Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

Song Gao; Jinmeng Rao; Yuhao Kang; Yunlei Liang; Jake Kruse; Dorte Dopfer; Ajay K. Sethi; Juan Francisco Mandujano Reyes; Brian S. Yandell; Jonathan A. Patz

doi:10.1001/jamanetworkopen.2020.20485

Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

Gao, Song; Rao, Jinmeng; Kang, Yuhao; Liang, Yunlei; Kruse, Jake; Dopfer, Dorte; Sethi, Ajay K.; Mandujano Reyes, Juan Francisco; Yandell, Brian S.; Patz, Jonathan A. 2020-09-08 00:00:00 Key Points Question Did human mobility patterns IMPORTANCE A stay-at-home social distancing mandate is a key nonpharmacological measure to change during stay-at-home orders and reduce the transmission rate of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), but were the mobility changes associated a high rate of adherence is needed. with the coronavirus disease 2019 (COVID-19) curve? OBJECTIVE To examine the association between the rate of human mobility changes and the rate of Findings This cross-sectional study confirmed cases of SARS-CoV-2 infection. using anonymous location data from more than 45 million mobile phones DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study used daily travel distance and found that median travel distance home dwell time derived from millions of anonymous mobile phone location data from March 11 to decreased and stay-at-home time April 10, 2020, provided by the Descartes Labs and SafeGraph to quantify the degree to which social increased across the nation, although distancing mandates were followed in the 50 US states and District of Columbia and the association there was geographic variation. State- of mobility changes with rates of coronavirus disease 2019 (COVID-19) cases. specific empirical doubling time of total COVID-19 cases increased (ie, the spread EXPOSURE State-level stay-at-home orders during the COVID-19 pandemic. reduced) significantly after stay-at- home orders were put in place. MAIN OUTCOMES AND MEASURES The main outcome was the association of state-specific rates of COVID-19 confirmed cases with the change rates of median travel distance and median home Meaning These findings suggest that dwell time of anonymous mobile phone users. The increase rates are measured by the exponent in stay-at-home social distancing curve fitting of the COVID-19 cumulative confirmed cases, while the mobility change (increase or mandates were associated with the decrease) rates were measured by the slope coefficient in curve fitting of median travel distance and reduced spread of COVID-19 when they median home dwell time for each state. were followed. RESULTS Data from more than 45 million anonymous mobile phone devices were analyzed. The Supplemental content correlation between the COVID-19 increase rate and travel distance decrease rate was –0.586 (95% CI, –0.742 to –0.370) and the correlation between COVID-19 increase rate and home dwell time Author affiliations and article information are listed at the end of this article. increase rate was 0.526 (95% CI, 0.293 to 0.700). Increases in state-specific doubling time of total cases ranged from 1.0 to 6.9 days (median [interquartile range], 2.7 [2.3-3.3] days) before stay-at- home orders were enacted to 3.7 to 30.3 days (median [interquartile range], 6.0 [4.8-7.1] days) after stay-at-home social distancing orders were put in place, consistent with pandemic modeling results. CONCLUSIONS AND RELEVANCE These findings suggest that stay-at-home social distancing mandates, when they were followed by measurable mobility changes, were associated with reduction in COVID-19 spread. These results come at a particularly critical period when US states are beginning to relax social distancing policies and reopen their economies. These findings support the efficacy of social distancing and could help inform future implementation of social distancing policies should they need to be reinstated during later periods of COVID-19 reemergence. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 Open Access. This is an open access article distributed under the terms of the CC-BY License. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 1/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Introduction The coronavirus disease 2019 (COVID-19) pandemic is a global threat with escalating health, economic, and social challenges. As of April 11, 2020, there were 492 416 total confirmed cases and 18 559 total deaths in the US, according to reports from the Centers for Disease Control and Prevention (CDC). People are still witnessing widespread community transmission of COVID-19 all over the world. To date, there is neither a vaccine nor pharmacological agent found to reduce the transmission of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the virus that causes COVID-19. Thus, the effects of nonpharmacological pandemic control and intervention measures, including travel restrictions, closures of schools and nonessential business services, wearing of face masks, testing, isolation, and timely quarantine on delaying the spread of COVID-19, have been 2-6 largely investigated and reported. To mitigate and ultimately contain the COVID-19 pandemic, one of the important nonpharmacological control measures to reduce the transmission rate of SARS- CoV-2 in the population is social (ie, physical) distancing. An interactive web-based mapping platform that provides timely quantitative information on how people in different counties and states reacted to state-at-home social distancing mandates has been developed (eAppendix 2 in the Supplement). It integrates geographic information systems and daily updated human mobility statistical patterns derived from millions of anonymized and aggregated smartphone location data at the county level in 7-10 the US. Reduced mobility and trips may help limit people’s exposure to large in-person gatherings. However, it is worth noting that reduced mobility does not necessarily ensure that social distancing in practice follows the CDC’s definition: “stay at least 6 feet (about 2 arms’ length) from other 11 12 people.” Due to the mobile phone Global Positioning System horizontal error and uncertainty, such physical distancing patterns cannot be directly identified from the user aggregated mobility data; that would require other wearable sensors or mobile phone Bluetooth trackers, which raise issues of personal data privacy and ethical concerns. Because COVID-19 is more contagious and far more deadly than seasonal flu, social distancing is critical in the fight to save lives and prevent illness. However, to what degree such guidelines have been followed from place to place before and after shelter-in-place orders across the US and the quantitative effect on flattening the curve are as yet unknown, to our knowledge. To this end, we used 2 human mobility metrics, the median of individual maximum travel distance and stay-at-home time derived from location data from millions of mobile phones, to assess the association of stay-at-home policies with reducing the spread of COVID-19. For each state, we examined these measures against the rate of SARS-CoV-2 infection cases. Methods A waiver of institutional review board review and informed consent was obtained from the University of Wisconsin–Madison because anonymized and aggregated data were used and our study does not involve human participants as defined. This study follows the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) reporting guideline. Data In this cross-sectional study, the epidemiological confirmed cases data were retrieved from the Corona Data Scraper open source project, which provides local-level and community-driven reports, and we conflated the data with the state-level department of health services official reports in each state to ensure the data quality. To understand how people reacted to the stay-at-home social distancing guidelines imposed during the COVID-19 pandemic, human mobility changes were considered in terms of changes in travel distance and stay-at-home dwell time. The travel distance mobility data were collected from an open-source repository released by Descartes Labs, while the home dwell time data derived from more than 45 million anonymous mobile phone users were JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 2/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US processed from SafeGraph. Both data sources were acquired at the county level and aggregated to the state level using median and interquartile range (IQR) values. To consider the socioeconomic factors that may be associated with statewide changes in human mobility, socioeconomic variables at the state level were also collected from the American Community Survey and the US Census Bureau. The following socioeconomic variables and geospatial data sets were retrieved and computed: population (ie, number of people), population density (measured by population divided by area of state), proportion of population with bachelor’s degree, proportions of population of different races/ethnicities, proportion of population of different age groups, median household income, and urban core area boundaries. Statistical Analysis Simple linear regression and multivariate linear regression analyses were performed using the scikit- learn package version 0.23.1 in Python. The Pearson correlation coefficient with 2-sided significance test, P < .05, was computed using the SciPy package version 1.4.0 in Python. Curve Fitting for Pandemic Spread, Travel Distance, and Home Dwell Time In the mathematical modeling process, we used a few types of mathematical formulas (eAppendix 1 in the Supplement) to fit the curve of the cumulative confirmed cases for the COVID-19 with respect to their temporal changes in each state and selected the following scaling law with a deivation term formula as the most appropriate: y (t)= t + k, in which y is the total number of confirmed cases in c c each state as a function of time, t is the number of days from March 11, 2020 (when the COVID-19 became a pandemic), and b and k are parameters we will estimate. By fitting the curve, we can compare the infection rates among different states using the coefficient b estimated from the model. Meanwhile, we used linear regression to detect the travel distance decreasing rate (represented by the slope estimated from the linear model) over time (eFigure 2 in the Supplement) and examined whether there was a correlation between the increase rate of cases and the distance decreasing rate. We also fitted the curve for the home dwell time changes for each state using the linear regression model. The linear model was selected from a few different models because it is the simplest one, and results of all fitted models were similar. We then calculated the correlation between the home dwell time increasing rate (the slope estimated from the linear model) and the increase rate of the number of confirmed cases. Evaluating Factors Associated With Changes in Travel Distance and Home Dwell Time To understand what socioeconomic factors were associated with travel distance changes and home dwell time changes, a multilinear regression model integrating socioeconomic factors was used to fit the mobility change rates that were represented by the slope estimates for each state. The R as goodness of fit and significance of variables are reported (eAppendix 1 in the Supplement). Calculating the Doubling Time of Total Confirmed Cases We investigated how the social distancing guidelines and stay-at-home orders (eTable 5 in the Supplement) were associated with the pandemic doubling time of COVID-19 confirmed cases from March 11 to April 10, 2020, in each state. We used mathematical curve fitting models and mechanistic epidemic models (eAppendix 1 in the Supplement) using Bayesian parametric estimation of the serial 18,19 interval distribution of successive cases to cross validate the conclusion. We calculated the doubling time of the number of cumulative confirmed cases (ie, the time intervals it takes for the cumulative confirmed cases to double in size ) to reflect the characteristics of the COVID-19 pandemic spread, especially how the stay-at-home orders in each state were associated with flattening the COVID-19 curve. The larger the doubling time, the smoother the pandemic increase curve. Within the time frame of our study, the state-level increase rates of COVID-19 cases in the US were either exponential or subexponential, thus we implemented an exponential model and a power-law model to fit the curve for calculating the doubling time. We also calculated the doubling JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 3/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US time based on empirical observations (model-free) to further explore how the doubling time differs in these methods. We used the effective date of the stay-at-home order to split the confirmed case data into 2 parts: before the order and after the order. We fitted each model on the data before the order and after the order, then we calculated the doubling times of the confirmed cases based on the model and empirical COVID-19 infection data. The doubling time of the cumulative confirmed cases in each state is defined as ln(2) d(t) = ln(1 + r[t]) In which d(t) represents the doubling time of the cumulative confirmed cases on date t in each state, ln(x) is the natural log of x, and r(t) represents the increase rate of the cumulative confirmed cases on date t in each state. In addition, we visualized and investigated the overall probability density distribution of the median doubling time before and after the stay-at-home order in each state to have a better understanding of the overall changes in the pandemic spread nationwide. Furthermore, we measured the similarity in probability density distribution of the median doubling time between the fitting results and the empirical data using the Jensen–Shannon Divergence. Results Trends of Human Mobility Changes Data from more than 45 million anonymous mobile phone devices were analyzed. The associations of stay-at-home policies with human mobility changes are illustrated in Figure 1, Figure 2, and the Table. Figure 1A shows the temporal changes of the median of individual maximum travel distances in the states with the highest infection rates (ie, New York, New Jersey, Michigan, California, and Massachusetts) by April 10, 2020. People’s daily mobility decreased significantly but with different temporal lags following the implementation of statewide stay-at-home orders across these states (Table). Figure 1B shows the state-specific temporal changes of median home dwell time. With the social distancing guidelines and shelter-at-home orders in place, the median home dwell time increased significantly in most states since March 23, 2020 (Table). Figure 2 shows the spatial distributions of confirmed cases per capita and the median of travel distances and median of home dwell time in 2 specific days as snapshots for comparison of mobility patterns with the COVID-19 infection rate before and after stay-at-home-orders: March 11 and April 10, 2020. The median travel Figure 1. Temporal Changes in Median of Individual Maximum Travel Distance and Median Home Dwell Time in the Most Infected US States From March 11 to April 10, 2020 A Travel distance B Stay-at-home dwell time New York 10 1100 New Jersey Michigan Massachusetts 1000 California Florida ALL 0 400 March March March March April April March March March March April April 11, 2020 17, 2020 23, 2020 29, 2020 4, 2020 10, 2020 11, 2020 17, 2020 23, 2020 29, 2020 4, 2020 10, 2020 Date Date JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 4/13 Median travel distance, km Home dwell time, min JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US distance decreased and the median home dwell time increased across the US during this period. In addition, we calculated the means of median travel distances before and after the stay-at-home- orders in each state. The median travel distances decreased in all states (Table). Implementation of stay-at-home social distancing policies were associated with human movement changes, that is, people generally reduced their daily travel distances and increased their home dwell time. Interestingly, the multiple linear regression model result for the increasing rate of home dwell time with the socioeconomic variables shows that the ratio of Asian individuals in each state was positively associated with longer home dwell time at the state level. The higher the proportion of Asian Figure 2. Comparison Among Confirmed Coronavirus Disease 2019 Cases Per Capita, Median of Individual Maximum Travel Distance, and Median Home Dwell Time From March 11 and April 10, 2020 A Confirmed cases Cases per capita >0.00 to 0.02 >0.02 to 0.08 >0.08 to 0.31 >0.31 to 15 775.90 March 11th April 10th B Median maximum travel distance Distance, km 0 to 3.38 >3.38 to 6.87 >6.87 to 9.89 >9.89 to 14.51 >14.51 to 684.76 March 11th April 10th C Stay-at-home dwell time Time, min 0.00 - 664.89 >664.89 to 747.50 >747.50 to 799.49 >799.49 to 856.61 >856.61 to 1185.02 March 11th April 10th JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 5/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Table. Empirical Doubling Time of Total Infected Cases and the Median Travel Distance and Home Dwell Time Before and After Stay-at-Home Orders Doubling time, d Travel distance, km Home dwell time, min Median (IQR) Median (IQR) Median (IQR) State Before order After order Change Before order After order Change Before order After order Change Alabama 3.3 6.5 3.2 6.651 4.311 –2.328 660.9 781.4 120.5 (2.3-4.4) (5.4-7.8) (5.356-8.278) (4.283-4.903) (576.3-695.4) (758.5-799.2) Alaska 6.9 30.3 23.4 1.369 0.091 –1.273 342.3 427.1 84.8 (3.5-9.2) (23.6-38.9) (0.168-2.450) (0.049-0.092) (282.5-376.2) (343.6-454.4) Arizona 2.5 6.8 4.3 3.227 1.037 –2.231 523.9 637.9 114.0 (2.0-3.8) (5.8-10.5) (1.82-5.071) (0.878-1.274) (489.3-560.7) (520.9-645.1) California 3.3 5.3 2.0 3.922 0.770 –3.207 748.1 833.0 84.9 (3.1-3.7) (3.8-7.3) (3.128-6.177) (0.259-0.986) (642.7-760.4) (754.4-874.0) Colorado 2.6 6.2 3.6 2.824 0.319 –2.496 529.6 676.5 146.9 (2.3-3.2) (5.7-9.3) (1.387-4.334) (0.095-0.464) (475.1-548.0) (575.7-692.6) Connecticut 1.7 4.5 2.8 3.100 0.396 –2.687 667.5 822.5 155.0 (1.3-2.8) (2.8-7.5) (2.174-4.482) (0.107-0.549) (583.0-725.4) (752.4-854.1) Delaware 2.9 4.7 1.8 4.102 0.641 –3.462 629.3 749.5 120.2 (1.5-4.7) (3.7-5.4) (2.888-5.704) (0.183-0.937) (546.9-667.6) (676.8-795.6) Florida 3.0 10.0 7.0 3.484 0.930 –2.622 559.6 694.3 134.7 (2.1-3.9) (8.8-11.1) (1.805-5.224) (0.655-1.208) (476.2-604.4) (680.0-717.5) Georgia 3.5 6.4 2.9 4.852 2.278 –2.758 636.6 759.4 122.9 (2.3-5.0) (6.1-10.7) (3.292-6.57) (1.818-3.04) (546.2-674.1) (732.3-784.3) Hawaii 2.0 7.3 5.3 4.294 1.147 –3.177 625.7 789.4 163.7 (1.6-2.4) (5.2-11.1) (3.131-6.057) (1.054-1.466) (541.1-649.5) (607.3-830.9) Idaho 1.3 4.8 3.5 3.424 1.286 –2.208 567.9 686.2 118.3 (1.0-2.6) (2.9-9.8) (2.661-4.599) (1.063-1.713) (499.7-604.7) (621.3-718.7) Illinois 1.9 4.7 2.8 4.214 0.784 –3.428 648.6 764.0 115.4 (1.9-2.4) (4.0-7.1) (3.046-6.604) (0.427-1.144) (599.5-694.8) (725.9-802.8) Indiana 2.7 3.7 1.0 4.513 1.64 –2.932 605.2 718.8 113.6 (2.0-3.0) (3.0-4.2) (3.41-6.232) (1.432-2.193) (525.7-634.2) (653.5-757.9) Kansas 2.7 5.8 3.1 3.589 1.897 –1.67 606.2 702.1 96.0 (1.8-3.5) (5.1-10.2) (2.300-4.657) (1.735-2.269) (553.6-644.3) (607.0-730.9) Kentucky 2.5 5.4 2.9 4.778 2.802 –2.041 630.5 744.3 113.8 (1.7-4.1) (4.3-9.1) (3.745-6.087) (2.333-3.256) (562.8-670.4) (686.1-764.2) Louisiana 2.1 4.6 2.5 6.242 3.176 –3.122 609.5 736.9 127.4 (1.9-2.3) (3.1-8.7) (5.925-8.289) (2.877-3.830) (515.4-631.4) (675.7-763.0) Maine 3.7 16.5 12.8 2.413 0.361 –2.014 553.2 690.0 136.7 (2.4-7.1) (11.6-17.6) (0.735-3.331) (0.094-0.705) (481.6-590.2) (638.8-703.4) Maryland 2.8 4.2 1.4 2.346 0.122 –2.271 688.5 794.9 106.4 (2.2-3.6) (3.4-6.1) (0.307-3.654) (0.045-0.092) (611.4-745.6) (676.6-824.0) Massachusetts 3.8 4.7 0.9 2.323 0.108 –2.213 640.4 780.8 140.5 (3.0-5.3) (4.5-6.3) (0.991-3.412) (0.045-0.104) (538.5-670.0) (692.0-812.3) Michigan 2.3 4.4 2.1 3.562 0.104 –3.454 566.6 734.3 167.6 (1.4-2.8) (3.7-7.1) (2.274-5.058) (0.046-0.131) (492.3-600.5) (649.8-764.3) Minnesota 3.0 8.7 5.7 2.927 0.482 –2.54 556.9 701.4 144.5 (1.7-4.9) (7.6-9.4) (1.359-4.268) (0.138-0.509) (500.4-607.0) (607.1-732.2) Mississippi 