ≤ 2.0 ? for the interface backbone atoms) increased from 21% with default Glide SP settings to 58% with the enhanced peptide sampling and scoring protocol in the case of redocking to the native protein structure. This approaches the accuracy of the recently developed Rosetta FlexPepDock method (63% success for these 19 peptides) while being over 100 times faster. Cross-docking was performed for a subset of cases where an unbound receptor structure was available, and in that case, 40% of peptides were docked successfully. We analyze the results and find that the optimized polypeptide protocol is most accurate for extended peptides of limited size and number of formal charges, defining a domain of applicability for this approach.","author":[{"dropping-particle":"","family":"Berger","given":"David","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Herkenhoff","given":"Kyle","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mongey","given":"Simon","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"An SEIR Infectious Disease Model with Testing and Conditional Quarantine","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1101/2020.05.14.20101667","abstract":"Since the emerging of the &quot;novel coronavirus&quot; SARS-CoV-2 and the corresponding respiratory disease COVID-19, the virus has spread all over the world. In Europe, Germany is currently one of the most affected countries. In March 2020, a &quot;lockdown&quot; was established to contain the virus spread, including the closure of schools and child day care facilities as well as forced social distancing and bans of any public gathering. The present study attempts to analyze whether these governmental interventions had an impact on the declared aim of &quot;flattening the curve&quot;, referring to the epidemic curve of new infections. This analysis is conducted from a regional perspective. On the level of the 412 German counties, logistic growth models were estimated based on reported cases of infections, aiming at determining the regional growth rate of infections and the point of inflection where infection rates begin to decrease and the curve flattens. All German counties exceeded the peak of new infections between the beginning of March and the middle of April. In a large majority of German counties, the epidemic curve has flattened before the social ban was established (March 23). In a minority of counties, the peak was already exceeded before school closures. The growth rates of infections vary spatially depending on the time the virus emerged. Counties belonging to states which established an additional curfew show no significant improvement with respect to growth rates and mortality. On the contrary, growth rates and mortality are significantly higher in Bavaria compared to whole Germany. The results raise the question whether social ban measures and curfews really contributed to the curve flattening. Furthermore, mortality varies strongly across German counties, which can be attributed to infections of people belonging to the “risk group”, especially residents of retirement peting Interest StatementThe authors have declared no competing interest.Funding StatementNo external funding was received.Author DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with …","author":[{"dropping-particle":"","family":"Wieland","given":"Thomas","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"medRxiv","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"page":"2020.05.14.20101667","publisher":"Cold Spring Harbor Laboratory Press","title":"Flatten the Curve! Modeling SARS-CoV-2/COVID-19 Growth in Germany on the County Level","type":"article-journal"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.2139/ssrn.3585921","ISSN":"1556-5068","abstract":"This paper provides insights for policymakers to evaluate the impact of staying at home and lockdown policies by investigating possible links between individual mobility and the spread of the COVID-19 virus in Italy. By relying on the daily data, the empirical evidence suggests that an increase in the number of visits to public spaces such as workspaces, parks, retail areas, and the use of public transportation is associated with an increase in the positive COVID-19 cases in a subsequent week. On the contrary, the increased intensity of staying in residential spaces is related to a decrease in the confirmed cases of COVID-19 significantly. Results are robust after controlling for the lockdown period. Empirical evidence underlines the importance of the lockdown decision. Further, there is substantial regional variation among the twenty regions of Italy. Individual presence in public vs. residential spaces has a more significant e?ect on the number of COVID-19 cases in the Lombardy region.","author":[{"dropping-particle":"","family":"Bilgin","given":"Nuriye Melisa","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-3","issued":{"date-parts":[["2020"]]},"title":"Tracing COVID-19 Spread in Italy with Mobility Data","type":"article-journal"},"uris":[""]},{"id":"ITEM-4","itemData":{"DOI":"10.2139/ssrn.3565703","ISSN":"1556-5068","abstract":"We combine GPS data on changes in average distance traveled by individuals at the county level with COVID-19 case data and other demographic information to estimate how individual mobility is affected by local disease prevalence and restriction orders to stay-at-home. We find that a rise of local infection rate from 0% to 0.003% 1 is associated with a reduction in mobility by 2.31%. An official stay-at-home restriction order corresponds to reducing mobility by 7.87%. Counties with larger shares of population over age 65, lower share of votes for the Republican Party in the 2016 Presidential Election, and higher population density are more responsive to disease prevalence and restriction orders.","author":[{"dropping-particle":"","family":"Engle","given":"Samuel","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Stromme","given":"John","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhou","given":"Anson","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-4","issued":{"date-parts":[["2020"]]},"title":"Staying at Home: Mobility Effects of COVID-19","type":"article-journal"},"uris":[""]},{"id":"ITEM-5","itemData":{"abstract":"In the absence of a vaccine or more effective treatment options, containing the spread of novel coronavirus disease 2019 (COVID-19) must rely on non-pharmaceutical interventions. All U.S. states adopted social-distancing measures in March and April of 2020, though they varied in both timing and scope. Kentucky began by closing public schools and restaurant dining rooms on March 16th before progressing to closing other non-essential businesses and eventually issuing a “Healthy at Home” order with restrictions similar to the shelter-in-place (SIPO) orders adopted by other states. We aim to quantify the impact of these measures on COVID-19 case growth in the state. An event-study model allows us to link adoption of social distancing measures across the Midwest and South to the growth rate of cases, allowing for effects to emerge gradually to account for the lag between infection and positive test result. We then use the results to predict how the number of cases would have evolved in Kentucky in the absence of these policy measures – in other words, if the state had relied on voluntary social distancing alone. We estimate that, by April 25, Kentucky would have had 44,482 confirmed COVID-19 cases without social distancing restrictions, as opposed to the 3,857 actually observed.","author":[{"dropping-particle":"","family":"Courtemanche","given":"Charles J","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Garuccio","given":"Joseph","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Le","given":"Anh","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Pinkston","given":"Joshua C","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Yelowitz","given":"Aaron","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Institute for the study of free enterprise working papers","id":"ITEM-5","issued":{"date-parts":[["2020"]]},"page":"1-23","title":"Did Social-Distancing Measures in Kentucky Help to Flatten the COVID-19 Curve?","type":"article-journal","volume":"4"},"uris":[""]},{"id":"ITEM-6","itemData":{"DOI":"10.1371/journal.pmed.1003166","ISSN":"15491676","PMID":"32692736","abstract":"Background The coronavirus disease (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has spread to nearly every country in the world since it first emerged in China in December 2019. Many countries have implemented social distancing as a measure to “flatten the curve” of the ongoing epidemics. Evaluation of the impact of government-imposed social distancing and of other measures to control further spread of COVID-19 is urgent, especially because of the large societal and economic impact of the former. The aim of this study was to compare the individual and combined effectiveness of self-imposed prevention measures and of short-term government-imposed social distancing in mitigating, delaying, or preventing a COVID-19 epidemic. Methods and findings We developed a deterministic compartmental transmission model of SARS-CoV-2 in a population stratified by disease status (susceptible, exposed, infectious with mild or severe disease, diagnosed, and recovered) and disease awareness status (aware and unaware) due to the spread of COVID-19. Self-imposed measures were assumed to be taken by disease-aware individuals and included handwashing, mask-wearing, and social distancing. Government-imposed social distancing reduced the contact rate of individuals irrespective of their disease or awareness status. The model was parameterized using current best estimates of key epidemiological parameters from COVID-19 clinical studies. The model outcomes included the peak number of diagnoses, attack rate, and time until the peak number of diagnoses. For fast awareness spread in the population, self-imposed measures can significantly reduce the attack rate and diminish and postpone the peak number of diagnoses. We estimate that a large epidemic can be prevented if the efficacy of these measures exceeds 50%. For slow awareness spread, self-imposed measures reduce the peak number of diagnoses and attack rate but do not affect the timing of the peak. Early implementation of short-term government-imposed social distancing alone is estimated to delay (by at most 7 months for a 3-month intervention) but not to reduce the peak. The delay can be even longer and the height of the peak can be additionally reduced if this intervention is combined with self-imposed measures that are continued after government-imposed social distancing has been lifted. Our analyses are limited in that they do not account for stochasticity, demographics, heterogeneities …","author":[{"dropping-particle":"","family":"Teslya","given":"Alexandra","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Pham","given":"Thi Mui","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Godijk","given":"Noortje G.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Kretzschmar","given":"Mirjam E.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Bootsma","given":"Martin C.J.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rozhnova","given":"Ganna","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"PLoS Medicine","id":"ITEM-6","issue":"7","issued":{"date-parts":[["2020"]]},"page":"e1003166","title":"Impact of self-imposed prevention measures and short-term government-imposed social distancing on mitigating and delaying a COVID-19 epidemic: A modelling study","type":"article-journal","volume":"17"},"uris":[""]},{"id":"ITEM-7","itemData":{"DOI":"10.1371/JOURNAL.PMED.1003244","ISSN":"15491676","PMID":"32780772","abstract":"Background Social distancing measures to address the US coronavirus disease 2019 (COVID-19) epidemic may have notable health and social impacts. Methods and findings We conducted a longitudinal pretest–posttest comparison group study to estimate the change in COVID-19 case growth before versus after implementation of statewide social distancing measures in the US. The primary exposure was time before (14 days prior to, and through 3 days after) versus after (beginning 4 days after, to up to 21 days after) implementation of the first statewide social distancing measures. Statewide restrictions on internal movement were examined as a secondary exposure. The primary outcome was the COVID-19 case growth rate. The secondary outcome was the COVID-19-attributed mortality growth rate. All states initiated social distancing measures between March 10 and March 25, 2020. The mean daily COVID-19 case growth rate decreased beginning 4 days after implementation of the first statewide social distancing measures, by 0.9% per day (95% CI ?1.4% to ?0.4%; P < 0.001). We did not observe a statistically significant difference in the mean daily case growth rate before versus after implementation of statewide restrictions on internal movement (0.1% per day; 95% CI ?0.04% to 0.3%; P = 0.14), but there is substantial difficulty in disentangling the unique associations with statewide restrictions on internal movement from the unique associations with the first social distancing measures. Beginning 7 days after social distancing, the COVID-19-attributed mortality growth rate decreased by 2.0% per day (95% CI ?3.0% to ?0.9%; P < 0.001). Our analysis is susceptible to potential bias resulting from the aggregate nature of the ecological data, potential confounding by contemporaneous changes (e.g., increases in testing), and potential underestimation of social distancing due to spillover effects from neighboring states. Conclusions Statewide social distancing measures were associated with a decrease in the COVID-19 case growth rate that was statistically significant. Statewide social distancing measures were also associated with a decrease in the COVID-19-attributed mortality growth rate beginning 7 days after implementation, although this decrease was no longer statistically significant by 10 days.","author":[{"dropping-particle":"","family":"Siedner","given":"Mark J.