2.8 9.4 6.6 7.103 4.751 –2.613 612.1 744.6 132.5 (1.7-5.1) (6.4-13.6) (5.675-8.868) (4.11-5.919) (514.0-654.5) (720.4-767.7) Montana 2.4 8.3 5.9 2.353 0.820 –1.475 443.3 559.6 116.3 (1.8-3.2) (7.4-14.5) (1.821-2.953) (0.405-1.158) (400.8-506.0) (477.2-577.8) Nevada 3.7 11.2 7.5 2.432 0.502 –1.962 516.0 611.5 95.6 (1.7-5.0) (8.5-12.6) (0.687-4.353) (0.253-0.764) (479.3-553.0) (596.7-620.5) New Hampshire 3.0 5.8 2.8 3.689 0.818 –3.014 585.0 735.4 150.4 (2.3-4.3) (4.3-11.7) (1.527-5.603) (0.266-1.073) (528.3-631.6) (623.7-752.7) New Jersey 1.8 4.2 2.4 3.244 0.095 –3.162 722.1 968.4 246.3 (1.3-2.0) (3.1-6.6) (1.972-5.362) (0.043-0.085) (671.7-819.5) (900.8-983.9) New Mexico 3.1 5.2 2.1 3.492 0.993 –2.519 467.8 577.5 109.8 (2.6-3.5) (4.4-6.9) (2.728-4.579) (0.873-1.275) (407.9-489.1) (488.1-596.8) New York 1.8 6.4 4.6 2.093 0.037 –2.056 580.0 767.4 187.4 (1.5-2.2) (4.4-9.5) (1.137-3.554) (0.032-0.039) (527.3-644.9) (669.5-785.6) North Carolina 2.7 6.3 3.6 5.220 2.679 –2.577 606.2 690.1 84.0 (2.1-3.5) (5.1-11.0) (3.935-7.065) (2.204-3.199) (545.8-633.2) (595.4-711.2) Ohio 2.1 5.3 3.2 4.076 1.202 –2.934 611.0 729.7 118.7 (1.9-2.5) (3.8-8.0) (3.275-6.096) (0.806-1.603) (547.3-653.0) (688.0-762.5) Oklahoma 2.4 5.6 3.2 5.962 3.550 –2.511 631.3 767.3 136.1 (1.6-3.1) (4.3-6.8) (4.864-7.734) (2.881-4.277) (563.2-664.9) (707.3-804.4) Oregon 3.8 6.7 2.9 2.667 0.571 –2.124 629.3 742.3 113.0 (3.2-4.3) (5.0-10.8) (1.930-3.900) (0.232-0.854) (575.8-663.4) (687.0-789.9) (continued) JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 6/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Table. Empirical Doubling Time of Total Infected Cases and the Median Travel Distance and Home Dwell Time Before and After Stay-at-Home Orders (continued) Doubling time, d Travel distance, km Home dwell time, min Median (IQR) Median (IQR) Median (IQR) State Before order After order Change Before order After order Change Before order After order Change Pennsylvania 2.5 5.8 3.3 1.798 0.184 –1.609 656.0 776.7 120.6 (2.1-3.3) (4.5-6.0) (0.078-2.445) (0.089-0.246) (562.7-704.5) (770.2-798.4) Rhode Island 1.9 4.6 2.7 2.286 0.256 –2.034 705.1 795.2 90.1 (1.4-3.5) (4.1-5.3) (0.804-3.592) (0.071-0.357) (587.1-733.2) (747.6-823.0) South Carolina 2.4 5.8 3.4 6.484 3.898 –2.664 586.9 700.2 113.2 (1.8-4.2) (3.9-8.1) (4.942-8.405) (3.651-4.390) (532.3-619.1) (626.4-712.3) Tennessee 3.3 10.3 7.0 5.679 3.442 –2.189 647.8 731.4 83.6 (1.8-4.0) (8.4-12.5) (4.094-7.368) (3.215-4.098) (590.0-683.3) (617.3-760.8) Texas 3.4 6.0 2.6 4.076 1.869 –2.249 589.5 728.8 139.2 (2.5-5.5) (5.5-7.1) (2.413-5.763) (1.837-2.326) (525.1-645.8) (722.2-759.4) Utah 2.5 6.7 4.2 3.351 1.369 –2.05 642.6 710.9 68.3 (1.9-4.1) (5.2-11.6) (2.094-4.933) (0.916-1.791) (560.2-666.3) (667.9-721.0) Vermont 2.3 7.1 4.8 2.716 0.166 –2.592 465.8 648.4 182.7 (1.6-2.8) (5.1-11.9) (0.843-4.386) (0.059-0.201) (414.6-515.0) (535.6-668.2) Virginia 3.4 4.8 1.4 3.261 1.029 –2.273 607.6 695.1 87.5 (2.4-4.9) (4.1-7.2) (1.454-4.669) (0.627-1.320) (556.1-645.8) (596.8-716.4) Washington 4.5 12.3 7.8 2.710 0.253 –2.501 683.5 811.6 128.1 (4.1-6.2) (5.2-14.2) (2.027-4.187) (0.054-0.332) (618.6-717.0) (760.0-848.3) Washington, DC 3.5 6.9 3.4 0.85 0.026 –0.823 615.7 716.7 101.0 (1.9-5.6) (4.3-7.2) (0.031-1.112) (0.024-0.027) (523.4-639.7) (696.9-722.0) West Virginia 1.0 4.4 3.4 4.611 1.691 –2.939 586.3 693.1 106.8 (1.0-1.3) (3.9-7.3) (3.573-6.217) (1.345-2.095) (488.6-626.1) (619.5-721.2) Wisconsin 2.3 7.0 4.7 3.233 0.753 –2.477 594.2 720.2 126.0 (1.9-2.6) (6.1-9.6) (2.061-4.871) (0.574-1.23) (549.6-631.4) (660.3-763.8) Wyoming 3.1 7.9 4.8 2.719 1.798 –0.867 478.3 617.7 139.4 (2.1-5.1) (5.5-11.0) (2.381-3.433) (1.218-2.198) (430.9-539.2) (491.8-636.6) Abbreviations: DC, District of Columbia; IQR, interquartile range. population, the longer the median home dwell time of residents in that state (eTable 7 in the Supplement). Association of Rate of Infection With Mobility Changes We fitted the curves for the state-specific COVID-19 confirmed cases using the scaling-law with a deviation term formula and identified the top 5 states with the largest increase rates of confirmed COVID-19 cases by April 10, 2020: New York, New Jersey, California, Michigan, and Massachusetts. Our fitting results corresponded to the up-to-date COVID-19 situation at that time (eTable 1 and eTable 2 in the Supplement). eFigure 1 in the Supplement shows the reported cases and the fitting curves in these 5 states using the scaling-law with a deviation term formula. The Pearson correlation coefficient between the cases increase rate and the distance decay rate was –0.586 (95% CI, –0.742 to –0.370; P < .001) (eTable 3 in the Supplement). Figure 3A shows the state-level correlation between the increase coefficients of confirmed cases and the travel distance decay coefficients across the nation. The moderate negative correlation indicates that in the states where the confirmed cases were increasing faster, people generally reduced their daily travel distance more quickly. Figure 3B shows the state-level correlation between the increase coefficients of confirmed cases and the home dwell time increment coefficients across the nation. The increase rates and the home dwell time rates (eTable 4 in the Supplement) had a positive correlation of 0.526 (95% CI, 0.293 to 0.700; P < .001), which suggests that in states with higher case increase rates, home dwell time of residents in this state were generally longer. These association analyses found that there was statistically significant mobility reduction associated with the increase rate of COVID-19 cases and that people in most states reduced their daily travel distance and increased stay-at-home time. In addition, the statistical variation of the mobility measures can be largely explained (travel 2 2 distance: R = 0.59; P < .001; home dwell time: R = 0.69; P < 001) by socioeconomic factors, including state policies, race/ethnicity, population density, age groups, and median household JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 7/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US income (eAppendix 1, eTable 6, and eTable 7 in the Supplement). Recent studies have also identified partisan differences in individual responses to stay-at-home social distancing guidelines during the COVID-19 pandemic (H. Alcott et al, unpublished data, July 2020). Pandemic Doubling Time Changes The fitted curves by an exponential model and a power-law model are shown in eFigure 3 and eFigure 4 in the Supplement. For the exponential model before the statewide stay-at-home orders, initial estimates of the increase rates of the number of confirmed cases for the pandemic in each state were 0.17 to 0.70 per day with a doubling time of 1.3 to 4.3 days (median [IQR], 2.6 [2.1-2.9] days). A similar result was found by fitting the power-law model, in which initial estimates of the case rates before the orders in each state were 0.12 to 0.71 cases per day with a doubling time of 1.3 to 6.2 days (median [IQR], 2.7 [2.2-3.1] days). The finding aligned well with the doubling time of 2.3 to 3.3 days in the early pandemic epicenter in Wuhan, China. After the implementation of stay-at-home orders, the estimates of the case rate in each state by the exponential model were reduced to 0.03 to 0.21 cases per day, with a doubling time increased to 3.7 to 27.7 days (median [IQR], 5.7 [4.7-6.9] days). Similarly, the estimates of the case rate in each state by the power-law model were reduced to 0.02 to 0.17 cases per day, with a doubling time increased to 4.3 to 29.8 days (median [IQR], 6.3 [5.4-7.9] days). The finding also aligned well (measured by Jensen–Shannon Divergence) with the result from the observed epidemiological data (Table), in which the empirical case rate in each state was 0.11 to 0.95 cases per day with a doubling time of 1.0 to 6.9 days (median [IQR], 2.7 [2.3-3.3] days) before the statewide stay-at-home orders, and reduced to 0.02 to 0.21 per cases day with a doubling time increased to 3.7 to 30.3 days (median [IQR], 6.0 [4.8-7.1] days) after the orders. The curve fitting results also matched the outcomes of mechanistic epidemic models (eFigure 7 in the 18 19 Supplement), such as the models reported by Cori et al and Thompson et al. These models used confirmed cases and the serial interval, that is, the days between 2 successive infection cases. In addition, we investigated the overall probability density distribution of the doubling time nationwide before and after the stay-at-home orders using the state-level median doubling time (Figure 4A; eFigure 5 and eFigure 6 in the Supplement). The doubling time nationwide increased after the stay-at-home orders (empirical observations: from median [IQR] 2.7 [2.3-3.3] days to median 6.0 [4.8-7.1] days). Our combined results on doubling times suggest that stay-at-home orders were associated with reduction of the COVID-19 pandemic spread and with flattening the curve. Similar findings have also been reported in a study by Sen et al on the association of stay-at-home orders with COVID-19 hospitalizations. In addition, the ten-hundred plot (Figure 4B) also shows that the case increase rate in each of the top 5 states (ie, New York, New Jersey, Michigan, California, Figure 3. State-Level Correlation Between the Increase Coefficients of Confirmed Cases, Travel Distance Decay Coefficients, and Home Dwell Time Increase Coefficients A B Correlation between the increase in coefficients of confirmed Correlation between the increase in coefficients of confirmed cases cases and the travel distance decay coefficients and the home dwell time increment coefficients –0.15 1.90 1.85 –0.20 1.80 –0.25 1.75 –0.30 1.70 –0.35 1.65 –0.40 1.60 –0.45 1.55 –0.50 1.50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Confirmed cases, coefficient Confirmed cases, coefficient JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 8/13 Travel distance, coefficient Home dwell time, coefficient JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US and Massachusetts) slowed down after the stay-at-home orders (approaching subexponential growth). The statistical variation of the mobility measures can be largely explained (travel distance: 2 2 R = 0.59; P < .05; home dwell time: R = 0.69; P < .05) (eAppendix 1 in the Supplement)by socioeconomic factors, including state policies, race/ethnicity, population density, age groups, and median household income (eTable 6 and eTable 7 in the Supplement). Recent studies have also identified partisan differences in individual responses to stay-at-home social distancing guidelines during the COVID-19 pandemic (H. Alcott et al, unpublished data, July 2020). Discussion These findings suggest that stay-at-home social distancing mandates, when they were followed by measurable mobility changes, were associated with reduction in COVID-19 case rates. Great efforts have been made in scientific research communities on the study of human mobility patterns using 26-31 various emerging data sources, including anonymized mobile phone call detail records, social 32,33 34-38 media (eg, Twitter), location-based services, and mobile applications. During the COVID-19 pandemic, both individual-level and aggregated-level human mobility patterns have been found 6,13,39,40 useful in pandemic modeling and digital contact tracing. However, technical challenges (eg, 41-44 location uncertainty), socioeconomic and sampling bias, privacy and ethical concerns have been 45-48 expressed by national and international societies. Moving forward, research efforts should continue exploring the balance of using such human mobility data at different geographic scales for public health and social good while preserving individual privacy and rights. Limitations This study has some limitations. Potential confounding issues relate to other control measures, such as varying state-level quarantine protocols, availability of personal protective equipment, and timely testing, but the detailed information was not available, and the consistency of our results across most states makes such confounding less likely. In addition, the variability in the curve fitting estimated parameters was not accounted for the correlation analysis. There are variations in human behaviors and risk perception even within a state. All these factors contribute to the potential endogeneity of findings and the limitations. Figure 4. Probability Density Distributions and Ten-Hundred Plot of Coronavirus Disease 2019 Spread Before and After Stay-at-Home Orders A B State-level median doubling time State-level confirmed case rate 0.6 30 Before order Before order After order After order 0.5 25 Trajectory 0.4 20 0.3 15 California California 0.2 10 Massachusetts New York New Jersey Michigan 0.1 5 New Jersey New York Michigan Massachusetts 0 0 0 2 4 6 8 10 12 14 0 5 10 15 20 25 30 Doubling time, d Latest date with N cases and the date with cases, d B. The lower-right region represents subexponential growth; the diagonal line, exponential growth; and the upper left region, super-exponential growth. The top 5 states with the most confirmed cases are labeled and their change rate changes are visualized as trajectories. N indicates the number of coronavirus disease 2019 confirmed cases on that date. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 9/13 Density N N Date with cases and the date with cases, d 10 100 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Conclusions This cross-sectional study found a statistically significant association of 2 human mobility measures (ie, travel distance and stay-at-home time) with the rates of COVID-19 cases across US states. This study found a reduction of the spread of COVID-19 after stay-at-home social distancing mandates were enacted in most states. The findings come at a particularly critical period, when US states are beginning to reopen their economies but COVID-19 cases are surging. At such a time, our study suggests the efficacy of stay-at-home social distancing measures and could inform future public health policy making. ARTICLE INFORMATION Accepted for Publication: July 31, 2020. Published: September 8, 2020. doi:10.1001/jamanetworkopen.2020.20485 Open Access: This is an open access article distributed under the terms of the CC-BY License.©2020GaoSetal. JAMA Network Open. Corresponding Authors: Song Gao, PhD (song.gao@wisc.edu) and Jonathan A. Patz, MD (patz@wisc.edu), University of Wisconsin–Madison, 550 N Park St, Madison, WI 53706. Author Affiliations: GeoDS Lab, Department of Geography, University of Wisconsin–Madison, Madison (Gao, Rao, Kang, Liang, Kruse); School of Veterinary Medicine, University of Wisconsin–Madison, Madison (Dopfer, Mandujano Reyes); School of Medicine and Public Health, University of Wisconsin–Madison, Madison (Sethi, Patz); Statistics and American Family Insurance Data Science Institute, University of Wisconsin–Madison, Madison (Yandell). Author Contributions: Dr Gao had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Gao, Kang, Sethi, Yandell, Patz. Acquisition, analysis, or interpretation of data: Gao, Rao, Kang, Liang, Kruse, Dopfer, Sethi, Mandujano Reyes. Drafting of the manuscript: Gao, Rao, Kang, Liang, Kruse, Dopfer, Sethi, Mandujano Reyes, Patz. Critical revision of the manuscript for important intellectual content: Gao, Rao, Kang, Sethi, Yandell, Patz. Statistical analysis: Gao, Rao, Kang, Liang, Dopfer, Mandujano Reyes, Yandell. Obtained funding: Gao. Administrative, technical, or material support: Gao, Liang, Yandell. Supervision: Gao. Conflict of Interest Disclosures: Dr Gao reported receiving grants from National Science Foundation during the conduct of the study. No other disclosures were reported. Funding/Support: Drs Gao and Patz received funding from grant No. BCS-2027375 from the National Science Foundation. Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Disclaimer: The opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. REFERENCES 1. Centers for Disease Control and Prevention. COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc. gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html 2. Pan A, Liu L, Wang C, et al. Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 3. Hartley DM, Perencevich EN. Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 4. Lai S, Ruktanonchai NW, Zhou L, et al. Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020. 03.03.20029843 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 10/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US 5. Chinazzi M, Davis JT, Ajelli M, et al. The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 6. Tian H, Liu Y, Li Y, et al. An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 7. Gao S, Rao J, Kang Y, Liang Y, Kruse J. Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824 8. Warren MS, Skillman SW. Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 9. Zhang L, Ghader S, Pack ML, et al. An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472 10. Zhou C, Su F, Pei T, et al. COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10. 1016/j.geosus.2020.03.005 11. Centers for Disease Control and Prevention. Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html 12. Gao S, Mai G. Mobile GIS and location-based services. In: Huang B, COVA TJ, Tsou MH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397. 13. Buckee CO, Balsari S, Chan J, et al. Aggregated mobility data could help fight COVID-19. Science. 2020;368 (6487):145-146. doi:10.1126/science.abb8021 14. Centers for Disease Control and Prevention. Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm 15. GitHub. Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper 16. US Census Bureau. American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs- surveys/acs 17. US Census Bureau. TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019. Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation- u-s-2010-census-urban-area-national 18. Cori A, Ferguson NM, Fraser C, Cauchemez S. A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 19. Thompson RN, Stockwin JE, van Gaalen RD, et al. Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 20. Vynnycky E, White R. An Introduction to Infectious Disease Modelling. Oxford University Press; 2010. 21. Lin J. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37 (1):145-151. doi:10.1109/18.61115 22. Maier BF, Brockmann D. Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 23. Sanche S, Lin YT, Xu C, Romero-Severson E, Hengartner N, Ke R. High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607. 24. Sen S, Karaca-Mandic P, Georgiou A. Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 25. Zhu J. The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/ index.html 26. González MC, Hidalgo CA, Barabási AL. Understanding individual human mobility patterns. Nature. 2008;453 (7196):779-782. doi:10.1038/nature06958 27. Song C, Qu Z, Blumm N, Barabási AL. Limits of predictability in human mobility. Science. 2010;327(5968): 1018-1021. doi:10.1126/science.1177170 28. Kang C, Ma X, Tong D, Liu Y. Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005 29. Gao S. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300 30. Yuan Y, Raubal M. Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 11/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US 31. Pullano G, Valdano E, Scarpa N, Rubrichi S, Colizza V. Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097 32. Huang Q, Wong DW. Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225 33. Huang X, Li Z, Jiang Y, Li X, Porter D. Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020. 34. Prestby T, App J, Kang Y, Gao S. Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911 35. Liang Y, Gao S, Cai Y, Foutz NZ, Wu L. Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624 36. Wu L, Zhi Y, Sui Z, Liu Y. Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 37. Shaw SL, Tsou MH, Ye X. Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9): 1687-1693. doi:10.1080/13658816.2016.1164317 38. Kang Y, Zhang F, Peng W, et al Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919 39. Ferretti L, Wymant C, Kendall M, et al. Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 40. Li R, Pei S, Chen B, et al. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 41. Zhao Z, Shaw SL, Xu Y, Lu F, Chen J, Yin L. Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298 42. Jiang J, Li Q, Tu W, Shaw SL, Yue Y. A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816. 