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Harling","given":"Guy","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Reynolds","given":"Zahra","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Gilbert","given":"Rebecca F.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Haneuse","given":"Sebastien","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Venkataramani","given":"Atheendar S.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Tsai","given":"Alexander C.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"PLoS Medicine","id":"ITEM-7","issue":"8 August","issued":{"date-parts":[["2020"]]},"page":"2020.04.03.20052373","publisher":"Cold Spring Harbor Laboratory Press","title":"Social distancing to slow the US COVID-19 epidemic: Longitudinal pretest–posttest comparison group study","type":"article-journal","volume":"17"},"uris":[""]}],"mendeley":{"formattedCitation":"(6,9,12,16,21–23)","plainTextFormattedCitation":"(6,9,12,16,21–23)","previouslyFormattedCitation":"(6,9,12,16,21–23)"},"properties":{"noteIndex":0},"schema":""}(6,9,12,16,21–23) while increased mobility resulted in increased COVID-19 transmissionADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/S1473-3099(20)30553-3","ISSN":"14744457","PMID":"32621869","abstract":"Background: Within 4 months of COVID-19 first being reported in the USA, it spread to every state and to more than 90% of all counties. During this period, the US COVID-19 response was highly decentralised, with stay-at-home directives issued by state and local officials, subject to varying levels of enforcement. The absence of a centralised policy and timeline combined with the complex dynamics of human mobility and the variable intensity of local outbreaks makes assessing the effect of large-scale social distancing on COVID-19 transmission in the USA a challenge. Methods: We used daily mobility data derived from aggregated and anonymised cell (mobile) phone data, provided by Teralytics (Zürich, Switzerland) from Jan 1 to April 20, 2020, to capture real-time trends in movement patterns for each US county, and used these data to generate a social distancing metric. We used epidemiological data to compute the COVID-19 growth rate ratio for a given county on a given day. Using these metrics, we evaluated how social distancing, measured by the relative change in mobility, affected the rate of new infections in the 25 counties in the USA with the highest number of confirmed cases on April 16, 2020, by fitting a statistical model for each county. Findings: Our analysis revealed that mobility patterns are strongly correlated with decreased COVID-19 case growth rates for the most affected counties in the USA, with Pearson correlation coefficients above 0·7 for 20 of the 25 counties evaluated. Additionally, the effect of changes in mobility patterns, which dropped by 35–63% relative to the normal conditions, on COVID-19 transmission are not likely to be perceptible for 9–12 days, and potentially up to 3 weeks, which is consistent with the incubation time of severe acute respiratory syndrome coronavirus 2 plus additional time for reporting. We also show evidence that behavioural changes were already underway in many US counties days to weeks before state-level or local-level stay-at-home policies were implemented, implying that individuals anticipated public health directives where social distancing was adopted, despite a mixed political message. Interpretation: This study strongly supports a role of social distancing as an effective way to mitigate COVID-19 transmission in the USA. Until a COVID-19 vaccine is widely available, social distancing will remain one of the primary measures to combat disease spread, and these findings should serve to support more time…","author":[{"dropping-particle":"","family":"Badr","given":"Hamada S.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Du","given":"Hongru","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Marshall","given":"Maximilian","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Dong","given":"Ensheng","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Squire","given":"Marietta M.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Gardner","given":"Lauren M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The Lancet Infectious Diseases","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Association between mobility patterns and COVID-19 transmission in the USA: a mathematical modelling study","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.31235/osf.io/hzj7a","abstract":"To what degree does social distancing have a causal effect on the spread of SARS-CoV-2? To generate causal evidence, we show that week to week changes in weather conditions provided a natural experiment that altered daily travel and movement outside the home, and thus affected social distancing in the first several weeks when Covid-19 began to spread in many U.S. counties. Using aggregated mobile phone location data and leveraging changes in social distancing driven by weekly weather conditions, we provide the first causal evidence on the effect of social distancing on the spread of SARS-CoV-2. Results show that a 1 percent increase in distance traveled leads to an 8.1 percent increase in new cases per capita in the following week, and a 1 percent increase in non-essential visits leads to a 6.9 percent increase in new cases per capita in the following week. Results are stronger in densely populated counties and close to zero in less densely populated counties.","author":[{"dropping-particle":"","family":"Sharkey","given":"Patrick","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wood","given":"George","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-2","issued":{"date-parts":[["2020"]]},"publisher":"SocArXiv","title":"The Causal Effect of Social Distancing on the Spread of SARS-CoV-2","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(14,24)","plainTextFormattedCitation":"(14,24)","previouslyFormattedCitation":"(14,24)"},"properties":{"noteIndex":0},"schema":""}(14,24). Finally, among health care infrastructure attributes, testing rate is linked with reduced risk of COVID-19 transmission ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3561142","ISSN":"1556-5068","abstract":"Predicting the binding mode of flexible polypeptides to proteins is an important task that falls outside the domain of applicability of most small molecule and protein?protein docking tools. Here, we test the small molecule flexible ligand docking program Glide on a set of 19 non-α-helical peptides and systematically improve pose prediction accuracy by enhancing Glide sampling for flexible polypeptides. In addition, scoring of the poses was improved by post-processing with physics-based implicit solvent MM- GBSA calculations. Using the best RMSD among the top 10 scoring poses as a metric, the success rate (RMSD ≤ 2.0 ? for the interface backbone atoms) increased from 21% with default Glide SP settings to 58% with the enhanced peptide sampling and scoring protocol in the case of redocking to the native protein structure. This approaches the accuracy of the recently developed Rosetta FlexPepDock method (63% success for these 19 peptides) while being over 100 times faster. Cross-docking was performed for a subset of cases where an unbound receptor structure was available, and in that case, 40% of peptides were docked successfully. We analyze the results and find that the optimized polypeptide protocol is most accurate for extended peptides of limited size and number of formal charges, defining a domain of applicability for this approach.","author":[{"dropping-particle":"","family":"Berger","given":"David","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Herkenhoff","given":"Kyle","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mongey","given":"Simon","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"An SEIR Infectious Disease Model with Testing and Conditional Quarantine","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1016/j.ijid.2020.04.021","ISSN":"18783511","PMID":"32320809","abstract":"Since the novel coronavirus disease (COVID-19) emerged in December 2019 in China, it has rapidly spread around the world, leading to one of the most significant pandemic events of recent history. Deriving reliable estimates of the COVID-19 epidemic growth rate is quite important to guide the timing and intensity of intervention strategies. Indeed, many studies have quantified the epidemic growth rate using time-series of reported cases during the early phase of the outbreak to estimate the basic reproduction number, R0. Using daily time series of COVID-19 incidence, we illustrate how epidemic curves of reported cases may not always reflect the true epidemic growth rate due to changes in testing rates, which could be influenced by limited diagnostic testing capacity during the early epidemic phase.","author":[{"dropping-particle":"","family":"Omori","given":"Ryosuke","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mizumoto","given":"Kenji","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Chowell","given":"Gerardo","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"International Journal of Infectious Diseases","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"page":"116-118","title":"Changes in testing rates could mask the novel coronavirus disease (COVID-19) growth rate","type":"article-journal","volume":"94"},"uris":[""]}],"mendeley":{"formattedCitation":"(21,25)","plainTextFormattedCitation":"(21,25)","previouslyFormattedCitation":"(21,25)"},"properties":{"noteIndex":0},"schema":""}(21,25). While earlier research efforts have considered the factors from all variable categories, it is important to recognize that each individual study focused only on a subset of variable groups (predominantly one or two) and have not controlled explicitly for other variable groups that can contribute to the COVID-19 transmission/mortality rate. The current study builds on earlier literature examining the factors affecting COVID-19 transmission and mortality rate and contributes along the following directions. First, we extensively enhance the spatial and temporal coverage of COVID-19 data in our analysis. Spatially, earlier research on COVID-19 aggregate data analysis has focused on a small number of counties (up to 100 counties). In our study, we consider all counties with total number of cases greater than 100 on August 4th. The 1,752 counties selected encompass 93% of the total population and 95% of the total confirmed COVID-19 cases. Temporally, earlier research has only considered data up to the month of April. While these studies are informative, cases in the US grew substantially in the recent months. Hence, in our study we have considered data from March 25th to August 4th, 2020. The longer period of data (133 days) also enables us to study/test for the evolution of variable effects over time. Second, earlier research studies have considered factors from one or two of the categories of variables identified above. Further, studies that tested health indicators employed one or two measures selectively. In our analysis, we conduct a comprehensive examination of factors affecting COVID-19 from all four categories of variables including (a) socio-demographics: distribution by age, gender, race, income, location (urban or rural), education status, income inequality and employment, (b) health indicators: percentage of population suffering from cancer, cardiovascular disease, hepatitis, Chronic Obstructive Pulmonary Disease (COPD); diabetes, obesity, Human Immunodeficiency Virus (HIV), heart disease, kidney disease, asthma; drinking and smoking habits, (c) mobility trends: daily average exposure, social distancing matrices, percentage of people staying at home, and (d) health care infrastructure attributes: hospitals per capita, ICU beds per capita, COVID-19 testing measures. Finally, the research study employs a robust modeling framework in terms of model structure and dependent variable representation. A mixed linear model system that addresses the limitations of the traditional linear regression framework for handling repeated measures is employed. For dependent variable, alternative functional forms of COVID-19 transmission – natural logarithm of daily cases per 100 thousand people and natural logarithm of 7-day moving average of cases per 100 thousand people - are considered in model estimation. The overall approach allows us to robustly quantify the impact of factors affecting COVID-19 transmission. MethodsData CollectionIndependent variables: Table 1 summarizes sample characteristics of the explanatory variables with the definition considered for final model estimation, the data source, and sample characteristics (minimum, maximum and mean values). The socio-demographic variables are collected from the American Community Survey (ACS) while information on the health indicator variables are gathered from the Centers for Disease Control and Prevention (CDC) systems. Using health indicator data, we ranked the 1,752 counties in a descending order of health metric and provided it in Fig 1. We performed ranking of the counties using multi-criteria decision analysis approach ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1111/j.1540-5915.1997.tb01306.x","ISSN":"00117315","abstract":"Often, data in multi-criteria decision making (MCDM) problems are imprecise and changeable. Therefore, an important step in many applications of MCDM is to perform a sensitivity analysis on the input data. This paper presents a methodology for performing a sensitivity analysis on the weights of the decision criteria and the performance values of the alternatives expressed in terms of the decision criteria. The proposed methodology is demonstrated on three widely used decision methods. These methods are the weighted sum model (WSM), the weighted product model (WPM), and the analytic hierarchy process (AHP). This paper formalizes a number of important issues on sensitivity analysis and derives some critical theoretical results. Also, a number of illustrative examples and computational experiments further illustrate the application of the proposed methodology.","author":[{"dropping-particle":"","family":"Triantaphyllou","given":"Evangelos","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Sánchez","given":"Alfonso","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Decision Sciences","id":"ITEM-1","issued":{"date-parts":[["1997"]]},"title":"A sensitivity analysis approach for some deterministic multi-criteria decision-making methods","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.14569/ijacsa.2017.080616","ISSN":"2158107X","abstract":"Wind energy is becoming a potential source for renewable and clean\nenergy. An important factor that contributes to efficient generation of\nwind power is the use of appropriate wind turbine. However, the task of\nselecting an appropriate, site-specific turbine is a complex problem.\nThe complexity is due to the presence of several conflicting decision\ncriteria in the decision process. Therefore, a decision is sought such\nthat best tradeoff is achieved between the selection criteria. With the\ninherent complexities encompassing the decision-making process, this\nstudy develops a multi-criteria decision model for turbine selection\nbased on the concepts of weighted sum approach. Results indicate that\nthe proposed methodology for finding the most suitable turbine from a\npool of 18 turbines is effective.","author":[{"dropping-particle":"","family":"Rehman","given":"Shafiqur","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"A.","given":"Salman","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"International Journal of Advanced Computer Science and Applications","id":"ITEM-2","issued":{"date-parts":[["2017"]]},"title":"Multi-Criteria Wind Turbine Selection using Weighted Sum Approach","type":"article-journal"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.1007/978-1-4471-2346-0_4","ISBN":"9781447123453","ISSN":"18653529","abstract":"In this chapter we look at two simple multi-criteria decision-making methods, the Weighted Sum method and the Weighted Product method. In the Weighted Sum method the score of an alternative is equal to the weighted sum of its evaluation ratings, where the weights are the importance weights associated with each attribute. In the Weighted Product method, instead of calculating sub-scores by multiplying performance scores times attribute importance, performance scores are raised to the power of the attribute importance weight.","author":[{"dropping-particle":"","family":"Mateo","given":"José Ramón San Cristóbal","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Green Energy and Technology","id":"ITEM-3","issued":{"date-parts":[["2012"]]},"title":"Weighted sum method and weighted product method","type":"paper-conference"},"uris":[""]}],"mendeley":{"formattedCitation":"(26–28)","plainTextFormattedCitation":"(26–28)","previouslyFormattedCitation":"(26–28)"},"properties":{"noteIndex":0},"schema":""}(26–28). Details on this approach are summarized in the supplementary materials. Further, we compute the average values for different health indicators across the healthiest and unhealthiest 10 counties to highlight the change in health conditions across the two groups. The values clearly emphasize the vulnerability of the unhealthiest counties relative to the healthiest counties. For instance, number of Cardio patients in the healthy counties are 28.44 while in the unhealthiest counties, it is almost 219% higher (90.69).Table 1 Descriptive Statistics of the Dependent and Independent VariablesVariablesSourceMeanMin/MaxSample SizeIndependent VariablesDemographic CharacteristicsPercentage of population aged 18 years and lowerACSa22.5587.155/35.9871752Percentage of population aged 65 years and overACS17.2566.724/56.9441752Percentage of African AmericanACS10.9940.113/80.5071752Percentage of HispanicACS10.3440.623/96.3231752Percentage of FemaleACS50.38637.041/54.4951752Ln (Median income)ACS10.87210.149/11.8221752Percentage of people less than high school educationACS14.1433.127/47.0531752Employment rate per capitaACS0.4410.190/0.6401752Income inequality ratio (80th percentile/20th percentile)CHRRb4.5472.988/9.1481752Health IndicatorsLn (HIV Prevalence Rate per 100K people)CHRR4.8700.723/7.8591752Hepatitis B Cases per 100K people in2017CDCc1.3380.000/11.7001752Hepatitis C Cases per 100K people in2017CDC1.0160.000/5.6001752Asthma % for >= 18 yearsCDC9.3327.400/12.3001752COPD % for >= 18 yearsCDC6.7573.300/13.7001752Reported cancer case per 100K peopleCDC455.651241.000/623.0001752Percentage of diabeticCHRR11.5273.300/20.4001752Percentage of obesity among adultsCHRR31.95113.600/46.7001752Cardiovascular Disease Hospitalization Rate per 1,000 Medicare BeneficiariesCDC63.4620.300/115.8001752Mobility TrendsLn (Daily Average Exposure), 10 days lagFrom April 25thCEId4.1760.591/7.048233,016% People staying at home14 days lagSafegraph0.1430.037/0.364233,016Healthcare Related AttributesHospitals per 100K peopleCHRR2.3720.000/15.6401752Number of ICU beds per capitaCHRR18.3340.000/171.8501752Ln (No of tests with?5 days lag)CTPe8.4310.000/12.0156,783Temporal FactorsDay is weekend--0.2850.000/1.000233,016Dependent VariablesLn (Daily COVID-19 transmission rate per 100K people)CSSEf1.4700.000/7.668233,016Ln (Total COVID-19 mortality rate per 100K people)CSSE2.8490.000/7.2371752a = American Community Surveyb = County Health Rankings & Roadmapsc= Central for Disease Control Systemd= COVID Exposure Indices (25)e= COVID-19 Tracking Project (26)f= Center for Systems Science and Engineering Coronavirus Resource Center at Johns Hopkins University (27)To incorporate mobility trends, we considered two variables: daily average exposure and social distancing metric to serve as a surrogate measure for the mobility patterns. The exposure variables provide information compiled based on smartphone movement data within and across the counties in US ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.3386/w27560","abstract":"Tracking human activity in real time and at fine spatial scale is particularly valuable during episodes such as the COVID-19 pandemic. In this paper, we discuss the suitability of smartphone data for quantifying movement and social contact. We show that these data cover broad sections of the US population and exhibit movement patterns similar to conventional survey data. We develop and make publicly available a location exposure index that summarizes county-to-county movements and a device exposure index that quantifies social contact within venues. We use these indices to document how pandemic-induced reductions in activity vary across people and places.","author":[{"dropping-particle":"","family":"Couture","given":"Victor","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Dingel","given":"Jonathan I","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Green","given":"Allison","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Handbury","given":"Jessie","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Williams","given":"Kevin R","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"NBER Working Paper Series","id":"ITEM-1","issue":"2241","issued":{"date-parts":[["2020"]]},"number":"35","title":"Measuring Movement and Social Contact With Smartphone Data : a Real-Time Application To Covid-19","type":"report","volume":"No. 27560"},"uris":[""]}],"mendeley":{"formattedCitation":"(30)","plainTextFormattedCitation":"(30)","previouslyFormattedCitation":"(30)"},"properties":{"noteIndex":0},"schema":""}(30). For our analysis, we confined our attention to the overlapping movements within the counties. From the movement data provided by PlaceIQ, for each smartphone device visiting a location, the total number of distinct devices visiting that location at that particular time is calculated ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.3386/w27560","abstract":"Tracking human activity in real time and at fine spatial scale is particularly valuable during episodes such as the COVID-19 pandemic. In this paper, we discuss the suitability of smartphone data for quantifying movement and social contact. We show that these data cover broad sections of the US population and exhibit movement patterns similar to conventional survey data. We develop and make publicly available a location exposure index that summarizes county-to-county movements and a device exposure index that quantifies social contact within venues. We use these indices to document how pandemic-induced reductions in activity vary across people and places.","author":[{"dropping-particle":"","family":"Couture","given":"Victor","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Dingel","given":"Jonathan I","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Green","given":"Allison","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Handbury","given":"Jessie","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Williams","given":"Kevin R","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"NBER Working Paper Series","id":"ITEM-1","issue":"2241","issued":{"date-parts":[["2020"]]},"number":"35","title":"Measuring Movement and Social Contact With Smartphone Data : a Real-Time Application To Covid-19","type":"report","volume":"No. 27560"},"uris":[""]}],"mendeley":{"formattedCitation":"(30)","plainTextFormattedCitation":"(30)","previouslyFormattedCitation":"(30)"},"properties":{"noteIndex":0},"schema":""}(30). These distinct devices will serve as exposure for the particular device. Similarly, one can compute the exposure for all the devices residing in a county and finally compute the daily average exposure at the count level. The reader would note that smartphone movement data is reported for counties with at least 1000 active devices in a day. The 1752 counties selected for analysis satisfied the requirement of minimum active devices. The second measure, information on social distancing is collected from Safegraph data (see Acknowledgement section for description of Safegraph data). These metrics provide information on the number of devices completely staying at home, mean/median distance travel from home, full time and part time work behavior at a daily basis for each county. Fig 2 provides a summary of both these measures at a state level from January 22nd to August 4th. From the figure, we can clearly see the reduction in average daily exposure in March as many states and local jurisdictions imposed lockdowns. By late April, exposure activity started to increase again across all the states while still being lower than the levels for February. In terms of the staying at home measure, as expected, we find an exactly opposite trend. Finally, within the healthcare infrastructure attributes, information about the hospitals and ICU beds are gathered from the County level health ranking data. COVID-19 testing measures are sourced from the COVID-19 tracking project ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","accessed":{"date-parts":[["2020","8","13"]]},"id":"ITEM-1","issued":{"date-parts":[["0"]]},"title":"Our Data: The COVID Tracking Project","type":"webpage"},"uris":[""]}],"mendeley":{"formattedCitation":"(31)","plainTextFormattedCitation":"(31)","previouslyFormattedCitation":"(31)"},"properties":{"noteIndex":0},"schema":""}(31) that provides a complete picture of testing as well the number of positive and negative cases for each county in the United States. Dependent variables: We analyze two county level dependent variables: (1) COVID-19 daily transmission rate per 100K population and (2) COVID-19 mortality rates per 100K population. For the transmission rate analysis, we tested two alternative functional forms: daily cases per 100 thousand people and 7-day moving average of cases per 100 thousand people. The moving average data is likely to be less volatile and serves as a stability test for the daily cases model. The reader would note that we used a natural logarithmic transformation for all the dependent variables. The COVID-19 dataset from Center for Systems Science and Engineering (CSSE) Coronavirus Resource Center