2018.1552964 43. Xu Y, Belyi A, Bojic I, Ratti C. Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001 44. Li M, Gao S, Lu F, Zhang H. Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346 45. de Montjoye YA, Hidalgo CA, Verleysen M, Blondel VD. Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 46. Tsou MH. Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251 47. McKenzie G, Keßler C, Andris C. Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10. 5311/JOSIS.2019.19.608 48. de Montjoye YA, Gambs S, Blondel V, et al. On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 49. Jewell NP, Lewnard JA, Jewell BL. Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 SUPPLEMENT. eAppendix 1. Supplementary Materials and Methods eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients eFigure 2. Linear Fit for Distance vs Days eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 12/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia eTable 6. Regression Results of Travel Distance Changes at States eTable 7. Regression Results of Dwell Time at Home at States eAppendix 2. Mapping Videos eReferences. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 13/13 Supplementary Online Content Gao S, Rao J, Kang Y, et al. Association of mobile phone location data indications of travel and stay-at- home mandates with COVID-19 infection rates in the US. JAMA Netw Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 eAppendix 1. Supplementary Materials and Methods eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients eFigure 2. Linear Fit for Distance vs Days eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March 31 eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia eTable 6. Regression Results of Travel Distance Changes at States eTable 7. Regression Results of Dwell Time at Home at States eAppendix 2. Mapping Videos eReferences This supplementary material has been provided by the authors to give readers additional information about their work. © 2020 Gao S et al. JAMA Network Open. eAppendix 1. Supplementary Materials and Methods Travel Distance Mobility Data and Home Dwell Time Data The travel distance mobility data are collected from an open-source repository released by the Descartes Labs [10], while the home dwell time data are collected from SafeGraph [1]. Both datasets are collected through tracking usages of mobile phone apps of sampled users. By aggregating large-scale (over 45 million) anonymized location data from smartphones, the average individual movement pattern for a given area can be derived. To enhance privacy, individual data are de-identified and aggregated and all analyses are performed at aggregated spatial units (e.g., census block groups, counties, and states). Travel distance mobility changes are derived using an index which measures the median of the maximum travel distances for all individuals in a given county on a given day [10, 3]. Such data are widely used to represent the dramatic human mobility changes in reaction to the COVID-19 [6, 3, 11]. To measure home dwell time, the home place for each individual is identified and the minutes for all sampled devices staying at that home place across the day are summed up. Median home dwell time for all observed devices is then aggregated in geographical units. Fitting the epidemic growth curves As the confirmed cases of COVID-19 keep growing, we aim to find a model that can describe such growth rates with respect to their temporal changes in different states. Therefore, we use four types of formulas to fit the curve of the confirmed cases for the COVID-19 epidemic. By fitting the curve, we can compare the growth rates among different states using the coefficients estimated from the model. The four models we tested are: yc(t) = t + k (1) yc(t) = t (2) yc(t) = at + k (3) bt yc(t) = ae (4) Where y is the total number of confirmed cases in each state as a function of time, t is the number of days from March 11 where the novel Coronavirus disease became a pandemic, a,b,k are parameters we will estimate. We fit the curve for two time periods: March 11 to March 31 and March 11 to April 10 with consideration of the starting date variation of stay-at-home orders in each state (in Table S5) and the incubation period ranging from 1-14 days with a median of 4 days according to clinical characteristics of COVID-19 patients [5], as well as the testing capacity at the beginning and the variability of exponential or sub-exponential growth of the COVID- 19 cases [8]. We analyzed two periods respectively. The fitting results for the period March 11 to March 31 fit better due to the fact that the majority changes of human responses in terms of mobility and home dwell time were observed after March 23, 2020 and with a short time lag across different states. The results of fitting the curve using the four formulas from March 11 to March 31 are listed in Table S1. The coefficient b is used to reflect the growth rate of the confirmed cases. We use the Mean Square Error (MSE) as the goodness of fit measure for the curve fitting. Based on the MSE, the third formula y (t) = at +k has the best fit curve for total number of confirmed cases in each state. The first formula and the fourth one have close goodness of fit and the second formula has the worst performance. We then looked at the coefficients for each state and Table S2 shows the top 10 states with the largest coefficients from each formula. A larger coefficient reflects a faster growth rate, so the top states in the Table S2 should be those that experienced the most rapid outbreaks of the COVID-19. When examining the top most infected states for the four models, the results for b b yc(t) = t + k and yc(t) = t better match the empirical observations in the study period. The top states from the two models that have the fastest growth rate are New York, New Jersey, California, and Michigan. Therefore, we use the coefficients generated from the approximate sub-exponential growth function y (t) = t + k as the indices representing the epidemic growth rate. Figure S1 shows the cases and the fitted curves using formulas yc(t) = t + k. © 2020 Gao S et al. JAMA Network Open. Fitting the mobility change rates In addition to fitting the epidemic growth curve, we also fit the curve for travel distance and home dwell time changes. More specifically, we calculate the coefficient that describes the overall trend of the median of max- distance individual mobility at each state since March 11, 2020. The coefficient found by fitting the distance curve is used to measure the speed of changes to daily mobility patterns. The analysis is conducted for each state and we fit the distance decay pattern over days from March 11 using a linear regression model: yd(t) = bt + k (5) Where y is travel distance as a function of time, and t is the number of days from March 11, 2020. The coefficient of the travel distance change is represented using the slope parameter b. In addition, we processed the travel distance data to cut off the days where the distances are smaller than 0.1 km for 3 consecutive days. The reason to do so is that the plot of distance vs. days has a long tail for some areas because most people stayed at home in late March in response to COVID-19, known as the social distancing inertia [4]. Since we try to capture the most dramatic change of potential travel distance decrease in response to the COVID- .1 km for three consecutive days. An example of fitting the distance data with and .1 km is shown in Figure S2. For some states that responded quickly and reduced the travel distance, we can capture the quick decreasing travel distance in this way. Table S3 shows the top five states with the largest absolute distance decreasing rates. From this analysis, Michigan, Washington D.C., and New York, respectively, are the states that responded quickest to COVID-19 in terms of decreased mobility. Furthermore, after we estimate the growth rate of confirmed cases and the decreasing rate of travel distance coefficient between the two coefficients using the data from March 11 to March 31 is -0.586 with a p-value <0.001, and the 95% confidence interval (CI) is -0.742 to -0.370. Figure 3A shows the two coefficients for all 50 states and DC. This negative relationship indicates that people have responded quickly to increases in confirmed cases. In places where COVID-19 cases are growing faster, people usually are quicker to reduce mobility and stay at home. We also describe the relationship between home dwell time (in minutes) and the number of confirmed cases by estimating the coefficients that represent the change of the home dwell time and compute the correlation between the dwell time coefficient and the cases coefficient. The following three types of formulas are used to fit the curve of home dwell time (y) vs. the days from March 11 (t). y(t) = t + k (6) bt (7) y(t) = ae y(t) = bt + k (8) Table S4 shows the results of fitting home dwell time. The coefficients represent human mobility changes from the perspective of how long people stay at home. The three formulas all have positive coefficients, meaning that as time goes by, home dwell time is increasing. The three formulas all have close MSE and we choose the linear fit y(t) = bt + k to represent the rate of change in home dwell time as we prefer a simpler formula when all formulas perform similarly. We then compute the correlation between the home dwell time coefficients and the confirmed cases coefficients to detect whether such mobility changes are associated with the growth of COVID-19 cases. Figure 3B shows that the growth rates of cases in each state and the associated dwell time fitted coefficients have a positive correlation of 0.526 (95% CI: [0.293, 0.700], p-value < 0.001). In addition, the correlation between the cases growth coefficients and the distance decay coefficients from the period March 11 to April 10 is very similar to the result of that for the period March 11 to March 31: -0.582 (95% CI: [-0.740, -0.365], p-value <0.001). Because we processed the distance data to cut off the days where the distance is < 0.1km and we can capture the distance decay rate even when the study period extended. For the correlation between the the cases growth coefficients and the home dwell time coefficients, however, the © 2020 Gao S et al. JAMA Network Open. correlation is 0.322 (95% CI: [0.051, -0.549], p-value <0.05), it is much smaller than the correlation we calculated for the period March 11 to March 31. Evaluating factors influencing changes in travel distance and home dwell time To understand how socio-economic factors influence changes in travel distance and home dwell time, we employed a multi-linear regression model involving the above-mentioned socio-economic variables as independent variables to model the mobility changes. The previously regressed change rates of travel distance b (Equation 5) and home dwell time b (Equation 8) are taken as the dependent variable of the following model: d t b(d,t) = a0 + a1x1 + a2x2 + a3x3 + ... + anxn (9) where x1,x2,...,xn are socio-economic factors that may influence these behaviour changes; and a0,a1,...,an are coefficients which illustrate to what degree these independent variables contribute to the behaviour changes. Since spatial scale plays an important role in socio-economic metrics, we aggregated and analyzed data at the state level. Although we perform the linear regression in two time periods, from March 11 to March 31, and from March 11 to April 10, only the results of former one are reported. Because there were more substantial mobility changes (both travel distance and home dwell time) during the former time period. It is worth noting that, at the state level, records with median travel distance less than 0.1 km are removed. In addition, in terms of different age groups and race groups, the proportion of population over age 65 and the proportion of other race groups are assigned as the reference group in the experiments. Table S6 shows the regression results for travel distance changes at the state level. The goodness-of-fit of the regression models is evaluated using R-squared (R ) and p-value. The R-squared is 0.59 for the regression, and the model is statistically significant (p-value < 0.05). Results show that behaviour changes can be largely explained by state policies and socio-economic variables. Further investigation into other confounding factors associated with travel distance reduction is necessary to quantify the effects of those covariates. Table S7 illustrates the regression results for home dwell time changes in response to socio-economic factors. The R-squared for multi-linear regression is 0.66 at state-level and the p-value is less than 0.05, which indicates that the results are statistically significant. In particular, the ratio of Asians has a positive standardized coefficient (6.959), p-value < 0.05, while the ratio of Hawaiians has a negative standardized coefficient (-2.863), p-value < 0.05, and both are significant at the state level. Accordingly, it can be inferred that the ratio of Asians is strongly correlated with longer home dwell time at the state level. The higher the proportion of Asian population, the longer the median home dwell time of residents in that state. Calculating the epidemic growth doubling time Based on the relationship between the number of confirmed cases and the date, we calculate the doubling time of the number of total confirmed cases (i.e., the time it takes for the confirmed cases to double in size) to reflect the characteristics of the COVID-19 epidemic growth. Additionally, we want to explore how the social distancing policy (e.g., stay-at-home orders) in each state makes a difference in flattening the curve. That is, the larger the doubling time, the less steep the epidemic growth curve. At the time of writing, the growth rates of COVID-19 cases in the U.S. states are either exponential or sub-exponential. We implement both of them to fit the curve for calculating the doubling time. We also calculated the doubling time based on empirical observations (model- free) to further explore how the doubling time differs in these methods. The two models we used are the exponential model (Equation 11) and the power-law model (Equation 12): at yc(t) = I(0)e (10) y (t) = bt + I(0) (11) Where yc is the total number of confirmed cases in each state; t represents the number of the days after the first case was confirmed in the state; I(0) represents the number of confirmed cases on the first-case day; a,b are the © 2020 Gao S et al. JAMA Network Open. coefficients of the models. Note that we introduced the intercept term I(0) into the two models to ensure that the initial case number on the first day is accurate. This is important since the accuracy of the thereafter calculated daily growth rate and doubling time relies on the number of initial cases. We investigate how stay-at-home orders in each state associated with the doubling time of confirmed cases in that state. Within the time frame of our study, as shown in Table S5, there were five states (North Dakota, ed a stay-at-home order, and three states (Oklahoma, Utah, and Wyoming) which only had partial orders issued locally by cities or counties. All other states, as well as the District of Columbia, issued the order for all residents. South Carolina issued partial orders effective March 26, and then issued the full order effective April 7. We focus on the confirmed case data from March 11 to April 10, and we use the effective date of the stay-at-home order to split the confirmed case data into two parts: before the order and after the order. For the states with only partial orders, we use the earliest order effective date in that state. For South Carolina, we also use the earliest effective date (i.e., March 26). We investigate the doubling time in the states with statewide orders (except for Missouri since the effective date of the stay-at- -order data in this state given the date range of our study) and the District of Columbia. We fit the models on the data before the order and after the order respectively, and then we calculate the doubling time of the confirmed cases based on the two models. The doubling time of the confirmed cases is defined as: (12) Where d(t) stands for the doubling time of the cumulative infection cases on date t in each state; r(t) represents the growth rate of the cumulative infection cases on date t in each state, and is defined as: (13) Where I(t) represents the number of cumulative infection cases on date t. For the exponential model, the daily growth rate is constant, and thus the doubling time is constant and can be calculated as: (14) Where d is the doubling time of the exponential model, and a is the exponent parameter in the exponential model (Equation 10). Note that Equation (14) can be regarded as a special case of Equation (12). For the power- law model (Equation 11), since the doubling time is not constant, we first manually calculated the time- dependent growth rate of the predicted cases by the model in each day for each state. Specifically, for each date icted case number, thereby obtaining the time- dependent daily growth rate r(t) (Equation 13). We then use the time-dependent growth rate to calculate the corresponding time-dependent doubling time of the power-law model, and we use the median of the time- dependent doubling time to represent the doubling time in that state during the before or after order period. For the two curve fitting models, the fitting results are shown in Figures S3 and S4. The green dashed line in each subplot represents the fitted curve on the data before the stay-at-home order in the state; the blue curve represents the fitted curve on the data after the stay-at-home order in the state; the vertical black dashed line indicates the date when the stay-at-home order takes effect in each state. All the states showed a large initial growth rate and small doubling time in the number of total confirmed cases. However, the growth starts to slow down in general after the effective date of the stay-at-home order in each state, which shows the efficiency of the social distancing policy in suppressing the transmission of the novel coronavirus. For empirical observations, we manually calculated the time-dependent growth rate of the reported cumulative confirmed cases in each day for each state. The empirical growth rate is also calculated following the same process as the power-law model. For each date except for the starting date, we subtract the previous divided by the © 2020 Gao S et al. JAMA Network Open. -dependent daily empirical growth rate. We then use the empirical time-dependent growth rate to calculate the corresponding empirical time- dependent doubling time, and we use the median of the time-dependent doubling time to represent the empirical doubling time in that state. In addition, we visualize and investigate the overall probability density distribution of the median doubling time before and after the order in each state to have a better understanding of the overall changes in the epidemic growth nationwide. We applied Kernel Density Estimation (KDE) on doubling times in all the states to derive the overall probability density distribution, where the doubling time for each state is calculated by the median of the daily growth rate in that state. Note that the doubling time for the exponential model is a constant, so the median equals that constant. As is shown in Figure 4A, S5 and S6, the doubling time of confirmed cases nationwide has significantly increased after the order in each state (empirical observations: from median 2.7 days, IQR: 1.0 to median 6.0 days, IQR: 2.3; exponential model: from median 2.6 days, IQR: 0.8 to median 5.7 days, IQR: 2.2; power-law model: from median 2.7 days, IQR: 0.9 to median 6.3 days, IQR: 2.5)), which reinforce the point that stay-at-home orders, or social distancing policy, associated with the reduced spread of the virus when they are compliant. Furthermore, we measure the difference in probability density distribution between the fitting results and the real data using Jensen Shannon Divergence (JSD). The JSD value ranges from 0 to 1 for two probability distributions; small JSD values indicate high similarity between two probability distributions. Compared with the probability density distribution of real data shown in Figure 4A, both the fitting results by the exponential model and power-law model show high similarity (i.e., with small JSD values) to the empirical data before the order (the exponential model: 0.11; the power-law model: 0.07) and after the order (the exponential model: 0.17; the power-law model: 0.18), respectively. This means that, despite different models, the discoveries are consistent: stay-at-home orders did associate with the slowdown of the COVID-19 case growth and with the increase of doubling time. Mechanistic prediction models The exponential equation approach was particularly suitable during the early outbreak phase but the sub- exponential growth fitted better after the city lock-downs [8], stay-at-home orders and following the social distancing regulations. Our curve fitting results matched the outcomes of mechanistic prediction models (Figure S7), such as the models reported by [2, 9]. The daily number of cumulative confirmed cases per U.S. state and the serial interval (SI), that is the number of days between two consecutive cases, were used assuming an average SI of 7.5 days with standard deviations (SD): 3.4 days as reported by [7], who estimated the SI using Bayesian parametric estimation from data in the city of Wuhan in China. In addition, the SI average and SD were sampled from normal distributions of mean 7.5 and SD 0.2, as well as SD mean of 3.4 and SD of 0.4 to sample from more SI distributions at random. Estimates of the instantaneous basic reproduction number (R0) were simulated by moving an 8-day sliding time interval across all data points since the start of the outbreak per state respectively. The instantaneous R0 on the last day before the start of the stay-at-home orders and from April 10, 2020, that is after the order went in, were used for projections of the confirmed number of cases per day. The resulting graphs are shown in Figure S7 for all states respectively where black dots represent observed values and the red dots represent the projected values for daily total confirmed cases. It shows that projected daily total confirmed cases from before the start date of the stay-at-home social distancing orders increase much more rapidly compared to the projected number of total confirmed cases starting April 10, 2020, that is after the orders went in. In summary, the mathematical curve fitting models and the mechanistic epidemic prediction models draw the same conclusion. © 2020 Gao S et al. JAMA Network Open. eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients A: New York; B: New Jersey; C: California; D: Michigan; E: Massachusetts. © 2020 Gao S et al. JAMA Network Open. eFigure 2. Linear Fit for Distance vs Days (a) (b) (a) using the original .1 km in New York state from March 11 to March 31. © 2020 Gao S et al. JAMA Network Open. eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State The green dashed line and the blue line represent the fitted curves on the data before and after the stay-at- home orders in each state, respectively; the vertical black dashed line indicates the effective date of the stay-at- home orders in each state. dt and dt represent the median doubling time before and after the order in before after each state. © 2020 Gao S et al. JAMA Network Open. eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State The green dashed line and the blue line represent the fitted curves on the data before and after the stay-at- home order in each state, respectively; the vertical black dashed line indicates the effective date of the stay-at- home order in each state. dt and dt represent the median doubling time before and after the order in before after each state. © 2020 Gao S et al. JAMA Network Open. eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model Figure S5: The overall changes of the doubling time nationwide before and after the order using the exponential fitting model. © 2020 Gao S et al. JAMA Network Open. eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model © 2020 Gao S et al. JAMA Network Open. eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders © 2020 Gao S et al. JAMA Network Open. eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 The fitting method Coefficient b (mean) MSE yc = t + k 2.284 2.694E+05 yc = t 2.259 3.779E+05 y = at + k 3.210 4.336E+04 bt yc = ae 0.191 2.960E+05 © 2020 Gao S et al. JAMA Network Open. eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March 31 b b b bt y = t + k yc = t y = at + k yc = ae c c New York New York Rhode Island Rhode Island New Jersey New Jersey Idaho Idaho California California Indiana Indiana Michigan Michigan Pennsylvania Pennsylvania Massachusetts Florida Massachusetts New Jersey Florida Massachusetts New Jersey Missouri Illinois Washington Maryland Massachusetts Louisiana Illinois Iowa Texas Washington Louisiana Missouri Maryland Pennsylvania Pennsylvania Texas Michigan © 2020 Gao S et al. JAMA Network Open. eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 State Distance coefficient Michigan -0.494 Washington, D.C. -0.485 New York -0.478 Missouri -0.452 Delaware -0.450 © 2020 Gao S et al. JAMA Network Open. eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home The fitting method Coefficient b MSE y = t + k 1.677 4850.671 bt y = ae 0.767 4791.662 y = bt + k 8.608 4864.841 © 2020 Gao S et al. JAMA Network Open. eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia State Name Order Type Earliest Effective Date California Full 3/19/2020 Illinois Full 3/21/2020 New Jersey Full 3/21/2020 New York Full 3/22/2020 Connecticut Full 3/23/2020 Louisiana Full 3/23/2020 Ohio Full 3/23/2020 Oregon Full 3/23/2020 Washington Full 3/23/2020 Delaware Full 3/24/2020 Indiana Full 3/24/2020 Massachusetts Full 3/24/2020 Michigan Full 3/24/2020 New Mexico Full 3/24/2020 West Virginia Full 3/24/2020 Hawaii Full 3/25/2020 Idaho Full 3/25/2020 Oklahoma Partial 3/25/2020 Vermont Full 3/25/2020 Wisconsin Full 3/25/2020 Colorado Full 3/26/2020 Kentucky Full 3/26/2020 South Carolina* Partial, Full 3/26/2020, 4/7/2020 Minnesota Full 3/27/2020 New Hampshire Full 3/27/2020 Utah Partial 3/27/2020 Alaska Full 3/28/2020 Montana Full 3/28/2020 Rhode Island Full 3/28/2020 Wyoming Partial 3/28/2020 Kansas Full 3/30/2020 Maryland Full 3/30/2020 North Carolina Full 3/30/2020 Virginia Full 3/30/2020 Arizona Full 3/31/2020 Tennessee Full 3/31/2020 Washington, D.C. Full 4/1/2020 Nevada Full 4/1/2020 Pennsylvania Full 4/1/2020 Maine Full 4/2/2020 Texas Full 4/2/2020 Florida Full 4/3/2020 Georgia Full 4/3/2020 Mississippi Full 4/3/2020 Alabama Full 4/4/2020 Missouri Full 4/6/2020 North Dakota No - South Dakota No - Nebraska No - Iowa No - © 2020 Gao S et al. JAMA Network Open. Arkansas No - -at-home order in this state. *South Carolina issued partial orders effective March 26, and then issued the full order effective April 7. We use the earliest effective date for South Carolina. © 2020 Gao S et al. JAMA Network Open. eTable 6. Regression Results of Travel Distance Changes at States Variables Coefficients Standard Error Standardized Coefficients Intercept 1.116 2.513 0.444 0.660 -0.364 Population 0.000 0.000 -1.688 0.100 -0.022 Median Age -0.025 0.029 -0.853 0.399 -0.053 Population Density 0.000 0.000 -1.177 0.247 -0.026 Household Income 0.000 0.000 1.042 0.304 0.014 Proportion of Population under Age 18 -2.438 2.801 -0.870 0.390 -0.051 Proportion of Population between Age 18 to 44 -1.680 2.133 -0.788 0.436 -0.041 Proportion of Population between Age 45 to 64 -3.147 1.670 -1.884 0.067 -0.057 Proportion of Population over Age 65 0.000 0.000 Proportion of Whites 1.029 0.623 1.652 0.107 0.140 Proportion of Blacks 0.676 0.583 1.160 0.253 0.073 Proportion of Asians 2.786 1.399 1.991 0.054 0.153 Proportion of Natives 2.473 0.954 2.593 0.014 0.066 Proportion of Hawaiian -3.191 2.250 -1.419 0.164 -0.046 Proportion of Other Race Groups 0.000 0.000 1.024 0.961 1.065 0.294 0.023 R : 0.59, p-value < 0.05 © 2020 Gao S et al. JAMA Network Open. eTable 7. Regression Results of Dwell Time at Home at States Variables Coefficients Standard Error Standardized Coefficients Intercept 61.218 70.426 0.869 0.390 8.814 Population 0.000 0.000 -1.476 0.148 -0.544 Median Age -0.716 0.824 -0.868 0.391 -1.508 Population Density 0.003 0.002 1.327 0.193 0.825 Household Income -0.007 0.008 -0.907 0.370 -0.351 Proportion of Population under Age 18 -39.540 78.514 -0.504 0.618 -0.825 Proportion of Population between Age 18 to 44 -79.376 59.786 -1.328 0.192 -1.925 Proportion of Population between Age 45 to 64 -7.773 46.809 -0.166 0.869 -0.142 Proportion of Population over Age 65 0.000 0.000 Proportion of Whites 26.084 17.457 1.494 0.144 3.551 Proportion of Blacks 25.046 16.326 1.534 0.134 2.715 Proportion of Asians 126.581 39.213 3.228 0.003 6.959 Proportion of Natives 25.069 26.732 0.938 0.354 0.672 Proportion of Hawaiian -200.726 63.055 -3.183 0.003 -2.863 Proportion of Other Race Groups 0.000 0.000 -20.269 26.941 -0.752 0.457 -0.449 R : 0.66, p-value < 0.05 © 2020 Gao S et al. JAMA Network Open. eAppendix 2. Mapping Videos (1) Mapping the COVID-19 infected areas in the U.S., available at: "https://geods.geography.wisc.edu/wp-content/uploads/2020/04/US_cases_animation_March.mov" (2) Mapping human mobility changes at the U.S. county level, available at: "https://geods.geography.wisc.edu/wp-content/uploads/2020/04/US_mobilitychanges_animation. mp4" © 2020 Gao S et al. JAMA Network Open. eReferences [1] SafeGraph Inc. Accessed April 12, 2020, https://www.safegraph.com/. [2] A. Cori, N. M. Ferguson, C. Fraser, and S. Cauchemez. A new framework and software to estimate timevarying reproduction numbers during epidemics. American Journal of Epidemiology, 178(9):1505 1512, 2013. [3] S. Gao, J. Rao, Y. Kang, Y. Liang, and J. Kruse. Mapping county-level mobility pattern changes in the united states in response to covid-19. SIGSPATIAL Special, 12(1):16 26, 2020. [4] S. Ghader, J. Zhao, M. Lee, W. Zhou, G. Zhao, and L. Zhang. Observed mobility behavior data reveal social distancing inertia. arXiv preprint arXiv:2004.14748, 2020. [5] W.-j. Guan, Z.-y. Ni, Y. Hu, W.-h. Liang, C.-q. Ou, J.-x. He, L. Liu, H. Shan, C.-l. Lei, D. S. Hui, et al. Clinical characteristics of coronavirus disease 2019 in China. New England Journal of Medicine, 2020. [6] J. R. Hipp and A. Boessen. The shape of mobility: Measuring the distance decay function of household mobility. The Professional Geographer, 69(1):32 44, 2017. [7] R. Li, S. Pei, B. Chen, Y. Song, T. Zhang, W. Yang, and J. Shaman. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science, 2020. [8] B. F. Maier and D. Brockmann. Effective containment explains sub-exponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science, 2020. [9] R. Thompson, J. Stockwin, R. van Gaalen, J. Polonsky, Z. Kamvar, P. Demarsh, E. Dahlqwist, S. Li, E. Miguel, T. Jombart, et al. Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics, 29:100356, 2019. [10] M. S. Warren and S. W. Skillman. Mobility changes in response to covid-19. arXiv preprint arXiv:2003.14228, 2020. [11] L. Zhang, S. Ghader, M. L. Pack, C. Xiong, A. Darzi, M. Yang, Q. Sun, A. Kabiri, and S. Hu. An interactive covid- 19 mobility impact and social distancing analysis platform. medRxiv, 2020. © 2020 Gao S et al. JAMA Network Open. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JAMA Network Open American Medical Association http://www.deepdyve.com/lp/american-medical-association/association-of-mobile-phone-location-data-indications-of-travel-and-dzad0WzF0y

Loading next page...

References (102)

M. Lemmens (2011)
Mobile GIS and Location-Based Services
Timothy Prestby, J. App, Yuhao Kang, Song Gao (2019)
Understanding neighborhood isolation through spatial interaction network analysis using location big data
Environment and Planning A: Economy and Space, 52
B. Maier, D. Brockmann (2020)
Effective containment explains subexponential growth in recent confirmed COVID-19 cases in China
Science (New York, N.y.), 368
G. Pullano, E. Valdano, Nicola Scarpa, Stefania Rubrichi, V. Colizza (2020)
Population mobility reductions during COVID-19 epidemic in France under lockdown
medRxiv
Ruiyun Li, S. Pei, Bin Chen, Yimeng Song, Tao Zhang, Wan Yang, J. Shaman (2020)
Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2)
Science (New York, N.y.), 368
Sepehr Ghader, Jun Zhao, Minha Lee, Weiyi Zhou, Guangchen Zhao, Lei Zhang (2020)
Observed mobility behavior data reveal social distancing inertia
ArXiv, abs/2004.14748
(JewellNP, LewnardJA, JewellBL Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 32297897)
JewellNP, LewnardJA, JewellBL Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 32297897
JewellNP, LewnardJA, JewellBL Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 32297897, JewellNP, LewnardJA, JewellBL Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 32297897
Yunlei Liang, Song Gao, Yuxin Cai, N. Foutz, L. Wu (2020)
Calibrating the dynamic Huff model for business analysis using location big data
Transactions in GIS, 24
Yang Xu, Alexander Belyi, I. Bojic, C. Ratti (2018)
Human mobility and socioeconomic status: Analysis of Singapore and Boston
Comput. Environ. Urban Syst., 72
(GaoS, MaiG. Mobile GIS and location-based services In: HuangB, COVATJ, TsouMH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397.)
GaoS, MaiG. Mobile GIS and location-based services In: HuangB, COVATJ, TsouMH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397.
GaoS, MaiG. Mobile GIS and location-based services In: HuangB, COVATJ, TsouMH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397., GaoS, MaiG. Mobile GIS and location-based services In: HuangB, COVATJ, TsouMH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397.
A. Cori, N. Ferguson, C. Fraser, S. Cauchemez (2013)
A New Framework and Software to Estimate Time-Varying Reproduction Numbers During Epidemics
American Journal of Epidemiology, 178
(US Census Bureau American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs-surveys/acs)
US Census Bureau American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs-surveys/acs
US Census Bureau American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs-surveys/acs, US Census Bureau American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs-surveys/acs
S. Sen, P. Karaca-Mandic, A. Georgiou (2020)
Association of Stay-at-Home Orders With COVID-19 Hospitalizations in 4 States.
JAMA
A. Pan, Li Liu, Chaolong Wang, Huan Guo, Xingjie Hao, Qi Wang, Jiao Huang, N. He, Hongjie Yu, Xihong Lin, Sheng Wei, Tangchun Wu (2020)
Association of Public Health Interventions With the Epidemiology of the COVID-19 Outbreak in Wuhan, China.
JAMA
Census Urban Area National
Yuhao Kang, Yuhao Kang, Fan Zhang, Wenzhe Peng, Song Gao, Jinmeng Rao, Fábio Duarte, Fábio Duarte, C. Ratti (2020)
Understanding house price appreciation using multi-source big geo-data and machine learning
Land Use Policy
B. Huang, K. Cao, Ea Silva (2017)
Comprehensive Geographic Information Systems
H. Tian, Yonghong Liu, Yidan Li, Chieh-Hsi Wu, Bin Chen, M. Kraemer, Bingying Li, Jun Cai, Bo Xu, Qiqi Yang, Ben Wang, P. Yang, Yujun Cui, Yimeng Song, P. Zheng, Quanyi Wang, O. Bjørnstad, Ruifu Yang, B. Grenfell, O. Pybus, C. Dye (2020)
An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China
Science (New York, N.y.), 368
Grant McKenzie, C. Kessler, Clio Andris (2019)
Geospatial Privacy and Security
J. Spatial Inf. Sci., 19
(2020)
Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2)
MobilePhoneLocationDataIndicationsofTravelandStay-at-HomeMandatesandCOVID-19InfectionRatesintheUS )
Mingxiao Li, Song Gao, F. Lu, Hengcai Zhang (2019)
Reconstruction of human movement trajectories from large-scale low-frequency mobile phone data
Comput. Environ. Urban Syst., 77
(KangC, MaX, TongD, LiuY Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005 )
KangC, MaX, TongD, LiuY Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005
KangC, MaX, TongD, LiuY Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005 , KangC, MaX, TongD, LiuY Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005
(Centers for Disease Control and Prevention Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html)
Centers for Disease Control and Prevention Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html
Centers for Disease Control and Prevention Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html, Centers for Disease Control and Prevention Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html
Matteo Chinazzi, Jessica Davis, M. Ajelli, C. Gioannini, M. Litvinova, S. Merler, A. Piontti, K. Mu, L. Rossi, K. Sun, C. Viboud, X. Xiong, Hongjie Yu, M. Halloran, I. Longini, A. Vespignani (2020)
The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak
Science (New York, N.y.), 368
(HuangQ, WongDW Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225 )
HuangQ, WongDW Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225
HuangQ, WongDW Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225 , HuangQ, WongDW Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225
(GonzálezMC, HidalgoCA, BarabásiAL Understanding individual human mobility patterns. Nature. 2008;453(7196):779-782. doi:10.1038/nature06958 18528393)
GonzálezMC, HidalgoCA, BarabásiAL Understanding individual human mobility patterns. Nature. 2008;453(7196):779-782. doi:10.1038/nature06958 18528393
GonzálezMC, HidalgoCA, BarabásiAL Understanding individual human mobility patterns. Nature. 2008;453(7196):779-782. doi:10.1038/nature06958 18528393, GonzálezMC, HidalgoCA, BarabásiAL Understanding individual human mobility patterns. Nature. 2008;453(7196):779-782. doi:10.1038/nature06958 18528393
(ChinazziM, DavisJT, AjelliM, The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 32144116)
ChinazziM, DavisJT, AjelliM, The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 32144116
ChinazziM, DavisJT, AjelliM, The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 32144116, ChinazziM, DavisJT, AjelliM, The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 32144116
(ZhangL, GhaderS, PackML, An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472)
ZhangL, GhaderS, PackML, An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472
ZhangL, GhaderS, PackML, An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472, ZhangL, GhaderS, PackML, An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472
(BuckeeCO, BalsariS, ChanJ, Aggregated mobility data could help fight COVID-19. Science. 2020;368(6487):145-146. doi:10.1126/science.abb802132205458)
BuckeeCO, BalsariS, ChanJ, Aggregated mobility data could help fight COVID-19. Science. 2020;368(6487):145-146. doi:10.1126/science.abb802132205458
BuckeeCO, BalsariS, ChanJ, Aggregated mobility data could help fight COVID-19. Science. 2020;368(6487):145-146. doi:10.1126/science.abb802132205458, BuckeeCO, BalsariS, ChanJ, Aggregated mobility data could help fight COVID-19. Science. 2020;368(6487):145-146. doi:10.1126/science.abb802132205458
Ziliang Zhao, S. Shaw, Yang Xu, F. Lu, Jie Chen, Ling Yin (2016)
Understanding the bias of call detail records in human mobility research
International Journal of Geographical Information Science, 30
C. Buckee, S. Balsari, J. Chan, M. Crosas, F. Dominici, Urs Gasser, Y. Grad, B. Grenfell, M. Halloran, M. Kraemer, M. Lipsitch, C. Metcalf, L. Meyers, T. Perkins, M. Santillana, S. Scarpino, C. Viboud, A. Wesolowski, A. Schroeder (2020)
Aggregated mobility data could help fight COVID-19
Science, 368
S. Shaw, Ming-Hsiang Tsou, X. Ye (2016)
Editorial: human dynamics in the mobile and big data era
International Journal of Geographical Information Science, 30
(SancheS, LinYT, XuC, Romero-SeversonE, HengartnerN, KeR High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607.200282 32255761)
SancheS, LinYT, XuC, Romero-SeversonE, HengartnerN, KeR High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607.200282 32255761
SancheS, LinYT, XuC, Romero-SeversonE, HengartnerN, KeR High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607.200282 32255761, SancheS, LinYT, XuC, Romero-SeversonE, HengartnerN, KeR High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607.200282 32255761
Jianhua Lin (1991)
Divergence measures based on the Shannon entropy
IEEE Trans. Inf. Theory, 37
David Hartley, E. Perencevich (2020)
Public Health Interventions for COVID-19: Emerging Evidence and Implications for an Evolving Public Health Crisis.