at Johns Hopkins UniversityADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","abstract":"Coronavirus COVID-19 Global Cases by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU) Continously updated world map with number of confirmed cases (infections)","accessed":{"date-parts":[["2020","7","11"]]},"author":[{"dropping-particle":"","family":"Johns Hopkins Coronavirus Resource Center","given":"","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Johns Hopkins Coronavirus Resource Center","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"COVID-19 Map - Johns Hopkins Coronavirus Resource Center","type":"webpage"},"uris":[""]}],"mendeley":{"formattedCitation":"(32)","plainTextFormattedCitation":"(32)","previouslyFormattedCitation":"(32)"},"properties":{"noteIndex":0},"schema":""}(32) provides information on the daily confirmed COVID-19 cases, number of people recovered (when available) and the number of deaths from COVID-19 starting from January 22nd to the current date across 3,142 counties in the United States. In our research, we confined our analysis to the cases between March 25th to August 4th resulting in 133 days of data. Further, we focus on counties that have at least 100 cases by August 4th and have available information on the mobility trends. With this requirement, a total of 1,752 counties are included in the analysis providing a coverage of 93% of the total population in the United States. For mortality rate, we considered the fatalities within the same time frame across all the 1,752 counties as the transmission rate variable. The summary statistics of the dependent variable are presented in bottom row panel of Table 1. Data Analysis (Modeling Framework)The two dependent variables: (a) COVID-19 daily transmission rate and (b) COVID-19 mortality rate are continuous in nature and linear regression model is the most traditional method to study such continuous responses. For the analysis of daily transmission rate, we have repeated measures of the variable (133 repetitions for each county). The traditional linear regression model is not appropriate to study data with multiple repeated observations ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.jtrangeo.2014.01.013","ISSN":"09666923","abstract":"Installed in 2009, BIXI is the first major public bicycle-sharing system in Montreal, Canada. The BIXI system has been a success, accounting for more than one million trips annually. This success has increased the interest in exploring the factors affecting bicycle-sharing flows and usage. Using data compiled as minute-by-minute readings of bicycle availability at all the stations of the BIXI system between April and August 2012, this study contributes to the literature on bicycle-sharing. We examine the influence of meteorological data, temporal characteristics, bicycle infrastructure, land use and built environment attributes on arrival and departure flows at the station level using a multilevel approach to statistical modeling, which could easily be applied to other regions. The findings allow us to identify factors contributing to increased usage of bicycle-sharing in Montreal and to provide recommendations pertaining to station size and location decisions. The developed methodology and findings can be of benefit to city planners and engineers who are designing or modifying bicycle-sharing systems with the goal of maximizing usage and availability.","author":[{"dropping-particle":"","family":"Faghih-Imani","given":"Ahmadreza","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Eluru","given":"Naveen","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"El-Geneidy","given":"Ahmed M.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rabbat","given":"Michael","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Haq","given":"Usama","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Journal of Transport Geography","id":"ITEM-1","issued":{"date-parts":[["2014"]]},"page":"306-314","title":"How land-use and urban form impact bicycle flows: Evidence from the bicycle-sharing system (BIXI) in Montreal","type":"article-journal","volume":"41"},"uris":[""]}],"mendeley":{"formattedCitation":"(33)","plainTextFormattedCitation":"(33)","previouslyFormattedCitation":"(33)"},"properties":{"noteIndex":0},"schema":""}(33). Hence, we employ a linear mixed modeling approach that builds on the linear regression model while incorporating the influence of repeated observations from the same county. By adopting the linear mixed model, we recognize the dependencies across COVID-19 cases occurring for the same county. A brief description of the linear mixed model is provided below:Let q = 1, 2, …, Q be an index to represent each county, and d = 1, 2, …, D be an index to represent the various days on which data (cases) was collected. The general form of the mixed linear regression model has the following structure:yqd= βX+ εqd (1)where yqd is the dependent variable representing the new COVID 19 cases per 100K population, X is the vector of attributes and β is the model coefficients. εqd is the random error term assumed to be normally distributed across the dataset.This ε term captures the dependencies across the repetition for each county. In our analysis, we estimate the correlation for different level of repetition measures: correlation for all records (133 repetitions), monthly level (31 repetitions) and weekly level (7 repetitions). The flexibility offered by the mixed model for testing dependencies enhances the model development exercise over its simpler form. In this structure, the data can be visualized as K (K = 133 or 31 or 7) records for each 1,752 counties. Estimating a full covariance matrix (up to 133*133) is computationally intensive while providing very little intuition. Hence, we parameterize the covariance matrix (Ω). For estimating a parsimonious specification, we tested first-order autoregressive (AR) and autoregressive moving average (ARMA) correlation structure within the mixed linear model. The reader would note that the final model was identified based on three criteria: autocorrelation function (ACF); a partial autocorrelation function (PACF) and Bayesian Information Criterion metric (BIC). All of these measures provide support to the ARMA model selection (see Supplementary Materials for more details). Therefore, in the current study, we will only discuss the framework for the ARMA model (due to space constraints). The ARMA correlation structure comprises three parameters σ, ρ, and φ as follows: Ω= σ21φρ φρ1 φρ2?φρK-1????? φρK-1? ?????1 (2)where, σ represents the error variance of ε, φ represents the common correlation factor across time periods K, ρ represents the dampening parameter that reduces the correlation with time and K represents the level of repetition. The correlation parameters φ and ρ, if significant, highlight the impact of county effects on the dependent variables. The models are estimated in SPSS using the restricted maximum likelihood estimation (RMLE) approach. For modeling the COVID 19 mortality rate, we rely on simple linear regression approach as the dependent variable here is the total number of COVID-19 deaths per 100K population at a county level. ResultsThe reader would note that prior to estimating the models, we checked for the multicollinearity issue across the independent variables as it is possible that county level characteristics are highly correlated. We did not find any significant impact of multicollinearity on our model estimates (see Supplemental Materials for more details)COVID-19 Transmission Rate Model Results The estimation results for the linear mixed model are presented in Table 2. From this point, we will use the term transmission rate for representing the natural logarithm of daily COVID-19 cases per 100K population. As discussed earlier, we also developed the same mixed linear model to estimate the 7-day moving average of COVID-19 cases per capita and find similar results as in the daily COVID-19 transmission model (results are available upon request from the authors). This further reinforces the stability of the transmission model.Table 2 Estimation Results for Daily COVID-19 Transmission Rate per 100K PopulationVariablesEstimatest-statisticp-valueConstant-4.882-18.307<0.001Demographics% of Female population0.0198.794<0.001% Young population (<=18 years)0.0096.097<0.001% of African-American population0.01027.055<0.001% of People less than high school education0.02222.738<0.001Ln (median income)0.32514.185<0.001Employment rate per capita0.9639.320<0.001Ln (% of People living in rural areas)-0.408-17.567<0.001Health IndicatorsLn (HIV rate per 100K People)0.0447.441<0.001Hepatitis C rate per 100K People0.0123.200<0.001Mobility TrendsLn (Daily Average Exposure), 10 days lagApril 25th to July 21st 0.02812.360<0.001After July 21st 0.17117.085<0.001% People staying at home (14 days lag)March 25th to July 21st -0.590-5.564<0.001After July 21st -3.952-13.023<0.001Health Care Infrastructure AttributesLn (Testing), 5 days lag?March 25th to May 10th0.0127.654<0.001After May 10th0.01915.350<0.001Temporal FactorsTemporal Lagged Variables7 days lag (March 25th to June 22nd)0.17769.165<0.0017 days lag (June 23rd to July 6th)0.28566.121<0.0017 days lag (After July 6th)0.362115.590<0.00114 days lag0.16777.272<0.001Day is Weekend-0.045-10.695<0.001Correlation σ0.988275.252<0.001 ρ0.959367.088<0.001 Φ0.286102.854<0.001Socio-demographics: We find several socio-demographic variables to have significant impact on the transmission rate. In terms of female population, we find that higher proportion of females in the population has a positive impact on transmission rated. At first glance , the result might appear to be contradicting earlier studies that show women are less likely to be affected by COVID-19 transmission relative to men ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3572329","ISSN":"1556-5068","abstract":"New York City is the hot spot of the Covid-19 pandemic in the United States. This paper merges information on the number of tests and the number of infections at the New York City zip code level with demographic and socioeconomic information from the decennial census and the American Community Surveys. People residing in poor or immigrant neighbourhoods were less likely to be tested; but the likelihood that a test was positive was larger in those neighbourhoods, as well as in neighbourhoods with larger households or predominantly black populations. The rate of infection in the population depends on both the frequency of tests and on the fraction of positive tests among those tested. The non-randomness in testing across New York City neighbourhoods indicates that the observed correlation between the rate of infection and the socioeconomic characteristics of a community tells an incomplete story of how the pandemic evolved in a congested urban setting.","author":[{"dropping-particle":"","family":"Borjas","given":"George J.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Demographic Determinants of Testing Incidence and COVID-19 Infections in New York City Neighborhoods","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(18)","plainTextFormattedCitation":"(18)","previouslyFormattedCitation":"(18)"},"properties":{"noteIndex":0},"schema":""}(18). However, the reader would note that this result only implies that counties with higher percentage of female population are likely to experience increased number of COVID-19 cases relative to other counties. The finding does not necessarily indicate that women are at a higher risk of being infected by COVID-19. For differences in proclivity for COVID-19 infection by gender, individual level data would be a more appropriate avenue for analysis. Among age distribution proportions, we found that increased percentage of younger individuals (<18 years) is associated with more transmission. In terms of racial distributions, counties with higher proportion of African-Americans are likely to have higher transmission rates (see earlier work for similar findings ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.scitotenv.2020.138884","ISSN":"18791026","PMID":"32335404","abstract":"During the first 90 days of the COVID-19 outbreak in the United States, over 675,000 confirmed cases of the disease have been reported, posing unprecedented socioeconomic burden to the country. Due to inadequate research on geographic modeling of COVID-19, we investigated county-level variations of disease incidence across the continental United States. We compiled a geodatabase of 35 environmental, socioeconomic, topographic, and demographic variables that could explain the spatial variability of disease incidence. Further, we employed spatial lag and spatial error models to investigate spatial dependence and geographically weighted regression (GWR) and multiscale GWR (MGWR) models to locally examine spatial non-stationarity. The results suggested that even though incorporating spatial autocorrelation could significantly improve the performance of the global ordinary least square model, these models still represent a significantly poor performance compared to the local models. Moreover, MGWR could explain the highest variations (adj. R2: 68.1%) with the lowest AICc compared to the others. Mapping the effects of significant explanatory variables (i.e., income inequality, median household income, the proportion of black females, and the proportion of nurse practitioners) on