JAMA
(Centers for Disease Control and Prevention COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html)
Centers for Disease Control and Prevention COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html
Centers for Disease Control and Prevention COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html, Centers for Disease Control and Prevention COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html
(GaoS, RaoJ, KangY, LiangY, KruseJ Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824 )
GaoS, RaoJ, KangY, LiangY, KruseJ Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824
GaoS, RaoJ, KangY, LiangY, KruseJ Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824 , GaoS, RaoJ, KangY, LiangY, KruseJ Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824
E. Vynnycky, R. White (2010)
An Introduction to Infectious Disease Modelling
(ZhaoZ, ShawSL, XuY, LuF, ChenJ, YinL Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298 )
ZhaoZ, ShawSL, XuY, LuF, ChenJ, YinL Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298
ZhaoZ, ShawSL, XuY, LuF, ChenJ, YinL Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298 , ZhaoZ, ShawSL, XuY, LuF, ChenJ, YinL Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298
(de MontjoyeYA, GambsS, BlondelV, On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 30532052)
de MontjoyeYA, GambsS, BlondelV, On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 30532052
de MontjoyeYA, GambsS, BlondelV, On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 30532052, de MontjoyeYA, GambsS, BlondelV, On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 30532052
Marta González, César Hidalgo, A. Barabási (2008)
Understanding individual human mobility patterns
Nature, 453
S. Sanche, Y. Lin, Chonggang Xu, E. Romero-Severson, N. Hengartner, R. Ke (2020)
High Contagiousness and Rapid Spread of Severe Acute Respiratory Syndrome Coronavirus 2
Emerging Infectious Diseases, 26
(TianH, LiuY, LiY, An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 32234804)
TianH, LiuY, LiY, An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 32234804
TianH, LiuY, LiY, An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 32234804, TianH, LiuY, LiY, An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 32234804
(YuanY, RaubalM Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555 )
YuanY, RaubalM Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555
YuanY, RaubalM Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555 , YuanY, RaubalM Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555
(LaiS, RuktanonchaiNW, ZhouL, Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020.03.03.20029843)
LaiS, RuktanonchaiNW, ZhouL, Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020.03.03.20029843
LaiS, RuktanonchaiNW, ZhouL, Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020.03.03.20029843, LaiS, RuktanonchaiNW, ZhouL, Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020.03.03.20029843
(TsouMH Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251)
TsouMH Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251
TsouMH Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251, TsouMH Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251
L. Ferretti, C. Wymant, M. Kendall, Lele Zhao, A. Nurtay, Lucie Abeler-Dörner, Michael Parker, D. Bonsall, C. Fraser (2020)
Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing
Science (New York, N.y.), 368
S. Lai, N. Ruktanonchai, Liangcai Zhou, O. Prosper, W. Luo, J. Floyd, A. Wesolowski, Chi Zhang, Xiangjun Du, Hongjie Yu, A. Tatem, W. Ruktanonchai (2020)
Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study
(HuangX, LiZ, JiangY, LiX, PorterD Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020.)
HuangX, LiZ, JiangY, LiX, PorterD Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020.
HuangX, LiZ, JiangY, LiX, PorterD Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020., HuangX, LiZ, JiangY, LiX, PorterD Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020.
Michael Warren, S. Skillman (2020)
Mobility Changes in Response to COVID-19
ArXiv, abs/2003.14228
(HartleyDM, PerencevichEN Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 32275299)
HartleyDM, PerencevichEN Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 32275299
HartleyDM, PerencevichEN Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 32275299, HartleyDM, PerencevichEN Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 32275299
(ZhouC, SuF, PeiT, COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10.1016/j.geosus.2020.03.005)
ZhouC, SuF, PeiT, COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10.1016/j.geosus.2020.03.005
ZhouC, SuF, PeiT, COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10.1016/j.geosus.2020.03.005, ZhouC, SuF, PeiT, COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10.1016/j.geosus.2020.03.005
Y. Montjoye, César Hidalgo, M. Verleysen, V. Blondel (2013)
Unique in the Crowd: The privacy bounds of human mobility
Scientific Reports, 3
(ZhuJ The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/index.html)
ZhuJ The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/index.html
ZhuJ The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/index.html, ZhuJ The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/index.html
John Hipp, Adam Boessen (2017)
The Shape of Mobility: Measuring the Distance Decay Function of Household Mobility
The Professional Geographer, 69
D. Huremović (2019)
Social Distancing, Quarantine, and Isolation
Psychiatry of Pandemics
(ShawSL, TsouMH, YeX Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9):1687-1693. doi:10.1080/13658816.2016.1164317 )
ShawSL, TsouMH, YeX Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9):1687-1693. doi:10.1080/13658816.2016.1164317
ShawSL, TsouMH, YeX Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9):1687-1693. doi:10.1080/13658816.2016.1164317 , ShawSL, TsouMH, YeX Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9):1687-1693. doi:10.1080/13658816.2016.1164317
(SongC, QuZ, BlummN, BarabásiAL Limits of predictability in human mobility. Science. 2010;327(5968):1018-1021. doi:10.1126/science.1177170 20167789)
SongC, QuZ, BlummN, BarabásiAL Limits of predictability in human mobility. Science. 2010;327(5968):1018-1021. doi:10.1126/science.1177170 20167789
SongC, QuZ, BlummN, BarabásiAL Limits of predictability in human mobility. Science. 2010;327(5968):1018-1021. doi:10.1126/science.1177170 20167789, SongC, QuZ, BlummN, BarabásiAL Limits of predictability in human mobility. Science. 2010;327(5968):1018-1021. doi:10.1126/science.1177170 20167789
(LiM, GaoS, LuF, ZhangH Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346 )
LiM, GaoS, LuF, ZhangH Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346
LiM, GaoS, LuF, ZhangH Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346 , LiM, GaoS, LuF, ZhangH Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346
Lei Zhang, Sepehr Ghader, Michael Pack, Chenfeng Xiong, Aref Darzi, Mofeng Yang, Qianqian Sun, Aliakbar Kabiri, Songhua Hu (2020)
AN INTERACTIVE COVID-19 MOBILITY IMPACT AND SOCIAL DISTANCING ANALYSIS PLATFORM
medRxiv
W. Guan, Z. Ni, Yu Hu, W. Liang, C. Ou, Jianxing He, Lei Liu, H. Shan, C. Lei, D. Hui, Bin Du, Lanjuan Li, G. Zeng, K. Yuen, Ru-chong Chen, C. Tang, Taojiao Wang, Ping-yan Chen, J. Xiang, Shi‐yue Li, Jin-lin Wang, Ziyao Liang, Yi-xiang Peng, Li Wei, Yong Liu, Ya-hua Hu, P. Peng, Jian-ming Wang, Ji-yang Liu, Zhong Chen, Gang Li, Zhi-jian Zheng, Shao-qin Qiu, Jie Luo, C. Ye, Shao-yong Zhu, N. Zhong (2020)
Clinical Characteristics of Coronavirus Disease 2019 in China
The New England Journal of Medicine
(LiangY, GaoS, CaiY, FoutzNZ, WuL Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624 )
LiangY, GaoS, CaiY, FoutzNZ, WuL Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624
LiangY, GaoS, CaiY, FoutzNZ, WuL Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624 , LiangY, GaoS, CaiY, FoutzNZ, WuL Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624
(LinJ. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37(1):145-151. doi:10.1109/18.61115 )
LinJ. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37(1):145-151. doi:10.1109/18.61115
LinJ. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37(1):145-151. doi:10.1109/18.61115 , LinJ. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37(1):145-151. doi:10.1109/18.61115
(PullanoG, ValdanoE, ScarpaN, RubrichiS, ColizzaV Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097)
PullanoG, ValdanoE, ScarpaN, RubrichiS, ColizzaV Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097
PullanoG, ValdanoE, ScarpaN, RubrichiS, ColizzaV Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097, PullanoG, ValdanoE, ScarpaN, RubrichiS, ColizzaV Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097
(JiangJ, LiQ, TuW, ShawSL, YueY A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816.2018.1552964 )
JiangJ, LiQ, TuW, ShawSL, YueY A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816.2018.1552964
JiangJ, LiQ, TuW, ShawSL, YueY A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816.2018.1552964 , JiangJ, LiQ, TuW, ShawSL, YueY A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816.2018.1552964
(XuY, BelyiA, BojicI, RattiC Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001 )
XuY, BelyiA, BojicI, RattiC Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001
XuY, BelyiA, BojicI, RattiC Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001 , XuY, BelyiA, BojicI, RattiC Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001
(Centers for Disease Control and Prevention Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm)
Centers for Disease Control and Prevention Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm
Centers for Disease Control and Prevention Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm, Centers for Disease Control and Prevention Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm
(CoriA, FergusonNM, FraserC, CauchemezS A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 24043437)
CoriA, FergusonNM, FraserC, CauchemezS A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 24043437
CoriA, FergusonNM, FraserC, CauchemezS A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 24043437, CoriA, FergusonNM, FraserC, CauchemezS A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 24043437
N. Jewell, J. Lewnard, B. Jewell (2020)
Predictive Mathematical Models of the COVID-19 Pandemic: Underlying Principles and Value of Projections.
JAMA
(US Census Bureau TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019 Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation-u-s-2010-census-urban-area-national)
US Census Bureau TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019 Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation-u-s-2010-census-urban-area-national
US Census Bureau TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019 Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation-u-s-2010-census-urban-area-national, US Census Bureau TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019 Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation-u-s-2010-census-urban-area-national
Qunying Huang, D. Wong (2016)
Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us?
International Journal of Geographical Information Science, 30
(LiR, PeiS, ChenB, Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 32179701)
LiR, PeiS, ChenB, Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 32179701
LiR, PeiS, ChenB, Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 32179701, LiR, PeiS, ChenB, Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 32179701
Song Gao
Spatial Cognition & Computation: an Interdisciplinary Journal Spatio-temporal Analytics for Exploring Human Mobility Patterns and Urban Dynamics in the Mobile Age Spatio-temporal Analytics for Exploring Human Mobility Patterns and Urban Dynamics in the Mobile Age
Chenghu Zhou, F. Su, T. Pei, A. Zhang, Yunyan Du, Bin Luo, Zhidong Cao, Juanle Wang, Wen Yuan, Yunqiang Zhu, Ci Song, Jie Chen, Jun Xu, Fujia Li, Ting Ma, Lili Jiang, Fengqin Yan, Jiawei Yi, Yunfeng Hu, Y. Liao, Han Xiao (2020)
COVID-19: Challenges to GIS with Big Data
Geography and Sustainability, 1
Taeuber Cm (2000)
The American community survey.
Population today, 28
Jincheng Jiang, Qingquan Li, Wei Tu, S. Shaw, Y. Yue (2018)
A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility
International Journal of Geographical Information Science, 33
(VynnyckyE, WhiteR An Introduction to Infectious Disease Modelling. Oxford University Press; 2010.)
VynnyckyE, WhiteR An Introduction to Infectious Disease Modelling. Oxford University Press; 2010.
VynnyckyE, WhiteR An Introduction to Infectious Disease Modelling. Oxford University Press; 2010., VynnyckyE, WhiteR An Introduction to Infectious Disease Modelling. Oxford University Press; 2010.
) Mapping human mobility changes at the U.S. county level, available at
Y. Montjoye, S. Gambs, V. Blondel, G. Canright, Nicolas Cordes, Sébastien Deletaille, Kenth Engø-Monsen, M. García-Herranz, J. Kendall, C. Kerry, G. Krings, E. Letouzé, M. Luengo-Oroz, N. Oliver, Luc Rocher, A. Rutherford, Z. Smoreda, J. Steele, Erik Wetter, A. Pentland, Linus Bengtsson (2018)
On the privacy-conscientious use of mobile phone data
Scientific Data, 5
Yihong Yuan, M. Raubal (2016)
Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study
International Journal of Geographical Information Science, 30
(KangY, ZhangF, PengW, Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919 )
KangY, ZhangF, PengW, Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919
KangY, ZhangF, PengW, Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919 , KangY, ZhangF, PengW, Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919
(PanA, LiuL, WangC, Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 32275295)
PanA, LiuL, WangC, Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 32275295
PanA, LiuL, WangC, Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 32275295, PanA, LiuL, WangC, Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 32275295
Lun Wu, Ye Zhi, Zhengwei Sui, Yu Liu (2014)
Intra-Urban Human Mobility and Activity Transition: Evidence from Social Media Check-In Data
PLoS ONE, 9
Chaoming Song, Zehui Qu, Nicholas Blumm, A. Barabási (2010)
Limits of Predictability in Human Mobility
Science, 327
(WuL, ZhiY, SuiZ, LiuY Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 24824892)
WuL, ZhiY, SuiZ, LiuY Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 24824892
WuL, ZhiY, SuiZ, LiuY Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 24824892, WuL, ZhiY, SuiZ, LiuY Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 24824892
(SenS, Karaca-MandicP, GeorgiouA Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 32459287)
SenS, Karaca-MandicP, GeorgiouA Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 32459287
SenS, Karaca-MandicP, GeorgiouA Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 32459287, SenS, Karaca-MandicP, GeorgiouA Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 32459287
(FerrettiL, WymantC, KendallM, Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 32234805)
FerrettiL, WymantC, KendallM, Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 32234805
FerrettiL, WymantC, KendallM, Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 32234805, FerrettiL, WymantC, KendallM, Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 32234805
(MaierBF, BrockmannD Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 32269067)
MaierBF, BrockmannD Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 32269067
MaierBF, BrockmannD Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 32269067, MaierBF, BrockmannD Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 32269067
(McKenzieG, KeßlerC, AndrisC Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10.5311/JOSIS.2019.19.608)
McKenzieG, KeßlerC, AndrisC Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10.5311/JOSIS.2019.19.608
McKenzieG, KeßlerC, AndrisC Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10.5311/JOSIS.2019.19.608, McKenzieG, KeßlerC, AndrisC Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10.5311/JOSIS.2019.19.608
(PrestbyT, AppJ, KangY, GaoS Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911)
PrestbyT, AppJ, KangY, GaoS Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911
PrestbyT, AppJ, KangY, GaoS Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911, PrestbyT, AppJ, KangY, GaoS Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911
Ming-Hsiang Tsou (2015)
Research challenges and opportunities in mapping social media and Big Data
Cartography and Geographic Information Science, 42
(GaoS. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300 )
GaoS. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300
GaoS. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300 , GaoS. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300
Song Gao, Jinmeng Rao, Yuhao Kang, Yunlei Liang, Jake Kruse (2020)
Mapping county-level mobility pattern changes in the United States in response to COVID-19
SIGSPATIAL Special, 12
(de MontjoyeYA, HidalgoCA, VerleysenM, BlondelVD Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 23524645)
de MontjoyeYA, HidalgoCA, VerleysenM, BlondelVD Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 23524645
de MontjoyeYA, HidalgoCA, VerleysenM, BlondelVD Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 23524645, de MontjoyeYA, HidalgoCA, VerleysenM, BlondelVD Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 23524645
(ThompsonRN, StockwinJE, van GaalenRD, Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 31624039)
ThompsonRN, StockwinJE, van GaalenRD, Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 31624039
ThompsonRN, StockwinJE, van GaalenRD, Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 31624039, ThompsonRN, StockwinJE, van GaalenRD, Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 31624039
Chaogui Kang, Xiujun Ma, D. Tong, Yu Liu (2012)
Intra-urban human mobility patterns: An urban morphology perspective
Physica A-statistical Mechanics and Its Applications, 391
Robin Thompson, Robin Thompson, J. Stockwin, R. Gaalen, J. Polonsky, J. Polonsky, Zhian Kamvar, P. Demarsh, P. deMarsh, Elisabeth Dahlqwist, S. Li, E. Miguel, T. Jombart, T. Jombart, J. Lessler, S. Cauchemez, A. Cori (2019)
Improved inference of time-varying reproduction numbers during infectious disease outbreaks
Epidemics, 29
(GitHub Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper)
GitHub Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper
GitHub Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper, GitHub Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper
(2020)
Bureau
Xiao Huang, Zhenlong Li, Yuqin Jiang, Xiaoming Li, D. Porter (2020)
Twitter, human mobility, and COVID-19
ArXiv, abs/2007.01100
(WarrenMS, SkillmanSW Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 2020.)
WarrenMS, SkillmanSW Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 2020.
WarrenMS, SkillmanSW Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 2020., WarrenMS, SkillmanSW Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 2020.

Publisher: American Medical Association
Copyright: Copyright 2020 Gao S et al. JAMA Network Open.
eISSN: 2574-3805
DOI: 10.1001/jamanetworkopen.2020.20485
Publisher site: See Article on Publisher Site

Abstract

Key Points Question Did human mobility patterns IMPORTANCE A stay-at-home social distancing mandate is a key nonpharmacological measure to change during stay-at-home orders and reduce the transmission rate of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), but were the mobility changes associated a high rate of adherence is needed. with the coronavirus disease 2019 (COVID-19) curve? OBJECTIVE To examine the association between the rate of human mobility changes and the rate of Findings This cross-sectional study confirmed cases of SARS-CoV-2 infection. using anonymous location data from more than 45 million mobile phones DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study used daily travel distance and found that median travel distance home dwell time derived from millions of anonymous mobile phone location data from March 11 to decreased and stay-at-home time April 10, 2020, provided by the Descartes Labs and SafeGraph to quantify the degree to which social increased across the nation, although distancing mandates were followed in the 50 US states and District of Columbia and the association there was geographic variation. State- of mobility changes with rates of coronavirus disease 2019 (COVID-19) cases. specific empirical doubling time of total COVID-19 cases increased (ie, the spread EXPOSURE State-level stay-at-home orders during the COVID-19 pandemic. reduced) significantly after stay-at- home orders were put in place. MAIN OUTCOMES AND MEASURES The main outcome was the association of state-specific rates of COVID-19 confirmed cases with the change rates of median travel distance and median home Meaning These findings suggest that dwell time of anonymous mobile phone users. The increase rates are measured by the exponent in stay-at-home social distancing curve fitting of the COVID-19 cumulative confirmed cases, while the mobility change (increase or mandates were associated with the decrease) rates were measured by the slope coefficient in curve fitting of median travel distance and reduced spread of COVID-19 when they median home dwell time for each state. were followed. RESULTS Data from more than 45 million anonymous mobile phone devices were analyzed. The Supplemental content correlation between the COVID-19 increase rate and travel distance decrease rate was –0.586 (95% CI, –0.742 to –0.370) and the correlation between COVID-19 increase rate and home dwell time Author affiliations and article information are listed at the end of this article. increase rate was 0.526 (95% CI, 0.293 to 0.700). Increases in state-specific doubling time of total cases ranged from 1.0 to 6.9 days (median [interquartile range], 2.7 [2.3-3.3] days) before stay-at- home orders were enacted to 3.7 to 30.3 days (median [interquartile range], 6.0 [4.8-7.1] days) after stay-at-home social distancing orders were put in place, consistent with pandemic modeling results. CONCLUSIONS AND RELEVANCE These findings suggest that stay-at-home social distancing mandates, when they were followed by measurable mobility changes, were associated with reduction in COVID-19 spread. These results come at a particularly critical period when US states are beginning to relax social distancing policies and reopen their economies. These findings support the efficacy of social distancing and could help inform future implementation of social distancing policies should they need to be reinstated during later periods of COVID-19 reemergence. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 Open Access. This is an open access article distributed under the terms of the CC-BY License. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 1/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Introduction The coronavirus disease 2019 (COVID-19) pandemic is a global threat with escalating health, economic, and social challenges. As of April 11, 2020, there were 492 416 total confirmed cases and 18 559 total deaths in the US, according to reports from the Centers for Disease Control and Prevention (CDC). People are still witnessing widespread community transmission of COVID-19 all over the world. To date, there is neither a vaccine nor pharmacological agent found to reduce the transmission of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the virus that causes COVID-19. Thus, the effects of nonpharmacological pandemic control and intervention measures, including travel restrictions, closures of schools and nonessential business services, wearing of face masks, testing, isolation, and timely quarantine on delaying the spread of COVID-19, have been 2-6 largely investigated and reported. To mitigate and ultimately contain the COVID-19 pandemic, one of the important nonpharmacological control measures to reduce the transmission rate of SARS- CoV-2 in the population is social (ie, physical) distancing. An interactive web-based mapping platform that provides timely quantitative information on how people in different counties and states reacted to state-at-home social distancing mandates has been developed (eAppendix 2 in the Supplement). It integrates geographic information systems and daily updated human mobility statistical patterns derived from millions of anonymized and aggregated smartphone location data at the county level in 7-10 the US. Reduced mobility and trips may help limit people’s exposure to large in-person gatherings. However, it is worth noting that reduced mobility does not necessarily ensure that social distancing in practice follows the CDC’s definition: “stay at least 6 feet (about 2 arms’ length) from other 11 12 people.” Due to the mobile phone Global Positioning System horizontal error and uncertainty, such physical distancing patterns cannot be directly identified from the user aggregated mobility data; that would require other wearable sensors or mobile phone Bluetooth trackers, which raise issues of personal data privacy and ethical concerns. Because COVID-19 is more contagious and far more deadly than seasonal flu, social distancing is critical in the fight to save lives and prevent illness. However, to what degree such guidelines have been followed from place to place before and after shelter-in-place orders across the US and the quantitative effect on flattening the curve are as yet unknown, to our knowledge. To this end, we used 2 human mobility metrics, the median of individual maximum travel distance and stay-at-home time derived from location data from millions of mobile phones, to assess the association of stay-at-home policies with reducing the spread of COVID-19. For each state, we examined these measures against the rate of SARS-CoV-2 infection cases. Methods A waiver of institutional review board review and informed consent was obtained from the University of Wisconsin–Madison because anonymized and aggregated data were used and our study does not involve human participants as defined. This study follows the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) reporting guideline. Data In this cross-sectional study, the epidemiological confirmed cases data were retrieved from the Corona Data Scraper open source project, which provides local-level and community-driven reports, and we conflated the data with the state-level department of health services official reports in each state to ensure the data quality. To understand how people reacted to the stay-at-home social distancing guidelines imposed during the COVID-19 pandemic, human mobility changes were considered in terms of changes in travel distance and stay-at-home dwell time. The travel distance mobility data were collected from an open-source repository released by Descartes Labs, while the home dwell time data derived from more than 45 million anonymous mobile phone users were JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 2/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US processed from SafeGraph. Both data sources were acquired at the county level and aggregated to the state level using median and interquartile range (IQR) values. To consider the socioeconomic factors that may be associated with statewide changes in human mobility, socioeconomic variables at the state level were also collected from the American Community Survey and the US Census Bureau. The following socioeconomic variables and geospatial data sets were retrieved and computed: population (ie, number of people), population density (measured by population divided by area of state), proportion of population with bachelor’s degree, proportions of population of different races/ethnicities, proportion of population of different age groups, median household income, and urban core area boundaries. Statistical Analysis Simple linear regression and multivariate linear regression analyses were performed using the scikit- learn package version 0.23.1 in Python. The Pearson correlation coefficient with 2-sided significance test, P < .05, was computed using the SciPy package version 1.4.0 in Python. Curve Fitting for Pandemic Spread, Travel Distance, and Home Dwell Time In the mathematical modeling process, we used a few types of mathematical formulas (eAppendix 1 in the Supplement) to fit the curve of the cumulative confirmed cases for the COVID-19 with respect to their temporal changes in each state and selected the following scaling law with a deivation term formula as the most appropriate: y (t)= t + k, in which y is the total number of confirmed cases in c c each state as a function of time, t is the number of days from March 11, 2020 (when the COVID-19 became a pandemic), and b and k are parameters we will estimate. By fitting the curve, we can compare the infection rates among different states using the coefficient b estimated from the model. Meanwhile, we used linear regression to detect the travel distance decreasing rate (represented by the slope estimated from the linear model) over time (eFigure 2 in the Supplement) and examined whether there was a correlation between the increase rate of cases and the distance decreasing rate. We also fitted the curve for the home dwell time changes for each state using the linear regression model. The linear model was selected from a few different models because it is the simplest one, and results of all fitted models were similar. We then calculated the correlation between the home dwell time increasing rate (the slope estimated from the linear model) and the increase rate of the number of confirmed cases. Evaluating Factors Associated With Changes in Travel Distance and Home Dwell Time To understand what socioeconomic factors were associated with travel distance changes and home dwell time changes, a multilinear regression model integrating socioeconomic factors was used to fit the mobility change rates that were represented by the slope estimates for each state. The R as goodness of fit and significance of variables are reported (eAppendix 1 in the Supplement). Calculating the Doubling Time of Total Confirmed Cases We investigated how the social distancing guidelines and stay-at-home orders (eTable 5 in the Supplement) were associated with the pandemic doubling time of COVID-19 confirmed cases from March 11 to April 10, 2020, in each state. We used mathematical curve fitting models and mechanistic epidemic models (eAppendix 1 in the Supplement) using Bayesian parametric estimation of the serial 18,19 interval distribution of successive cases to cross validate the conclusion. We calculated the doubling time of the number of cumulative confirmed cases (ie, the time intervals it takes for the cumulative confirmed cases to double in size ) to reflect the characteristics of the COVID-19 pandemic spread, especially how the stay-at-home orders in each state were associated with flattening the COVID-19 curve. The larger the doubling time, the smoother the pandemic increase curve. Within the time frame of our study, the state-level increase rates of COVID-19 cases in the US were either exponential or subexponential, thus we implemented an exponential model and a power-law model to fit the curve for calculating the doubling time. We also calculated the doubling JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 3/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US time based on empirical observations (model-free) to further explore how the doubling time differs in these methods. We used the effective date of the stay-at-home order to split the confirmed case data into 2 parts: before the order and after the order. We fitted each model on the data before the order and after the order, then we calculated the doubling times of the confirmed cases based on the model and empirical COVID-19 infection data. The doubling time of the cumulative confirmed cases in each state is defined as ln(2) d(t) = ln(1 + r[t]) In which d(t) represents the doubling time of the cumulative confirmed cases on date t in each state, ln(x) is the natural log of x, and r(t) represents the increase rate of the cumulative confirmed cases on date t in each state. In addition, we visualized and investigated the overall probability density distribution of the median doubling time before and after the stay-at-home order in each state to have a better understanding of the overall changes in the pandemic spread nationwide. Furthermore, we measured the similarity in probability density distribution of the median doubling time between the fitting results and the empirical data using the Jensen–Shannon Divergence. Results Trends of Human Mobility Changes Data from more than 45 million anonymous mobile phone devices were analyzed. The associations of stay-at-home policies with human mobility changes are illustrated in Figure 1, Figure 2, and the Table. Figure 1A shows the temporal changes of the median of individual maximum travel distances in the states with the highest infection rates (ie, New York, New Jersey, Michigan, California, and Massachusetts) by April 10, 2020. People’s daily mobility decreased significantly but with different temporal lags following the implementation of statewide stay-at-home orders across these states (Table). Figure 1B shows the state-specific temporal changes of median home dwell time. With the social distancing guidelines and shelter-at-home orders in place, the median home dwell time increased significantly in most states since March 23, 2020 (Table). Figure 2 shows the spatial distributions of confirmed cases per capita and the median of travel distances and median of home dwell time in 2 specific days as snapshots for comparison of mobility patterns with the COVID-19 infection rate before and after stay-at-home-orders: March 11 and April 10, 2020. The median travel Figure 1. Temporal Changes in Median of Individual Maximum Travel Distance and Median Home Dwell Time in the Most Infected US States From March 11 to April 10, 2020 A Travel distance B Stay-at-home dwell time New York 10 1100 New Jersey Michigan Massachusetts 1000 California Florida ALL 0 400 March March March March April April March March March March April April 11, 2020 17, 2020 23, 2020 29, 2020 4, 2020 10, 2020 11, 2020 17, 2020 23, 2020 29, 2020 4, 2020 10, 2020 Date Date JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 4/13 Median travel distance, km Home dwell time, min JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US distance decreased and the median home dwell time increased across the US during this period. In addition, we calculated the means of median travel distances before and after the stay-at-home- orders in each state. The median travel distances decreased in all states (Table). Implementation of stay-at-home social distancing policies were associated with human movement changes, that is, people generally reduced their daily travel distances and increased their home dwell time. Interestingly, the multiple linear regression model result for the increasing rate of home dwell time with the socioeconomic variables shows that the ratio of Asian individuals in each state was positively associated with longer home dwell time at the state level. The higher the proportion of Asian Figure 2. Comparison Among Confirmed Coronavirus Disease 2019 Cases Per Capita, Median of Individual Maximum Travel Distance, and Median Home Dwell Time From March 11 and April 10, 2020 A Confirmed cases Cases per capita >0.00 to 0.02 >0.02 to 0.08 >0.08 to 0.31 >0.31 to 15 775.90 March 11th April 10th B Median maximum travel distance Distance, km 0 to 3.38 >3.38 to 6.87 >6.87 to 9.89 >9.89 to 14.51 >14.51 to 684.76 March 11th April 10th C Stay-at-home dwell time Time, min 0.00 - 664.89 >664.89 to 747.50 >747.50 to 799.49 >799.49 to 856.61 >856.61 to 1185.02 March 11th April 10th JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 5/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Table. Empirical Doubling Time of Total Infected Cases and the Median Travel Distance and Home Dwell Time Before and After Stay-at-Home Orders Doubling time, d Travel distance, km Home dwell time, min Median (IQR) Median (IQR) Median (IQR) State Before order After order Change Before order After order Change Before order After order Change Alabama 3.3 6.5 3.2 6.651 4.311 –2.328 660.9 781.4 120.5 (2.3-4.4) (5.4-7.8) (5.356-8.278) (4.283-4.903) (576.3-695.4) (758.5-799.2) Alaska 6.9 30.3 23.4 1.369 0.091 –1.273 342.3 427.1 84.8 (3.5-9.2) (23.6-38.9) (0.168-2.450) (0.049-0.092) (282.5-376.2) (343.6-454.4) Arizona 2.5 6.8 4.3 3.227 1.037 –2.231 523.9 637.9 114.0 (2.0-3.8) (5.8-10.5) (1.82-5.071) (0.878-1.274) (489.3-560.7) (520.9-645.1) California 3.3 5.3 2.0 3.922 0.770 –3.207 748.1 833.0 84.9 (3.1-3.7) (3.8-7.3) (3.128-6.177) (0.259-0.986) (642.7-760.4) (754.4-874.0) Colorado 2.6 6.2 3.6 2.824 0.319 –2.496 529.6 676.5 146.9 (2.3-3.2) (5.7-9.3) (1.387-4.334) (0.095-0.464) (475.1-548.0) (575.7-692.6) Connecticut 1.7 4.5 2.8 3.100 0.396 –2.687 667.5 822.5 155.0 (1.3-2.8) (2.8-7.5) (2.174-4.482) (0.107-0.549) (583.0-725.4) (752.4-854.1) Delaware 2.9 4.7 1.8 4.102 0.641 –3.462 629.3 749.5 120.2 (1.5-4.7) (3.7-5.4) (2.888-5.704) (0.183-0.937) (546.9-667.6) (676.8-795.6) Florida 3.0 10.0 7.0 3.484 0.930 –2.622 559.6 694.3 134.7 (2.1-3.9) (8.8-11.1) (1.805-5.224) (0.655-1.208) (476.2-604.4) (680.0-717.5) Georgia 3.5 6.4 2.9 4.852 2.278 –2.758 636.6 759.4 122.9 (2.3-5.0) (6.1-10.7) (3.292-6.57) (1.818-3.04) (546.2-674.1) (732.3-784.3) Hawaii 2.0 7.3 5.3 4.294 1.147 –3.177 625.7 789.4 163.7 (1.6-2.4) (5.2-11.1) (3.131-6.057) (1.054-1.466) (541.1-649.5) (607.3-830.9) Idaho 1.3 4.8 3.5 3.424 1.286 –2.208 567.9 686.2 118.3 (1.0-2.6) (2.9-9.8) (2.661-4.599) (1.063-1.713) (499.7-604.7) (621.3-718.7) Illinois 1.9 4.7 2.8 4.214 0.784 –3.428 648.6 764.0 115.4 (1.9-2.4) (4.0-7.1) (3.046-6.604) (0.427-1.144) (599.5-694.8) (725.9-802.8) Indiana 2.7 3.7 1.0 4.513 1.64 –2.932 605.2 718.8 113.6 (2.0-3.0) (3.0-4.2) (3.41-6.232) (1.432-2.193) (525.7-634.2) (653.5-757.9) Kansas 2.7 5.8 3.1 3.589 1.897 –1.67 606.2 702.1 96.0 (1.8-3.5) (5.1-10.2) (2.300-4.657) (1.735-2.269) (553.6-644.3) (607.0-730.9) Kentucky 2.5 5.4 2.9 4.778 2.802 –2.041 630.5 744.3 113.8 (1.7-4.1) (4.3-9.1) (3.745-6.087) (2.333-3.256) (562.8-670.4) (686.1-764.2) Louisiana 2.1 4.6 2.5 6.242 3.176 –3.122 609.5 736.9 127.4 (1.9-2.3) (3.1-8.7) (5.925-8.289) (2.877-3.830) (515.4-631.4) (675.7-763.0) Maine 3.7 16.5 12.8 2.413 0.361 –2.014 553.2 690.0 136.7 (2.4-7.1) (11.6-17.6) (0.735-3.331) (0.094-0.705) (481.6-590.2) (638.8-703.4) Maryland 2.8 4.2 1.4 