spatial variability of COVID-19 incidence rates using MGWR could provide useful insights to policymakers for targeted interventions.","author":[{"dropping-particle":"","family":"Mollalo","given":"Abolfazl","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Vahedi","given":"Behzad","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rivera","given":"Kiara M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"GIS-based spatial modeling of COVID-19 incidence rate in the continental United States","type":"article-journal","volume":"728"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1101/2020.05.06.20093195","abstract":"Introduction With the pandemic of COVID-19, the number of confirmed cases and related deaths are increasing in the US. We aimed to understand the potential impact of health and demographic factors on the infection and mortality rates of COVID-19 at the population level. Methods We collected total number of confirmed cases and deaths related to COVID-19 at the county level in the US from January 21, 2020 to April 23, 2020. We extracted health and demographic measures for each US county. Multivariable linear mixed effects models were used to investigate potential correlations of health and demographic characteristics with the infection and mortality rates of COVID-19 in US counties. Results Our models showed that several health and demographic factors were positively correlated with the infection rate of COVID-19, such as low education level and percentage of Black. In contrast, several factors, including percentage of smokers and percentage of food insecure, were negatively correlated with the infection rate of COVID-19. While the number of days since first confirmed case and the infection rate of COVID-19 were negatively correlated with the mortality rate of COVID-19, percentage of elders (65 and above) and percentage of rural were positively correlated with the mortality rate of COVID-19. Conclusions At the population level, health and demographic factors could impact the infection and mortality rates of COVID-19 in US counties.","author":[{"dropping-particle":"","family":"Xie","given":"Zidian","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Li","given":"Dongmei","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Medrxiv","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"number-of-pages":"1-14","title":"Health and Demographic Impact on COVID-19 Infection and Mortality in US Counties","type":"report"},"uris":[""]}],"mendeley":{"formattedCitation":"(13,20)","plainTextFormattedCitation":"(13,20)","previouslyFormattedCitation":"(13,20)"},"properties":{"noteIndex":0},"schema":""}(13,20)). It has been postulated that African-Americans in general reside in densely populated low income neighborhoods with lower access to amenities and are employed in industries that requires more public exposure ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3571699","ISSN":"1556-5068","abstract":"COVID-19 is killing African Americans at a rate 7% to 193% higher than the general population. Understanding why, as well as the reasons behind the wide variation, is paramount to saving lives. Here, we test two potential explanations for this effect. On the one hand, African Americans might be dying more because they have a lower average income (‘the Socioeconomic Hypothesis’). On the other hand, they might be dying more because their skin is more resistant to UV radiation, as we previously showed that COVID-19 infections and deaths decrease with higher irradiance (‘the Irradiance Hypothesis’). The two hypotheses are not mutually exclusive. We show that the overrepresentation of African Americans among COVID-19 deaths shows a significant negative correlation with mean solar irradiance, with a 20% decrease in Global Horizontal Irradiance leading to a 76% increase in the overrepresentation of African Americans amongst COVID-19 deaths. We then show that in Michigan, one of the US states with the lowest irradiance in early April, the % of each county’s population that is black, more than its median income, median age or % of the population above 65 years old, predicts COVID-19 morbidity and mortality rates. These results suggest a susceptibility linked to low irradiance may play a large role in African American vulnerability to COVID-19, and that black populations in (darker) locations with lower irradiance may benefit from sunlight exposure during the COVID-19 pandemic.","author":[{"dropping-particle":"","family":"Backer","given":"Alex","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Why COVID-19 May Be Disproportionately Killing African Americans: Black Overrepresentation among COVID-19 Mortality Increases with Lower Irradiance, Where Ethnicity Is More Predictive of COVID-19 Infection and Mortality Than Median Income","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(19)","plainTextFormattedCitation":"(19)","previouslyFormattedCitation":"(19)"},"properties":{"noteIndex":0},"schema":""}(19). Educational status in a county also plays an important role in influencing the COVID-19 transmission. The counties with higher share of individuals with less than high school education are likely to report increased incidence of COVID-19. In terms of income, we find that higher median income in a county leads to rise in daily COVID-19 incidence. The effect of income might appear counter-intuitive at first glance. However, it is possible that higher income individuals are more likely to get tested (even in the absence of symptoms) due to higher health insurance affordability. Low income individuals are more likely to lose their jobs and health insurance coverage due to COVID-19 pandemic ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.scitotenv.2020.138884","ISSN":"18791026","PMID":"32335404","abstract":"During the first 90 days of the COVID-19 outbreak in the United States, over 675,000 confirmed cases of the disease have been reported, posing unprecedented socioeconomic burden to the country. Due to inadequate research on geographic modeling of COVID-19, we investigated county-level variations of disease incidence across the continental United States. We compiled a geodatabase of 35 environmental, socioeconomic, topographic, and demographic variables that could explain the spatial variability of disease incidence. Further, we employed spatial lag and spatial error models to investigate spatial dependence and geographically weighted regression (GWR) and multiscale GWR (MGWR) models to locally examine spatial non-stationarity. The results suggested that even though incorporating spatial autocorrelation could significantly improve the performance of the global ordinary least square model, these models still represent a significantly poor performance compared to the local models. Moreover, MGWR could explain the highest variations (adj. R2: 68.1%) with the lowest AICc compared to the others. Mapping the effects of significant explanatory variables (i.e., income inequality, median household income, the proportion of black females, and the proportion of nurse practitioners) on spatial variability of COVID-19 incidence rates using MGWR could provide useful insights to policymakers for targeted interventions.","author":[{"dropping-particle":"","family":"Mollalo","given":"Abolfazl","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Vahedi","given":"Behzad","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Rivera","given":"Kiara M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Science of the Total Environment","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"GIS-based spatial modeling of COVID-19 incidence rate in the continental United States","type":"article-journal","volume":"728"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1016/S2468-2667(20)30085-2","ISSN":"24682667","PMID":"32247329","author":[{"dropping-particle":"","family":"Ahmed","given":"Faheem","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Ahmed","given":"Na'eem","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Pissarides","given":"Christopher","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Stiglitz","given":"Joseph","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The Lancet Public Health","id":"ITEM-2","issue":"5","issued":{"date-parts":[["2020"]]},"page":"e240","title":"Why inequality could spread COVID-19","type":"article","volume":"5"},"uris":[""]}],"mendeley":{"formattedCitation":"(13,34)","plainTextFormattedCitation":"(13,34)","previouslyFormattedCitation":"(13,34)"},"properties":{"noteIndex":0},"schema":""}(13,34). With respect to employment rate, counties with higher employment rate reflect more exposure and have a positive association with transmission. The percentage of people living in rural area offers a negative association with the daily COVID-19 incidence. This indicates that people living in rural areas are less affected by COVID-19. This is intuitive as rural areas are sparsely populated and hence have more opportunity for social distancing thus lowering transmission rates. Health indicators: With respect to health indicators, we tried several variables in the transmission rate model. Of these, two variables number of people suffering from HIV and hepatitis C in a county offered significant impacts. We observe that counties with higher percentage of HIV and hepatitis C patients have an increased incidence of COVID-19 transmission. Individuals with these diseases have weaker immune systems and hence are more susceptible to COVID-19 transmission. Mobility Trends: In terms of mobility trends, we tested two measures: daily average exposure and percentage of people staying at home. In considering these variables in the model, we recognize that exposure will have a lagged effect on transmission i.e. exposure to virus today is likely to manifest as a case in the next 5 to 14 days. In our analysis, we tested several lag combinations and selected the 10 day lag exposure as it offered the best fit. Similarly, for people staying at home, the 14 day lag offered the best fit. The exposure variable offers interesting results. Until April 25th exposure variable does not have any impact on transmission. This is strongly coinciding with the lower exposure trends (see Fig 2). After April 25th, increased exposure is associated with higher transmission rates 10 days into the future (see Hamada and colleagues ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/S1473-3099(20)30553-3","ISSN":"14744457","PMID":"32621869","abstract":"Background: Within 4 months of COVID-19 first being reported in the USA, it spread to every state and to more than 90% of all counties. During this period, the US COVID-19 response was highly decentralised, with stay-at-home directives issued by state and local officials, subject to varying levels of enforcement. The absence of a centralised policy and timeline combined with the complex dynamics of human mobility and the variable intensity of local outbreaks makes assessing the effect of large-scale social distancing on COVID-19 transmission in the USA a challenge. Methods: We used daily mobility data derived from aggregated and anonymised cell (mobile) phone data, provided by Teralytics (Zürich, Switzerland) from Jan 1 to April 20, 2020, to capture real-time trends in movement patterns for each US county, and used these data to generate a social distancing metric. We used epidemiological data to compute the COVID-19 growth rate ratio for a given county on a given day. Using these metrics, we evaluated how social distancing, measured by the relative change in mobility, affected the rate of new infections in the 25 counties in the USA with the highest number of confirmed cases on April 16, 2020, by fitting a statistical model for each county. Findings: Our analysis revealed that mobility patterns are strongly correlated with decreased COVID-19 case growth rates for the most affected counties in the USA, with Pearson correlation coefficients above 0·7 for 20 of the 25 counties evaluated. Additionally, the effect of changes in mobility patterns, which dropped by 35–63% relative to the normal conditions, on COVID-19 transmission are not likely to be perceptible for 9–12 days, and potentially up to 3 weeks, which is consistent with the incubation time of severe acute respiratory syndrome coronavirus 2 plus additional time for reporting. We also show evidence that behavioural changes were already underway in many US counties days to weeks before state-level or local-level stay-at-home policies were implemented, implying that individuals anticipated public health directives where social distancing was adopted, despite a mixed political message. Interpretation: This study strongly supports a role of social distancing as an effective way to mitigate COVID-19 transmission in the USA. Until a COVID-19 vaccine is widely available, social distancing will remain one of the primary measures to combat disease spread, and these findings should serve to support more time…","author":[{"dropping-particle":"","family":"Badr","given":"Hamada S.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Du","given":"Hongru","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Marshall","given":"Maximilian","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Dong","given":"Ensheng","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Squire","given":"Marietta M.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Gardner","given":"Lauren M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The Lancet Infectious Diseases","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Association between mobility patterns and COVID-19 transmission in the USA: a mathematical modelling study","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(24)","plainTextFormattedCitation":"(24)","previouslyFormattedCitation":"(24)"},"properties":{"noteIndex":0},"schema":""}(24) for similar findings). Further, the influence of exposure is substantially larger after July 21st indicating a higher risk of exposure for COVID-19 transmission. For the second measure, staying at home with 14 days lag, we find that daily transmission rates are negatively affected as expected ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3561142","ISSN":"1556-5068","abstract":"Predicting the binding mode of flexible polypeptides to proteins is an important task that falls outside the domain of applicability of most small molecule and protein?protein docking tools. Here, we test the small molecule flexible ligand docking program Glide on a set of 19 non-α-helical peptides and systematically improve pose prediction accuracy by enhancing Glide sampling for flexible polypeptides. In addition, scoring of the poses was improved by post-processing with physics-based implicit solvent MM- GBSA calculations. Using the best RMSD among the top 10 scoring poses as a metric, the success rate (RMSD ≤ 2.0 ? for the interface backbone atoms) increased from 21% with default Glide SP settings to 58% with the enhanced peptide sampling and scoring protocol in the case of redocking to the native protein structure. This approaches the accuracy of the recently developed Rosetta FlexPepDock method (63% success for these 19 peptides) while being over 100 times faster. Cross-docking was performed for a subset of cases where an unbound receptor structure was available, and in that case, 40% of peptides were docked successfully. We analyze the results and find that the optimized polypeptide protocol is most accurate for extended peptides of limited size and number of formal charges, defining a domain of applicability for this approach.","author":[{"dropping-particle":"","family":"Berger","given":"David","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Herkenhoff","given":"Kyle","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mongey","given":"Simon","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"An SEIR Infectious Disease Model with Testing and Conditional Quarantine","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.2139/ssrn.3585921","ISSN":"1556-5068","abstract":"This paper provides insights for policymakers to evaluate the impact of staying at home and lockdown policies by investigating possible links between individual mobility and the spread of the COVID-19 virus in Italy. By relying on the daily data, the empirical evidence suggests that an increase in the number of visits to public spaces such as workspaces, parks, retail areas, and the use of public transportation is associated with an increase in the positive COVID-19 cases in a subsequent week. On the contrary, the increased intensity of staying in residential spaces is related to a decrease in the confirmed cases of COVID-19 significantly. Results are robust after controlling for the lockdown period. Empirical evidence underlines the importance of the lockdown decision. Further, there is substantial regional variation among the twenty regions of Italy. Individual presence in public vs. residential spaces has a more significant e?ect on the number of COVID-19 cases in the Lombardy region.","author":[{"dropping-particle":"","family":"Bilgin","given":"Nuriye Melisa","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"title":"Tracing COVID-19 Spread in Italy with Mobility Data","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(12,21)","plainTextFormattedCitation":"(12,21)","previouslyFormattedCitation":"(12,21)"},"properties":{"noteIndex":0},"schema":""}(12,21). The impact of staying at home percentage is particularly stronger in recent weeks as indicated by the higher negative impact from July 21st. The two variable effects since July 21st reflect the influence of increased exposure to COVID-19 in recent weeks across the country. The reader would note that the two measures considered were not found to be strongly correlated (see Supplementary Materials for details) and thus were simultaneously considered in the model. Health Care Infrastructure Attributes: The only set of variables found to have a significant impact of COVID-19 transmission rate within this category correspond to COVID-19 testing effects. Again, we select a 5 day lag as we believe testing results are provided in 3-5 days. The coefficient of this variable is positive as expected and highly significant ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3561142","ISSN":"1556-5068","abstract":"Predicting the binding mode of flexible polypeptides to proteins is an important task that falls outside the domain of applicability of most small molecule and protein?protein docking tools. Here, we test the small molecule flexible ligand docking program Glide on a set of 19 non-α-helical peptides and systematically improve pose prediction accuracy by enhancing Glide sampling for flexible polypeptides. In addition, scoring of the poses was improved by post-processing with physics-based implicit solvent MM- GBSA calculations. Using the best RMSD among the top 10 scoring poses as a metric, the success rate (RMSD ≤ 2.0 ? for the interface backbone atoms) increased from 21% with default Glide SP settings to 58% with the enhanced peptide sampling and scoring protocol in the case of redocking to the native protein structure. This approaches the accuracy of the recently developed Rosetta FlexPepDock method (63% success for these 19 peptides) while being over 100 times faster. Cross-docking was performed for a subset of cases where an unbound receptor structure was available, and in that case, 40% of peptides were docked successfully. We analyze the results and find that the optimized polypeptide protocol is most accurate for extended peptides of limited size and number of formal charges, defining a domain of applicability for this approach.","author":[{"dropping-particle":"","family":"Berger","given":"David","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Herkenhoff","given":"Kyle","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mongey","given":"Simon","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"An SEIR Infectious Disease Model with Testing and Conditional Quarantine","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(21)","plainTextFormattedCitation":"(21)","previouslyFormattedCitation":"(21)"},"properties":{"noteIndex":0},"schema":""}(21). However, after May10th, the effect has a higher magnitude which suggests that compared to the previous time period (before May 10th), higher testing rate will increase the daily COVID-19 transmission at a marginally higher rate. Temporal factors: With data available for 133 days, we can evaluate the effect of the transmission rate in previous time period on the current time period. As expected, we find a positive association between the daily COVID-19 transmission rate and the temporal lagged variables in the previous time period for 7 and 14 days. The result suggests higher transmission rate in previous time periods (7 and 14 days earlier) is likely to result in increased transmission. However, the effect is higher for the 7 day lagged variable, as evidenced by the higher magnitude associated with the corresponding time period in Table 2. Further, the 7 day lagged transmission rate after June 21st and July 7th time period offer larger positive impacts. Unsurprisingly, the effect for July 7th and later is significantly larger than the other variable effect. The result is aligned with the sudden surge in COVID-19 cases since beginning of July. Finally, the weekend variable highlights that the COVID-19 transmission rate is lower during weekends possibly because of reduced testing rate on weekends ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","accessed":{"date-parts":[["2020","7","11"]]},"id":"ITEM-1","issued":{"date-parts":[["0"]]},"title":"Reduced testing suggested as reason for weekend drop in confirmed COVID-19 deaths | Michigan Radio","type":"webpage"},"uris":[""]}],"mendeley":{"formattedCitation":"(35)","plainTextFormattedCitation":"(35)","previouslyFormattedCitation":"(35)"},"properties":{"noteIndex":0},"schema":""}(35). Correlation: As indicated earlier, we developed the mixed linear model for estimating the daily COVID-19 transmission rate per 100,000 people while incorporating the dependencies across each county for multiple repetition levels. Of these different models, we selected the model that provides best result in terms of statistical data fit and variable interpretation. We found that the model accommodating weekly correlations provided the best result. The final set of variables in table 2 corresponds to the correlation parameter across every 7 days within a county. All the parameters are highly significant highlighting the role of common unobserved factors influencing the daily COVID-19 transmission rate over a week across the counties. COVID-19 Mortality RateAs opposed to the transmission rate model, we adopted a simple linear regression approach to study the determinants of the COVID-19 mortality rate at a county level. The coefficients in table 3 represent the effect of different independent variables on the COVID-19 mortality rate (total number of deaths per 100K population in 3 months period) at a county level. Table 3 Estimation Results for COVID-19 Mortality Rate per 100K PopulationVariablesEstimatest-statisticp-valueConstant-6.467-3.741<0.001DemographicsOlder people % (>65 years old)0.0536.663<0.001% of African-American population0.0218.077<0.001% of People less than high school education0.07010.730<0.001Income inequality ratio0.1683.700<0.001Employment rate per capita6.3817.953<0.001Ln (% of People living in rural areas)-1.335-7.061<0.001Health IndicatorsLn (HIV rate per 100K people)0.2004.889<0.001Cancer rate per 100K people0.2561.9190?036Hepatitis A rate per 100K People0.0512.1570?031Ln (Cardiovascular disease per 1K people)0.3863.0640?002Health Care Infrastructure AttributesICU beds per capita-0.007-4.382<0.001Socio-demographics: With respect to socio-demographic variables, we find several attributes to have a significant impact on the COVID-19 mortality rate. For instance, higher percentage of older people in a county leads to an increased COVID-19 mortality rate as indicated by the positive coefficient in the Table 3. Similar results are also observed in earlier studies ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1101/2020.05.14.20101667","abstract":"Since the emerging of the &quot;novel coronavirus&quot; SARS-CoV-2 and the corresponding respiratory disease COVID-19, the virus has spread all over the world. In Europe, Germany is currently one of the most affected countries. In March 2020, a &quot;lockdown&quot; was established to contain the virus spread, including the closure of schools and child day care facilities as well as forced social distancing and bans of any public gathering. The present study attempts to analyze whether these governmental interventions had an impact on the declared aim of &quot;flattening the curve&quot;, referring to the epidemic curve of new infections. This analysis is conducted from a regional perspective. On the level of the 412 German counties, logistic growth models were estimated based on reported cases of infections, aiming at determining the regional growth rate of infections and the point of inflection where infection rates begin to decrease and the curve flattens. All German counties exceeded the peak of new infections between the beginning of March and the middle of April. In a large majority of German counties, the epidemic curve has flattened before the social ban was established (March 23). In a minority of counties, the peak was already exceeded before school closures. The growth rates of infections vary spatially depending on the time the virus emerged. Counties belonging to states which established an additional curfew show no significant improvement with respect to growth rates and mortality. On the contrary, growth rates and mortality are significantly higher in Bavaria compared to whole Germany. The results raise the question whether social ban measures and curfews really contributed to the curve flattening. Furthermore, mortality varies strongly across German counties, which can be attributed to infections of people belonging to the “risk group”, especially residents of retirement peting Interest StatementThe authors have declared no competing interest.Funding StatementNo external funding was received.Author DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with …","author":[{"dropping-particle":"","family":"Wieland","given":"Thomas","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"medRxiv","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"page":"2020.05.14.20101667","publisher":"Cold Spring Harbor Laboratory Press","title":"Flatten the Curve! Modeling SARS-CoV-2/COVID-19 Growth in Germany on the County Level","type":"article-journal"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1101/2020.05.06.20093195","abstract":"Introduction With the pandemic of COVID-19, the number of confirmed cases and related deaths are increasing in the US. We aimed to understand the potential impact of health and demographic factors on the infection and mortality rates of COVID-19 at the population level. Methods We collected total number of confirmed cases and deaths related to COVID-19 at the county level in the US from January 21, 2020 to April 23, 2020. We extracted health and demographic measures for each US county. Multivariable linear mixed effects models were used to investigate potential correlations of health and demographic characteristics with the infection and mortality rates of COVID-19 in US counties. Results Our models showed that several health and demographic factors were positively correlated with the infection rate of COVID-19, such as low education level and percentage of Black. In contrast, several factors, including percentage of smokers and percentage of food insecure, were negatively correlated with the infection rate of COVID-19. While the number of days since first confirmed case and the infection rate of COVID-19 were negatively correlated with the mortality rate of COVID-19, percentage of elders (65 and above) and percentage of rural were positively correlated with the mortality rate of COVID-19. Conclusions At the population level, health and demographic factors could impact the infection and mortality rates of COVID-19 in US counties.","author":[{"dropping-particle":"","family":"Xie","given":"Zidian","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Li","given":"Dongmei","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Medrxiv","id":"ITEM-2","issued":{"date-parts":[["2020"]]},"number-of-pages":"1-14","title":"Health and Demographic Impact on COVID-19 Infection and Mortality in US Counties","type":"report"},"uris":[""]}],"mendeley":{"formattedCitation":"(16,20)","plainTextFormattedCitation":"(16,20)","previouslyFormattedCitation":"(16,20)"},"properties":{"noteIndex":0},"schema":""}(16,20). Further, consistent with previous research ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2139/ssrn.3571699","ISSN":"1556-5068","abstract":"COVID-19 is killing African Americans at a rate 7% to 193% higher than the general population. Understanding why, as well as the reasons behind the wide variation, is paramount to saving lives. Here, we test two potential explanations for this effect. On the one hand, African Americans might be dying more because they have a lower average income (‘the Socioeconomic Hypothesis’). On the other hand, they might be dying more because their skin is more resistant to UV radiation, as we previously showed that COVID-19 infections and deaths decrease with higher irradiance (‘the Irradiance Hypothesis’). The two hypotheses are not mutually exclusive. We show that the overrepresentation of African Americans among COVID-19 deaths shows a significant negative correlation with mean solar irradiance, with a 20% decrease in Global Horizontal Irradiance leading to a 76% increase in the overrepresentation of African Americans amongst COVID-19 deaths. We then show that in Michigan, one of the US states with the lowest irradiance in early April, the % of each county’s population that is black, more than its median income, median age or % of the population above 65 years old, predicts COVID-19 morbidity and mortality rates. These results suggest a susceptibility linked to low irradiance may play a large role in African American vulnerability to COVID-19, and that black populations in (darker) locations with lower irradiance may benefit from sunlight exposure during the COVID-19 pandemic.","author":[{"dropping-particle":"","family":"Backer","given":"Alex","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"SSRN Electronic Journal","id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Why COVID-19 May Be Disproportionately Killing African Americans: Black Overrepresentation among COVID-19 Mortality Increases with Lower Irradiance, Where Ethnicity Is More Predictive of COVID-19 Infection and Mortality Than Median Income","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(19)","plainTextFormattedCitation":"(19)","previouslyFormattedCitation":"(19)"},"properties":{"noteIndex":0},"schema":""}(19), the current analysis also found a positive coefficient associated with the percentage of African-American people revealing a higher COVID-19 mortality rate in counties with higher proportion of African-American people. The variable specific to education status indicates that the likelihood of COVID-19 mortality increases with increasing share of people with less than high school education in a county. From the estimated results presented in table 3, we find that counties with higher income inequality ratio are more likely to experience higher number of COVID-19 deaths per capita relative to the counties with lower income disparities. Higher income inequality mainly reflects a significant share of low-income workers who possibly need to continue their daily activities despite the risk of COVID-19 transmission. Further, they usually have less access to the health care system and thus have an increased risk of mortality ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","accessed":{"date-parts":[["2020","7","11"]]},"id":"ITEM-1","issued":{"date-parts":[["0"]]},"title":"Income and wealth inequality in the U.S. has fueled COVID-19 deaths - MarketWatch","type":"webpage"},"uris":[""]}],"mendeley":{"formattedCitation":"(36)","plainTextFormattedCitation":"(36)","previouslyFormattedCitation":"(36)"},"properties":{"noteIndex":0},"schema":""}(36). Moreover, we find a positive association between the employment rate and COVID-19 mortality rate in a county. As discussed in the transmission model, high employment rate mainly reflects increased exposure which eventually increases the risk of COVID transmission resulting in higher risk of COVID-19 mortality. Finally, the last variable in the demographic category corresponds to the percentage of people living in rural areas that implies a negative effect on COVID-19 mortality rate indicating a reduced COVID-19 mortality rate in a county with more people living in the rural regions.Health Indicators: Among the health indicators, we found several variables significantly influence the COVID-19 mortality rate in a county. For instance, in comparison to other counties, counties with higher number of HIV, cancer, hepatitis A and cardiovascular patients are more likely to have higher number of COVID-19 deaths. This is expected as people with such conditions usually have weaker immune system which makes them vulnerable to the disease. The results are in line with a number of earlier studies ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"URL":"","accessed":{"date-parts":[["2020","7","11"]]},"author":[{"dropping-particle":"","family":"American Cancer Society","given":"","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["2020"]]},"title":"Common Questions About the New Coronavirus Outbreak","type":"webpage"},"uris":[""]},{"id":"ITEM-2","itemData":{"DOI":"10.1016/j.annonc.2020.03.296","ISSN":"15698041","PMID":"32224151","abstract":"Background: Cancer patients are regarded as a highly vulnerable group in the current Coronavirus Disease 2019 (COVID-19) pandemic. To date, the clinical characteristics of COVID-19-infected cancer patients remain largely unknown. Patients and methods: In this retrospective cohort study, we included cancer patients with laboratory-confirmed COVID-19 from three designated hospitals in Wuhan, China. Clinical data were collected from medical records from 13 January 2020 to 26 February 2020. Univariate and multivariate analyses were carried out to assess the risk factors associated with severe events defined as a condition requiring admission to an intensive care unit, the use of mechanical ventilation, or death. Results: A total of 28 COVID-19-infected cancer patients were included; 17 (60.7%) patients were male. Median (interquartile range) age was 65.0 (56.0–70.0) years. Lung cancer was the most frequent cancer type (n = 7; 25.0%). Eight (28.6%) patients were suspected to have hospital-associated transmission. The following clinical features were shown in our cohort: fever (n = 23, 82.1%), dry cough (n = 22, 81%), and dyspnoea (n = 14, 50.0%), along with lymphopaenia (n = 23, 82.1%), high level of high-sensitivity C-reactive protein (n = 23, 82.1%), anaemia (n = 21, 75.0%), and hypoproteinaemia (n = 25, 89.3%). The common chest computed tomography (CT) findings were ground-glass opacity (n = 21, 75.0%) and patchy consolidation (n = 13, 46.3%). A total of 15 (53.6%) patients had severe events and the mortality rate was 28.6%. If the last antitumour treatment was within 14 days, it significantly increased the risk of developing severe events [hazard ratio (HR) = 4.079, 95% confidence interval (CI) 1.086–15.322, P = 0.037]. Furthermore, patchy consolidation on CT on admission was associated with a higher risk of developing severe events (HR = 5.438, 95% CI 1.498–19.748, P = 0.010). Conclusions: Cancer patients show deteriorating conditions and poor outcomes from the COVID-19 infection. It is recommended that cancer patients receiving antitumour treatments should have vigorous screening for COVID-19 infection and should avoid treatments causing immunosuppression or have their dosages decreased in case of COVID-19 coinfection.","author":[{"dropping-particle":"","family":"Zhang","given":"L.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhu","given":"F.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Xie","given":"L.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wang","given":"C.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wang","given":"J.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Chen","given":"R.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Jia","given":"P.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Guan","given":"H. Q.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Peng","given":"L.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Chen","given":"Y.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Peng","given":"P.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhang","given":"P.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Chu","given":"Q.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Shen","given":"Q.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wang","given":"Y.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Xu","given":"S. Y.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhao","given":"J. P.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Zhou","given":"M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Annals of Oncology","id":"ITEM-2","issue":"7","issued":{"date-parts":[["2020"]]},"page":"894-901","title":"Clinical characteristics of COVID-19-infected cancer patients: a retrospective case study in three hospitals within Wuhan, China","type":"article-journal","volume":"31"},"uris":[""]},{"id":"ITEM-3","itemData":{"DOI":"10.1016/j.dsx.2020.03.013","ISSN":"18780334","PMID":"32247212","abstract":"Background and aims: Many patients with coronavirus disease 2019 (COVID-19) have underlying cardiovascular (CV) disease or develop acute cardiac injury during the course of the illness. Adequate understanding of the interplay between COVID-19 and CV disease is required for optimum management of these patients. Methods: A literature search was done using PubMed and Google search engines to prepare a narrative review on this topic. Results: Respiratory illness is the dominant clinical manifestation of COVID-19; CV involvement occurs much less commonly. Acute cardiac injury, defined as significant elevation of cardiac troponins, is the most commonly reported cardiac abnormality in COVID-19. It occurs in approximately 8–12% of all patients. Direct myocardial injury due to viral involvement of cardiomyocytes and the effect of systemic inflammation appear to be the most common mechanisms responsible for cardiac injury. The information about other CV manifestations in COVID-19 is very limited at present. Nonetheless, it has been consistently shown that the presence of pre-existing CV disease and/or development of acute cardiac injury are associated with significantly worse outcome in these patients. Conclusions: Most of the current reports on COVID-19 have only briefly described CV manifestations in these patients. Given the enormous burden posed by this illness and the significant adverse prognostic impact of cardiac involvement, further research is required to understand the incidence, mechanisms, clinical presentation and outcomes of various CV manifestations in COVID-19 patients.","author":[{"dropping-particle":"","family":"Bansal","given":"Manish","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Diabetes and Metabolic Syndrome: Clinical Research and Reviews","id":"ITEM-3","issue":"3","issued":{"date-parts":[["2020"]]},"page":"247-250","title":"Cardiovascular disease and COVID-19","type":"article-journal","volume":"14"},"uris":[""]}],"mendeley":{"formattedCitation":"(5,37,38)","plainTextFormattedCitation":"(5,37,38)","previouslyFormattedCitation":"(5,37,38)"},"properties":{"noteIndex":0},"schema":""}(5,37,38). Health Care Infrastructure Attributes: Finally, among health care infrastructure attributes, number of ICU beds per capita at a county is found to have a negative impact on COVID-19 mortality rate suggesting a reduced death rate with higher number of ICU bed per person in a county. The result is intuitive as more ICU bed per capita indicates the county is well equipped to handle higher patient demand and treatment is accessible to more COVID-19 patients. Policy ImplicationsTo illustrate the applicability of the proposed COVD-19 transmission model, we conduct a scenario analysis exercise by imposing hypothetical mobility restrictions. While earlier researchers explored the influence of mobility measures, these models did not account for county level factors such as socio-demographics, health indicators and hospital infrastructure attributes. In our framework, the sensitivity analysis is conducted while controlling for several other factors. The hypothetical restrictions on mobility are considered through the following changes to two variables: (1) county level average daily exposure reduced by 10%, 25% and 50% (2) county level percentage of stay at home population increased to 40%, 50% and 60%. The changes to the independent variables were used to predict the transformed dependent variable. Subsequently, the transformed variable was converted to the daily cases per 100 thousand people. The results from this exercise are presented in Table 4. We present the average change in cases for all counties (1,752), and for the 100 counties with the highest overall transmission rates. From table 4, two important observations can be made. First, changes to average daily exposure and stay at home population influence COVID-19 transmission significantly. In fact, by increasing stay at home population share to 50%, the model predicts a reduction of the number of cases by about 33%. Further, mobility restriction results in suppressed COVID-19 transmission as indicated by the negative values from Table 4. Second, the benefit from mobility restrictions and staying at home is slightly higher for the worst 100 counties with higher overall cases. The two observations provide evidence that issuing lockdown orders in counties with a recent surge is a potential mitigation measure to curb future transmission. The COVID-19 total mortality rate model can be employed to identify vulnerable counties that need to be prioritized for vaccination programs (when available). While prioritizing the counties based on mortality rate might be a potential approach, it might be feasible. To elaborate, vaccination programs have to be planned well in advance (say 2 months) of the vaccine availability. As total mortality rates for 2 months into the future are unavailable, we need a model to predict total mortality into the future. The estimated mortality rate model provides a framework for such analysis. To be sure, it would be prudent to update the proposed model with the latest data to develop a more accurate prediction system. Table 4 Policy Scenario Analysis of Social Distancing in COVID-19 Transmission Rate per 100K PopulationHypothetical Scenarios1,752 CountiesWorst 100 Counties1: daily average exposure reduced by 10%-0.636-0.6402: daily average exposure reduced by 25%-1.716-1.7263: daily average exposure reduced by 50%-4.030-4.0554: 40% people stay at home-26.423-26.6545: 50% people stay at home-33.082-33.2586: 60% people stay at home-38.561-38.700DiscussionThe current study develops a comprehensive framework for examining COVID-19 transmission and mortality rates in the United States at a county level including an exhaustive set of independent variables: socio-demographics, health indicators, mobility trends and health care infrastructure attributes. In our analysis, we consider all counties with total number of cases greater than 100 on August 4th and analyze daily cases data from March 25th to August 4th, 2020. The COVID-19 transmission rate is modeled at a daily basis using a linear mixed method while the total mortality rate is analyzed adopting a linear regression approach. Several county level factors including proportion of African-Americans, income inequality, health indicators associated with Asthma, Cancer, HIV and heart disease, percentage of stay at home individuals, testing infrastructure and Intensive Care Unit capacity impact transmission and/or mortality rates. The results clearly support our hypothesis of considering a universal set of factors in analyzing the COVID-19 data. Further we conducted policy scenario analysis to evaluate the influence of social distancing on the COVID-19 transmission rate. The results highlight the effectiveness of social distancing in mitigating the virus transmission. In fact, we found that by increasing stay at home population share to 50% the model predicts a reduction of the number of cases by about 33%. The finding provides evidence that issuing lockdown orders in counties with a recent surge is a potential mitigation measure to curb future transmission. To be sure, the study is not without limitations. The study is focused on county level analysis and is intended to reflect associations as opposed to causation. However, for the causation based analysis, data from individuals would be more suitable. As with any area level analysis, there is a small possibility that some of the estimated parameters might be spurious associations due to aggregation bias. However, in the absence of individual level data, these area level models offer a valid and useful tool for epidemiologists and planners. Further, the inherent aggregation of the data at a county level would initiate some form of spatial heterogeneity which we did not account for in our analysis. In future, it would be interesting to accommodate these effects separately while considering the temporal correlation. Further, the proposed model can be enhanced using more detailed information such as percentage of health workers in the workforce, number of hospital beds and mask mandate dates. While exposure data were reasonably addressed, data was not available for mask wearing behavior across all counties. Finally, the data on transmission and mortality are updated for few counties to correct for errors or omissions. These were carefully considered in our data preparation. However, it is possible that further updates might be made after we finished our analysis. . ContributorsNE conceptualized the study. TB and NE finalized the study design. TB, SD and NC conducted the literature review. TB, SD and NC collected the data. TB, SD, NC, and NE analyzed and interpreted the model results. TB, NC and SD prepared the figures. TB, SD, NC and NE drafted the main manuscript. All authors reviewed the results and approved the final version of the manuscript.Declaration of InterestsWe declare no competing interests. AcknowledgementThe authors would like to gratefully acknowledge SafeGraph COVID-19 Data Consortium, County Health Ranking and Road Maps, Centers for Disease Control System for providing access to the data at county level for United States. SafeGraph is a data company that aggregates anonymized location data from numerous applications in order to provide insights about physical places. To enhance privacy, SafeGraph excludes census block group information if fewer than five devices visited an establishment in a month from a given census block groupReferences1. Worldometer. Coronavirus Cases & Mortality [Internet]. Worldometer. 2020 [cited 2020 Jul 12]. p. 1–22. Available from: . The Global Economic Outlook During the COVID-19 Pandemic: A Changed World [Internet]. [cited 2020 Jul 12]. Available from: . Bhowmik T, Eluru N. A Comprehensive County Level Framework to Identify Factors Affecting Hospital Capacity and Predict Future Hospital Demand. [cited 2021 Mar 9]; Available from: . 4. Centers for Disease Control and Prevention (CDC). Cases in the U.S. [Internet]. Vol. 2019, Coronavirus Disease 2019 (COVID-19). 2020 [cited 2020 Jul 12]. p. 1–4. Available from: 5. Bansal M. Cardiovascular disease and COVID-19. Diabetes Metab Syndr Clin Res Rev. 2020;14(3):247–50. 6. Engle S, Stromme J, Zhou A. Staying at Home: Mobility Effects of COVID-19. SSRN Electron J. 2020; 7. Mukandavire Z, Nyabadza F, Malunguza NJ, Cuadros DF, Shiri T, Musuka G. Quantifying early COVID-19 outbreak transmission in South Africa and exploring vaccine efficacy scenarios. PLoS One. 2020;15(7 July):e0236003. 8. Yuan X, Xu J, Hussain S, Wang H, Gao N, Zhang L. Trends and Prediction in Daily New Cases and Deaths of COVID-19 in the United States: An Internet Search-Interest Based Model. Explor Res Hypothesis Med. 2020;000(000):1–6. 9. Courtemanche CJ, Garuccio J, Le A, Pinkston JC, Yelowitz A. Did Social-Distancing Measures in Kentucky Help to Flatten the COVID-19 Curve? Inst study Free Enterp Work Pap [Internet]. 2020;4:1–23. Available from: . Zhan C, Tse CK, Lai Z, Hao T, Su J. Prediction of COVID-19 spreading profiles in South Korea, Italy and Iran by data-driven coding. PLoS One. 2020;15(7 July). 11. Ci C, Zhou B, Xi L. A support vector machine based scheduling approach for a material handling system. In: Proceedings - 2010 6th International Conference on Natural Computation, ICNC 2010. 2010. p. 3768–72. 12. Bilgin NM. Tracing COVID-19 Spread in Italy with Mobility Data. SSRN Electron J. 2020; 13. Mollalo A, Vahedi B, Rivera KM. GIS-based spatial modeling of COVID-19 incidence rate in the continental United States. Sci Total Environ. 2020;728. 14. Sharkey P, Wood G. The Causal Effect of Social Distancing on the Spread of SARS-CoV-2. 2020; 15. Wu X, Nethery RC, Sabath BM, Braun D, Dominici F. Exposure to air pollution and COVID-19 mortality in the United States: A nationwide cross-sectional study. medRxiv Prepr Serv Heal Sci. 2020; 16. Wieland T. Flatten the Curve! Modeling SARS-CoV-2/COVID-19 Growth in Germany on the County Level. medRxiv [Internet]. 2020;2020.05.14.20101667. Available from: . Xie J, Zhu Y. Association between ambient temperature and COVID-19 infection in 122 cities from China. Sci Total Environ. 2020;724. 18. Borjas GJ. Demographic Determinants of Testing Incidence and COVID-19 Infections in New York City Neighborhoods. SSRN Electron J. 2020; 19. Backer A. Why COVID-19 May Be Disproportionately Killing African Americans: Black Overrepresentation among COVID-19 Mortality Increases with Lower Irradiance, Where Ethnicity Is More Predictive of COVID-19 Infection and Mortality Than Median Income. SSRN Electron J. 2020; 20. Xie Z, Li D. Health and Demographic Impact on COVID-19 Infection and Mortality in US Counties [Internet]. Medrxiv. 2020. Available from: . Berger D, Herkenhoff K, Mongey S. An SEIR Infectious Disease Model with Testing and Conditional Quarantine. SSRN Electron J. 2020; 22. Teslya A, Pham TM, Godijk NG, Kretzschmar ME, Bootsma MCJ, Rozhnova G. Impact of self-imposed prevention measures and short-term government-imposed social distancing on mitigating and delaying a COVID-19 epidemic: A modelling study. PLoS Med. 2020;17(7):e1003166. 23. Siedner MJ, Harling G, Reynolds Z, Gilbert RF, Haneuse S, Venkataramani AS, et al. Social distancing to slow the US COVID-19 epidemic: Longitudinal pretest–posttest comparison group study. PLoS Med [Internet]. 2020;17(8 August):2020.04.03.20052373. Available from: . Badr HS, Du H, Marshall M, Dong E, Squire MM, Gardner LM. Association between mobility patterns and COVID-19 transmission in the USA: a mathematical modelling study. Lancet Infect Dis. 2020; 25. Omori R, Mizumoto K, Chowell G. Changes in testing rates could mask the novel coronavirus disease (COVID-19) growth rate. Int J Infect Dis. 2020;94:116–8. 26. Triantaphyllou E, Sánchez A. A sensitivity analysis approach for some deterministic multi-criteria decision-making methods. Decis Sci. 1997; 27. Rehman S, A. S. Multi-Criteria Wind Turbine Selection using Weighted Sum Approach. Int J Adv Comput Sci Appl. 2017; 28. Mateo JRSC. Weighted sum method and weighted product method. In: Green Energy and Technology. 2012. 29. USA Counties | ArcGIS Hub [Internet]. [cited 2020 Dec 11]. Available from: . Couture V, Dingel JI, Green A, Handbury J, Williams KR. Measuring Movement and Social Contact With Smartphone Data?: a Real-Time Application To Covid-19 [Internet]. Vol. No. 27560, NBER Working Paper Series. 2020. Report No.: 35. Available from: . The COVID-19 tracking project [cited 2020 June 12]. Available from: . COVID-19 Map - Johns Hopkins Coronavirus Resource Center [cited 2020 Jul 11]. Available from: . Faghih-Imani A, Eluru N, El-Geneidy AM, Rabbat M, Haq U. How land-use and urban form impact bicycle flows: Evidence from the bicycle-sharing system (BIXI) in Montreal. J Transp Geogr. 2014;41:306–14. 34. Ahmed F, Ahmed N, Pissarides C, Stiglitz J. Why inequality could spread COVID-19. Vol. 5, The Lancet Public Health. 2020. p. e240. 35. Reduced testing suggested as reason for weekend drop in confirmed COVID-19 deaths | Michigan Radio [cited 2020 Jul 11]. Available from: . Income and wealth inequality in the U.S. has fueled COVID-19 deaths - MarketWatch [cited 2020 Jul 11]. Available from: . Common Questions About the New Coronavirus Outbreak [cited 2020 Jul 11]. Available from: HYPERLINK "" 38. Zhang L, Zhu F, Xie L, Wang C, Wang J, Chen R, et al. Clinical characteristics of COVID-19-infected cancer patients: a retrospective case study in three hospitals within Wuhan, China. Ann Oncol. 2020;31(7):894–901. 39. Robinson, G. M.. Time Series Analysis. In International Encyclopedia of Human Geography. 2009; 285–293. Elsevier Inc. . Ringle CM, Wende S, Becker JM. SmartPLS 3. Boenningstedt: SmartPLS GmbH. 2015. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download