2.346 0.122 –2.271 688.5 794.9 106.4 (2.2-3.6) (3.4-6.1) (0.307-3.654) (0.045-0.092) (611.4-745.6) (676.6-824.0) Massachusetts 3.8 4.7 0.9 2.323 0.108 –2.213 640.4 780.8 140.5 (3.0-5.3) (4.5-6.3) (0.991-3.412) (0.045-0.104) (538.5-670.0) (692.0-812.3) Michigan 2.3 4.4 2.1 3.562 0.104 –3.454 566.6 734.3 167.6 (1.4-2.8) (3.7-7.1) (2.274-5.058) (0.046-0.131) (492.3-600.5) (649.8-764.3) Minnesota 3.0 8.7 5.7 2.927 0.482 –2.54 556.9 701.4 144.5 (1.7-4.9) (7.6-9.4) (1.359-4.268) (0.138-0.509) (500.4-607.0) (607.1-732.2) Mississippi 2.8 9.4 6.6 7.103 4.751 –2.613 612.1 744.6 132.5 (1.7-5.1) (6.4-13.6) (5.675-8.868) (4.11-5.919) (514.0-654.5) (720.4-767.7) Montana 2.4 8.3 5.9 2.353 0.820 –1.475 443.3 559.6 116.3 (1.8-3.2) (7.4-14.5) (1.821-2.953) (0.405-1.158) (400.8-506.0) (477.2-577.8) Nevada 3.7 11.2 7.5 2.432 0.502 –1.962 516.0 611.5 95.6 (1.7-5.0) (8.5-12.6) (0.687-4.353) (0.253-0.764) (479.3-553.0) (596.7-620.5) New Hampshire 3.0 5.8 2.8 3.689 0.818 –3.014 585.0 735.4 150.4 (2.3-4.3) (4.3-11.7) (1.527-5.603) (0.266-1.073) (528.3-631.6) (623.7-752.7) New Jersey 1.8 4.2 2.4 3.244 0.095 –3.162 722.1 968.4 246.3 (1.3-2.0) (3.1-6.6) (1.972-5.362) (0.043-0.085) (671.7-819.5) (900.8-983.9) New Mexico 3.1 5.2 2.1 3.492 0.993 –2.519 467.8 577.5 109.8 (2.6-3.5) (4.4-6.9) (2.728-4.579) (0.873-1.275) (407.9-489.1) (488.1-596.8) New York 1.8 6.4 4.6 2.093 0.037 –2.056 580.0 767.4 187.4 (1.5-2.2) (4.4-9.5) (1.137-3.554) (0.032-0.039) (527.3-644.9) (669.5-785.6) North Carolina 2.7 6.3 3.6 5.220 2.679 –2.577 606.2 690.1 84.0 (2.1-3.5) (5.1-11.0) (3.935-7.065) (2.204-3.199) (545.8-633.2) (595.4-711.2) Ohio 2.1 5.3 3.2 4.076 1.202 –2.934 611.0 729.7 118.7 (1.9-2.5) (3.8-8.0) (3.275-6.096) (0.806-1.603) (547.3-653.0) (688.0-762.5) Oklahoma 2.4 5.6 3.2 5.962 3.550 –2.511 631.3 767.3 136.1 (1.6-3.1) (4.3-6.8) (4.864-7.734) (2.881-4.277) (563.2-664.9) (707.3-804.4) Oregon 3.8 6.7 2.9 2.667 0.571 –2.124 629.3 742.3 113.0 (3.2-4.3) (5.0-10.8) (1.930-3.900) (0.232-0.854) (575.8-663.4) (687.0-789.9) (continued) JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 6/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Table. Empirical Doubling Time of Total Infected Cases and the Median Travel Distance and Home Dwell Time Before and After Stay-at-Home Orders (continued) Doubling time, d Travel distance, km Home dwell time, min Median (IQR) Median (IQR) Median (IQR) State Before order After order Change Before order After order Change Before order After order Change Pennsylvania 2.5 5.8 3.3 1.798 0.184 –1.609 656.0 776.7 120.6 (2.1-3.3) (4.5-6.0) (0.078-2.445) (0.089-0.246) (562.7-704.5) (770.2-798.4) Rhode Island 1.9 4.6 2.7 2.286 0.256 –2.034 705.1 795.2 90.1 (1.4-3.5) (4.1-5.3) (0.804-3.592) (0.071-0.357) (587.1-733.2) (747.6-823.0) South Carolina 2.4 5.8 3.4 6.484 3.898 –2.664 586.9 700.2 113.2 (1.8-4.2) (3.9-8.1) (4.942-8.405) (3.651-4.390) (532.3-619.1) (626.4-712.3) Tennessee 3.3 10.3 7.0 5.679 3.442 –2.189 647.8 731.4 83.6 (1.8-4.0) (8.4-12.5) (4.094-7.368) (3.215-4.098) (590.0-683.3) (617.3-760.8) Texas 3.4 6.0 2.6 4.076 1.869 –2.249 589.5 728.8 139.2 (2.5-5.5) (5.5-7.1) (2.413-5.763) (1.837-2.326) (525.1-645.8) (722.2-759.4) Utah 2.5 6.7 4.2 3.351 1.369 –2.05 642.6 710.9 68.3 (1.9-4.1) (5.2-11.6) (2.094-4.933) (0.916-1.791) (560.2-666.3) (667.9-721.0) Vermont 2.3 7.1 4.8 2.716 0.166 –2.592 465.8 648.4 182.7 (1.6-2.8) (5.1-11.9) (0.843-4.386) (0.059-0.201) (414.6-515.0) (535.6-668.2) Virginia 3.4 4.8 1.4 3.261 1.029 –2.273 607.6 695.1 87.5 (2.4-4.9) (4.1-7.2) (1.454-4.669) (0.627-1.320) (556.1-645.8) (596.8-716.4) Washington 4.5 12.3 7.8 2.710 0.253 –2.501 683.5 811.6 128.1 (4.1-6.2) (5.2-14.2) (2.027-4.187) (0.054-0.332) (618.6-717.0) (760.0-848.3) Washington, DC 3.5 6.9 3.4 0.85 0.026 –0.823 615.7 716.7 101.0 (1.9-5.6) (4.3-7.2) (0.031-1.112) (0.024-0.027) (523.4-639.7) (696.9-722.0) West Virginia 1.0 4.4 3.4 4.611 1.691 –2.939 586.3 693.1 106.8 (1.0-1.3) (3.9-7.3) (3.573-6.217) (1.345-2.095) (488.6-626.1) (619.5-721.2) Wisconsin 2.3 7.0 4.7 3.233 0.753 –2.477 594.2 720.2 126.0 (1.9-2.6) (6.1-9.6) (2.061-4.871) (0.574-1.23) (549.6-631.4) (660.3-763.8) Wyoming 3.1 7.9 4.8 2.719 1.798 –0.867 478.3 617.7 139.4 (2.1-5.1) (5.5-11.0) (2.381-3.433) (1.218-2.198) (430.9-539.2) (491.8-636.6) Abbreviations: DC, District of Columbia; IQR, interquartile range. population, the longer the median home dwell time of residents in that state (eTable 7 in the Supplement). Association of Rate of Infection With Mobility Changes We fitted the curves for the state-specific COVID-19 confirmed cases using the scaling-law with a deviation term formula and identified the top 5 states with the largest increase rates of confirmed COVID-19 cases by April 10, 2020: New York, New Jersey, California, Michigan, and Massachusetts. Our fitting results corresponded to the up-to-date COVID-19 situation at that time (eTable 1 and eTable 2 in the Supplement). eFigure 1 in the Supplement shows the reported cases and the fitting curves in these 5 states using the scaling-law with a deviation term formula. The Pearson correlation coefficient between the cases increase rate and the distance decay rate was –0.586 (95% CI, –0.742 to –0.370; P < .001) (eTable 3 in the Supplement). Figure 3A shows the state-level correlation between the increase coefficients of confirmed cases and the travel distance decay coefficients across the nation. The moderate negative correlation indicates that in the states where the confirmed cases were increasing faster, people generally reduced their daily travel distance more quickly. Figure 3B shows the state-level correlation between the increase coefficients of confirmed cases and the home dwell time increment coefficients across the nation. The increase rates and the home dwell time rates (eTable 4 in the Supplement) had a positive correlation of 0.526 (95% CI, 0.293 to 0.700; P < .001), which suggests that in states with higher case increase rates, home dwell time of residents in this state were generally longer. These association analyses found that there was statistically significant mobility reduction associated with the increase rate of COVID-19 cases and that people in most states reduced their daily travel distance and increased stay-at-home time. In addition, the statistical variation of the mobility measures can be largely explained (travel 2 2 distance: R = 0.59; P < .001; home dwell time: R = 0.69; P < 001) by socioeconomic factors, including state policies, race/ethnicity, population density, age groups, and median household JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 7/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US income (eAppendix 1, eTable 6, and eTable 7 in the Supplement). Recent studies have also identified partisan differences in individual responses to stay-at-home social distancing guidelines during the COVID-19 pandemic (H. Alcott et al, unpublished data, July 2020). Pandemic Doubling Time Changes The fitted curves by an exponential model and a power-law model are shown in eFigure 3 and eFigure 4 in the Supplement. For the exponential model before the statewide stay-at-home orders, initial estimates of the increase rates of the number of confirmed cases for the pandemic in each state were 0.17 to 0.70 per day with a doubling time of 1.3 to 4.3 days (median [IQR], 2.6 [2.1-2.9] days). A similar result was found by fitting the power-law model, in which initial estimates of the case rates before the orders in each state were 0.12 to 0.71 cases per day with a doubling time of 1.3 to 6.2 days (median [IQR], 2.7 [2.2-3.1] days). The finding aligned well with the doubling time of 2.3 to 3.3 days in the early pandemic epicenter in Wuhan, China. After the implementation of stay-at-home orders, the estimates of the case rate in each state by the exponential model were reduced to 0.03 to 0.21 cases per day, with a doubling time increased to 3.7 to 27.7 days (median [IQR], 5.7 [4.7-6.9] days). Similarly, the estimates of the case rate in each state by the power-law model were reduced to 0.02 to 0.17 cases per day, with a doubling time increased to 4.3 to 29.8 days (median [IQR], 6.3 [5.4-7.9] days). The finding also aligned well (measured by Jensen–Shannon Divergence) with the result from the observed epidemiological data (Table), in which the empirical case rate in each state was 0.11 to 0.95 cases per day with a doubling time of 1.0 to 6.9 days (median [IQR], 2.7 [2.3-3.3] days) before the statewide stay-at-home orders, and reduced to 0.02 to 0.21 per cases day with a doubling time increased to 3.7 to 30.3 days (median [IQR], 6.0 [4.8-7.1] days) after the orders. The curve fitting results also matched the outcomes of mechanistic epidemic models (eFigure 7 in the 18 19 Supplement), such as the models reported by Cori et al and Thompson et al. These models used confirmed cases and the serial interval, that is, the days between 2 successive infection cases. In addition, we investigated the overall probability density distribution of the doubling time nationwide before and after the stay-at-home orders using the state-level median doubling time (Figure 4A; eFigure 5 and eFigure 6 in the Supplement). The doubling time nationwide increased after the stay-at-home orders (empirical observations: from median [IQR] 2.7 [2.3-3.3] days to median 6.0 [4.8-7.1] days). Our combined results on doubling times suggest that stay-at-home orders were associated with reduction of the COVID-19 pandemic spread and with flattening the curve. Similar findings have also been reported in a study by Sen et al on the association of stay-at-home orders with COVID-19 hospitalizations. In addition, the ten-hundred plot (Figure 4B) also shows that the case increase rate in each of the top 5 states (ie, New York, New Jersey, Michigan, California, Figure 3. State-Level Correlation Between the Increase Coefficients of Confirmed Cases, Travel Distance Decay Coefficients, and Home Dwell Time Increase Coefficients A B Correlation between the increase in coefficients of confirmed Correlation between the increase in coefficients of confirmed cases cases and the travel distance decay coefficients and the home dwell time increment coefficients –0.15 1.90 1.85 –0.20 1.80 –0.25 1.75 –0.30 1.70 –0.35 1.65 –0.40 1.60 –0.45 1.55 –0.50 1.50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Confirmed cases, coefficient Confirmed cases, coefficient JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 8/13 Travel distance, coefficient Home dwell time, coefficient JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US and Massachusetts) slowed down after the stay-at-home orders (approaching subexponential growth). The statistical variation of the mobility measures can be largely explained (travel distance: 2 2 R = 0.59; P < .05; home dwell time: R = 0.69; P < .05) (eAppendix 1 in the Supplement)by socioeconomic factors, including state policies, race/ethnicity, population density, age groups, and median household income (eTable 6 and eTable 7 in the Supplement). Recent studies have also identified partisan differences in individual responses to stay-at-home social distancing guidelines during the COVID-19 pandemic (H. Alcott et al, unpublished data, July 2020). Discussion These findings suggest that stay-at-home social distancing mandates, when they were followed by measurable mobility changes, were associated with reduction in COVID-19 case rates. Great efforts have been made in scientific research communities on the study of human mobility patterns using 26-31 various emerging data sources, including anonymized mobile phone call detail records, social 32,33 34-38 media (eg, Twitter), location-based services, and mobile applications. During the COVID-19 pandemic, both individual-level and aggregated-level human mobility patterns have been found 6,13,39,40 useful in pandemic modeling and digital contact tracing. However, technical challenges (eg, 41-44 location uncertainty), socioeconomic and sampling bias, privacy and ethical concerns have been 45-48 expressed by national and international societies. Moving forward, research efforts should continue exploring the balance of using such human mobility data at different geographic scales for public health and social good while preserving individual privacy and rights. Limitations This study has some limitations. Potential confounding issues relate to other control measures, such as varying state-level quarantine protocols, availability of personal protective equipment, and timely testing, but the detailed information was not available, and the consistency of our results across most states makes such confounding less likely. In addition, the variability in the curve fitting estimated parameters was not accounted for the correlation analysis. There are variations in human behaviors and risk perception even within a state. All these factors contribute to the potential endogeneity of findings and the limitations. Figure 4. Probability Density Distributions and Ten-Hundred Plot of Coronavirus Disease 2019 Spread Before and After Stay-at-Home Orders A B State-level median doubling time State-level confirmed case rate 0.6 30 Before order Before order After order After order 0.5 25 Trajectory 0.4 20 0.3 15 California California 0.2 10 Massachusetts New York New Jersey Michigan 0.1 5 New Jersey New York Michigan Massachusetts 0 0 0 2 4 6 8 10 12 14 0 5 10 15 20 25 30 Doubling time, d Latest date with N cases and the date with cases, d B. The lower-right region represents subexponential growth; the diagonal line, exponential growth; and the upper left region, super-exponential growth. The top 5 states with the most confirmed cases are labeled and their change rate changes are visualized as trajectories. N indicates the number of coronavirus disease 2019 confirmed cases on that date. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 9/13 Density N N Date with cases and the date with cases, d 10 100 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US Conclusions This cross-sectional study found a statistically significant association of 2 human mobility measures (ie, travel distance and stay-at-home time) with the rates of COVID-19 cases across US states. This study found a reduction of the spread of COVID-19 after stay-at-home social distancing mandates were enacted in most states. The findings come at a particularly critical period, when US states are beginning to reopen their economies but COVID-19 cases are surging. At such a time, our study suggests the efficacy of stay-at-home social distancing measures and could inform future public health policy making. ARTICLE INFORMATION Accepted for Publication: July 31, 2020. Published: September 8, 2020. doi:10.1001/jamanetworkopen.2020.20485 Open Access: This is an open access article distributed under the terms of the CC-BY License.©2020GaoSetal. JAMA Network Open. Corresponding Authors: Song Gao, PhD (song.gao@wisc.edu) and Jonathan A. Patz, MD (patz@wisc.edu), University of Wisconsin–Madison, 550 N Park St, Madison, WI 53706. Author Affiliations: GeoDS Lab, Department of Geography, University of Wisconsin–Madison, Madison (Gao, Rao, Kang, Liang, Kruse); School of Veterinary Medicine, University of Wisconsin–Madison, Madison (Dopfer, Mandujano Reyes); School of Medicine and Public Health, University of Wisconsin–Madison, Madison (Sethi, Patz); Statistics and American Family Insurance Data Science Institute, University of Wisconsin–Madison, Madison (Yandell). Author Contributions: Dr Gao had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Gao, Kang, Sethi, Yandell, Patz. Acquisition, analysis, or interpretation of data: Gao, Rao, Kang, Liang, Kruse, Dopfer, Sethi, Mandujano Reyes. Drafting of the manuscript: Gao, Rao, Kang, Liang, Kruse, Dopfer, Sethi, Mandujano Reyes, Patz. Critical revision of the manuscript for important intellectual content: Gao, Rao, Kang, Sethi, Yandell, Patz. Statistical analysis: Gao, Rao, Kang, Liang, Dopfer, Mandujano Reyes, Yandell. Obtained funding: Gao. Administrative, technical, or material support: Gao, Liang, Yandell. Supervision: Gao. Conflict of Interest Disclosures: Dr Gao reported receiving grants from National Science Foundation during the conduct of the study. No other disclosures were reported. Funding/Support: Drs Gao and Patz received funding from grant No. BCS-2027375 from the National Science Foundation. Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Disclaimer: The opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. REFERENCES 1. Centers for Disease Control and Prevention. COVID-19 cases in the U.S. Accessed April 12, 2020. https://www.cdc. gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html 2. Pan A, Liu L, Wang C, et al. Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China. JAMA. 2020;323(19):1915-1923. doi:10.1001/jama.2020.6130 3. Hartley DM, Perencevich EN. Public health interventions for COVID-19: emerging evidence and implications for an evolving public health crisis. JAMA. 2020;323(19):1908-1909. doi:10.1001/jama.2020.5910 4. Lai S, Ruktanonchai NW, Zhou L, et al. Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study. medRxiv. Preprint posted online March 13, 2020. doi:10.1101/2020. 03.03.20029843 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 10/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US 5. Chinazzi M, Davis JT, Ajelli M, et al. The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science. 2020;368(6489):395-400. doi:10.1126/science.aba9757 6. Tian H, Liu Y, Li Y, et al. An investigation of transmission control measures during the first 50 days of the COVID-19 epidemic in China. Science. 2020;368(6491):638-642. doi:10.1126/science.abb6105 7. Gao S, Rao J, Kang Y, Liang Y, Kruse J. Mapping county-level mobility pattern changes in the United States in response to COVID-19. SIGSPATIAL Special. 2020;12(1):16-26. doi:10.1145/3404820.3404824 8. Warren MS, Skillman SW. Mobility changes in response to COVID-19. arXiv. Preprint posted online March 31, 9. Zhang L, Ghader S, Pack ML, et al. An interactive COVID-19 mobility impact and social distancing analysis platform. medRxiv. Preprint posted online May 5, 2020. doi:10.1101/2020.04.29.20085472 10. Zhou C, Su F, Pei T, et al. COVID-19: Challenges to GIS with Big Data. Geogr Sustainability. 2020;1(1):77-87. doi:10. 1016/j.geosus.2020.03.005 11. Centers for Disease Control and Prevention. Social distancing, quarantine, and isolation. Accessed April 12, 2020. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/social-distancing.html 12. Gao S, Mai G. Mobile GIS and location-based services. In: Huang B, COVA TJ, Tsou MH, eds. Comprehensive Geographic Information Systems. Elsevier; 2017:384-397. 13. Buckee CO, Balsari S, Chan J, et al. Aggregated mobility data could help fight COVID-19. Science. 2020;368 (6487):145-146. doi:10.1126/science.abb8021 14. Centers for Disease Control and Prevention. Similarities and differences between flu and COVID-19. Accessed July 10, 2020. https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm 15. GitHub. Coronadatascraper. Accessed April 12, 2020. https://github.com/covidatlas/coronadatascraper 16. US Census Bureau. American Community Survey. Accessed April 12, 2020. https://www.census.gov/programs- surveys/acs 17. US Census Bureau. TIGER/Line Shapefile, 2017, 2010 nation, U.S. Census Urban Area National. Updated October 17, 2019. Accessed April 12, 2020. https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-nation- u-s-2010-census-urban-area-national 18. Cori A, Ferguson NM, Fraser C, Cauchemez S. A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013;178(9):1505-1512. doi:10.1093/aje/kwt133 19. Thompson RN, Stockwin JE, van Gaalen RD, et al. Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics. 2019;29:100356. doi:10.1016/j.epidem.2019.100356 20. Vynnycky E, White R. An Introduction to Infectious Disease Modelling. Oxford University Press; 2010. 21. Lin J. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory. 1991;37 (1):145-151. doi:10.1109/18.61115 22. Maier BF, Brockmann D. Effective containment explains subexponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science. 2020;368(6492):742-746. doi:10.1126/science.abb4557 23. Sanche S, Lin YT, Xu C, Romero-Severson E, Hengartner N, Ke R. High contagiousness and rapid spread of severe acute respiratory syndrome coronavirus 2. Emerg Infect Dis. 2020;26(7):1470-1477. doi:10.3201/eid2607. 24. Sen S, Karaca-Mandic P, Georgiou A. Association of stay-at-home orders with COVID-19 hospitalizations in 4 states. JAMA. 2020;323(24):2522-2524. doi:10.1001/jama.2020.9176 25. Zhu J. The ten-hundred plot on COVID-19. Accessed April 12, 2020. http://pages.cs.wisc.edu/~jerryzhu/COVID19/ index.html 26. González MC, Hidalgo CA, Barabási AL. Understanding individual human mobility patterns. Nature. 2008;453 (7196):779-782. doi:10.1038/nature06958 27. Song C, Qu Z, Blumm N, Barabási AL. Limits of predictability in human mobility. Science. 2010;327(5968): 1018-1021. doi:10.1126/science.1177170 28. Kang C, Ma X, Tong D, Liu Y. Intra-urban human mobility patterns: an urban morphology perspective. Physica A. 2012;391(4):1702-1717. doi:10.1016/j.physa.2011.11.005 29. Gao S. Spatio-temporal analytics for exploring human mobility patterns and urban dynamics in the mobile age. Spatial Cogn Comp. 2015;15(2):86-114. doi:10.1080/13875868.2014.984300 30. Yuan Y, Raubal M. Analyzing the distribution of human activity space from mobile phone usage: an individual and urban-oriented study. Int J Geogr Inf Sci. 2016;30(8):1594–1621. doi:10.1080/13658816.2016.1143555 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 11/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US 31. Pullano G, Valdano E, Scarpa N, Rubrichi S, Colizza V. Population mobility reductions during COVID-19 epidemic in France under lockdown. medRxiv. Preprint published online June 1, 2020. doi:10.1101/2020.05.29.20097097 32. Huang Q, Wong DW. Activity patterns, socioeconomic status and urban spatial structure: what can social media data tell us? Int J Geogr Inf Sci. 2016;30(9):1873-1898. doi:10.1080/13658816.2016.1145225 33. Huang X, Li Z, Jiang Y, Li X, Porter D. Twitter, human mobility, and COVID-19. arXiv. Preprint published online June 24, 2020. 34. Prestby T, App J, Kang Y, Gao S. Understanding neighborhood isolation through spatial interaction network analysis using location big data. Environ Plann A. 2020;52(6):1027-1031. doi:10.1177/0308518X19891911 35. Liang Y, Gao S, Cai Y, Foutz NZ, Wu L. Calibrating the dynamic Huff model for business analysis using location big data. Trans GIS. 2020;24(3):681-703. doi:10.1111/tgis.12624 36. Wu L, Zhi Y, Sui Z, Liu Y. Intra-urban human mobility and activity transition: evidence from social media check-in data. PLoS One. 2014;9(5):e97010. doi:10.1371/journal.pone.0097010 37. Shaw SL, Tsou MH, Ye X. Human dynamics in the mobile and big data era. Int J Geogr Inf Sci. 2016;30(9): 1687-1693. doi:10.1080/13658816.2016.1164317 38. Kang Y, Zhang F, Peng W, et al Understanding house price appreciation using multi-source big geo-data and machine learning. Land Use Policy. 2020:104919. doi:10.1016/j.landusepol.2020.104919 39. Ferretti L, Wymant C, Kendall M, et al. Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020;368(6491):eabb6936. doi:10.1126/science.abb6936 40. Li R, Pei S, Chen B, et al. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science. 2020;368(6490):489-493. doi:10.1126/science.abb3221 41. Zhao Z, Shaw SL, Xu Y, Lu F, Chen J, Yin L. Understanding the bias of call detail records in human mobility research. Int J Geogr Inf Sci. 2016;30(9):1738-1762. doi:10.1080/13658816.2015.1137298 42. Jiang J, Li Q, Tu W, Shaw SL, Yue Y. A simple and direct method to analyse the influences of sampling fractions on modelling intra-city human mobility. Int J Geogr Inf Sci. 2019;33(3):618–644. doi:10.1080/13658816. 2018.1552964 43. Xu Y, Belyi A, Bojic I, Ratti C. Human mobility and socioeconomic status: analysis of Singapore and Boston. Comput Environ Urban Syst. 2018;72:51-67. doi:10.1016/j.compenvurbsys.2018.04.001 44. Li M, Gao S, Lu F, Zhang H. Reconstruction of human movement trajectories from largescale low-frequency mobile phone data. Comput Environ Urban Syst. 2019;77:101346. doi:10.1016/j.compenvurbsys.2019.101346 45. de Montjoye YA, Hidalgo CA, Verleysen M, Blondel VD. Unique in the crowd: the privacy bounds of human mobility. Sci Rep. 2013;3:1376. doi:10.1038/srep01376 46. Tsou MH. Research challenges and opportunities in mapping social media and Big Data. Cartography Geogr Inf Sci. 2015;42(supp 1):70-74. doi:10.1080/15230406.2015.1059251 47. McKenzie G, Keßler C, Andris C. Geospatial privacy and security. J Spatial Inf Sci. 2019;2019(19):53-55. doi:10. 5311/JOSIS.2019.19.608 48. de Montjoye YA, Gambs S, Blondel V, et al. On the privacy-conscientious use of mobile phone data. Sci Data. 2018;5(1):180286. doi:10.1038/sdata.2018.286 49. Jewell NP, Lewnard JA, Jewell BL. Predictive mathematical models of the COVID-19 pandemic: underlying principles and value of projections. JAMA. 2020;323(19):1893-1894. doi:10.1001/jama.2020.6585 SUPPLEMENT. eAppendix 1. Supplementary Materials and Methods eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients eFigure 2. Linear Fit for Distance vs Days eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 12/13 JAMA Network Open | Public Health Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates and COVID-19 Infection Rates in the US eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia eTable 6. Regression Results of Travel Distance Changes at States eTable 7. Regression Results of Dwell Time at Home at States eAppendix 2. Mapping Videos eReferences. JAMA Network Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 (Reprinted) September 8, 2020 13/13 Supplementary Online Content Gao S, Rao J, Kang Y, et al. Association of mobile phone location data indications of travel and stay-at- home mandates with COVID-19 infection rates in the US. JAMA Netw Open. 2020;3(9):e2020485. doi:10.1001/jamanetworkopen.2020.20485 eAppendix 1. Supplementary Materials and Methods eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients eFigure 2. Linear Fit for Distance vs Days eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March 31 eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia eTable 6. Regression Results of Travel Distance Changes at States eTable 7. Regression Results of Dwell Time at Home at States eAppendix 2. Mapping Videos eReferences This supplementary material has been provided by the authors to give readers additional information about their work. © 2020 Gao S et al. JAMA Network Open. eAppendix 1. Supplementary Materials and Methods Travel Distance Mobility Data and Home Dwell Time Data The travel distance mobility data are collected from an open-source repository released by the Descartes Labs [10], while the home dwell time data are collected from SafeGraph [1]. Both datasets are collected through tracking usages of mobile phone apps of sampled users. By aggregating large-scale (over 45 million) anonymized location data from smartphones, the average individual movement pattern for a given area can be derived. To enhance privacy, individual data are de-identified and aggregated and all analyses are performed at aggregated spatial units (e.g., census block groups, counties, and states). Travel distance mobility changes are derived using an index which measures the median of the maximum travel distances for all individuals in a given county on a given day [10, 3]. Such data are widely used to represent the dramatic human mobility changes in reaction to the COVID-19 [6, 3, 11]. To measure home dwell time, the home place for each individual is identified and the minutes for all sampled devices staying at that home place across the day are summed up. Median home dwell time for all observed devices is then aggregated in geographical units. Fitting the epidemic growth curves As the confirmed cases of COVID-19 keep growing, we aim to find a model that can describe such growth rates with respect to their temporal changes in different states. Therefore, we use four types of formulas to fit the curve of the confirmed cases for the COVID-19 epidemic. By fitting the curve, we can compare the growth rates among different states using the coefficients estimated from the model. The four models we tested are: yc(t) = t + k (1) yc(t) = t (2) yc(t) = at + k (3) bt yc(t) = ae (4) Where y is the total number of confirmed cases in each state as a function of time, t is the number of days from March 11 where the novel Coronavirus disease became a pandemic, a,b,k are parameters we will estimate. We fit the curve for two time periods: March 11 to March 31 and March 11 to April 10 with consideration of the starting date variation of stay-at-home orders in each state (in Table S5) and the incubation period ranging from 1-14 days with a median of 4 days according to clinical characteristics of COVID-19 patients [5], as well as the testing capacity at the beginning and the variability of exponential or sub-exponential growth of the COVID- 19 cases [8]. We analyzed two periods respectively. The fitting results for the period March 11 to March 31 fit better due to the fact that the majority changes of human responses in terms of mobility and home dwell time were observed after March 23, 2020 and with a short time lag across different states. The results of fitting the curve using the four formulas from March 11 to March 31 are listed in Table S1. The coefficient b is used to reflect the growth rate of the confirmed cases. We use the Mean Square Error (MSE) as the goodness of fit measure for the curve fitting. Based on the MSE, the third formula y (t) = at +k has the best fit curve for total number of confirmed cases in each state. The first formula and the fourth one have close goodness of fit and the second formula has the worst performance. We then looked at the coefficients for each state and Table S2 shows the top 10 states with the largest coefficients from each formula. A larger coefficient reflects a faster growth rate, so the top states in the Table S2 should be those that experienced the most rapid outbreaks of the COVID-19. When examining the top most infected states for the four models, the results for b b yc(t) = t + k and yc(t) = t better match the empirical observations in the study period. The top states from the two models that have the fastest growth rate are New York, New Jersey, California, and Michigan. Therefore, we use the coefficients generated from the approximate sub-exponential growth function y (t) = t + k as the indices representing the epidemic growth rate. Figure S1 shows the cases and the fitted curves using formulas yc(t) = t + k. © 2020 Gao S et al. JAMA Network Open. Fitting the mobility change rates In addition to fitting the epidemic growth curve, we also fit the curve for travel distance and home dwell time changes. More specifically, we calculate the coefficient that describes the overall trend of the median of max- distance individual mobility at each state since March 11, 2020. The coefficient found by fitting the distance curve is used to measure the speed of changes to daily mobility patterns. The analysis is conducted for each state and we fit the distance decay pattern over days from March 11 using a linear regression model: yd(t) = bt + k (5) Where y is travel distance as a function of time, and t is the number of days from March 11, 2020. The coefficient of the travel distance change is represented using the slope parameter b. In addition, we processed the travel distance data to cut off the days where the distances are smaller than 0.1 km for 3 consecutive days. The reason to do so is that the plot of distance vs. days has a long tail for some areas because most people stayed at home in late March in response to COVID-19, known as the social distancing inertia [4]. Since we try to capture the most dramatic change of potential travel distance decrease in response to the COVID- .1 km for three consecutive days. An example of fitting the distance data with and .1 km is shown in Figure S2. For some states that responded quickly and reduced the travel distance, we can capture the quick decreasing travel distance in this way. Table S3 shows the top five states with the largest absolute distance decreasing rates. From this analysis, Michigan, Washington D.C., and New York, respectively, are the states that responded quickest to COVID-19 in terms of decreased mobility. Furthermore, after we estimate the growth rate of confirmed cases and the decreasing rate of travel distance coefficient between the two coefficients using the data from March 11 to March 31 is -0.586 with a p-value <0.001, and the 95% confidence interval (CI) is -0.742 to -0.370. Figure 3A shows the two coefficients for all 50 states and DC. This negative relationship indicates that people have responded quickly to increases in confirmed cases. In places where COVID-19 cases are growing faster, people usually are quicker to reduce mobility and stay at home. We also describe the relationship between home dwell time (in minutes) and the number of confirmed cases by estimating the coefficients that represent the change of the home dwell time and compute the correlation between the dwell time coefficient and the cases coefficient. The following three types of formulas are used to fit the curve of home dwell time (y) vs. the days from March 11 (t). y(t) = t + k (6) bt (7) y(t) = ae y(t) = bt + k (8) Table S4 shows the results of fitting home dwell time. The coefficients represent human mobility changes from the perspective of how long people stay at home. The three formulas all have positive coefficients, meaning that as time goes by, home dwell time is increasing. The three formulas all have close MSE and we choose the linear fit y(t) = bt + k to represent the rate of change in home dwell time as we prefer a simpler formula when all formulas perform similarly. We then compute the correlation between the home dwell time coefficients and the confirmed cases coefficients to detect whether such mobility changes are associated with the growth of COVID-19 cases. Figure 3B shows that the growth rates of cases in each state and the associated dwell time fitted coefficients have a positive correlation of 0.526 (95% CI: [0.293, 0.700], p-value < 0.001). In addition, the correlation between the cases growth coefficients and the distance decay coefficients from the period March 11 to April 10 is very similar to the result of that for the period March 11 to March 31: -0.582 (95% CI: [-0.740, -0.365], p-value <0.001). Because we processed the distance data to cut off the days where the distance is < 0.1km and we can capture the distance decay rate even when the study period extended. For the correlation between the the cases growth coefficients and the home dwell time coefficients, however, the © 2020 Gao S et al. JAMA Network Open. correlation is 0.322 (95% CI: [0.051, -0.549], p-value <0.05), it is much smaller than the correlation we calculated for the period March 11 to March 31. Evaluating factors influencing changes in travel distance and home dwell time To understand how socio-economic factors influence changes in travel distance and home dwell time, we employed a multi-linear regression model involving the above-mentioned socio-economic variables as independent variables to model the mobility changes. The previously regressed change rates of travel distance b (Equation 5) and home dwell time b (Equation 8) are taken as the dependent variable of the following model: d t b(d,t) = a0 + a1x1 + a2x2 + a3x3 + ... + anxn (9) where x1,x2,...,xn are socio-economic factors that may influence these behaviour changes; and a0,a1,...,an are coefficients which illustrate to what degree these independent variables contribute to the behaviour changes. Since spatial scale plays an important role in socio-economic metrics, we aggregated and analyzed data at the state level. Although we perform the linear regression in two time periods, from March 11 to March 31, and from March 11 to April 10, only the results of former one are reported. Because there were more substantial mobility changes (both travel distance and home dwell time) during the former time period. It is worth noting that, at the state level, records with median travel distance less than 0.1 km are removed. In addition, in terms of different age groups and race groups, the proportion of population over age 65 and the proportion of other race groups are assigned as the reference group in the experiments. Table S6 shows the regression results for travel distance changes at the state level. The goodness-of-fit of the regression models is evaluated using R-squared (R ) and p-value. The R-squared is 0.59 for the regression, and the model is statistically significant (p-value < 0.05). Results show that behaviour changes can be largely explained by state policies and socio-economic variables. Further investigation into other confounding factors associated with travel distance reduction is necessary to quantify the effects of those covariates. Table S7 illustrates the regression results for home dwell time changes in response to socio-economic factors. The R-squared for multi-linear regression is 0.66 at state-level and the p-value is less than 0.05, which indicates that the results are statistically significant. In particular, the ratio of Asians has a positive standardized coefficient (6.959), p-value < 0.05, while the ratio of Hawaiians has a negative standardized coefficient (-2.863), p-value < 0.05, and both are significant at the state level. Accordingly, it can be inferred that the ratio of Asians is strongly correlated with longer home dwell time at the state level. The higher the proportion of Asian population, the longer the median home dwell time of residents in that state. Calculating the epidemic growth doubling time Based on the relationship between the number of confirmed cases and the date, we calculate the doubling time of the number of total confirmed cases (i.e., the time it takes for the confirmed cases to double in size) to reflect the characteristics of the COVID-19 epidemic growth. Additionally, we want to explore how the social distancing policy (e.g., stay-at-home orders) in each state makes a difference in flattening the curve. That is, the larger the doubling time, the less steep the epidemic growth curve. At the time of writing, the growth rates of COVID-19 cases in the U.S. states are either exponential or sub-exponential. We implement both of them to fit the curve for calculating the doubling time. We also calculated the doubling time based on empirical observations (model- free) to further explore how the doubling time differs in these methods. The two models we used are the exponential model (Equation 11) and the power-law model (Equation 12): at yc(t) = I(0)e (10) y (t) = bt + I(0) (11) Where yc is the total number of confirmed cases in each state; t represents the number of the days after the first case was confirmed in the state; I(0) represents the number of confirmed cases on the first-case day; a,b are the © 2020 Gao S et al. JAMA Network Open. coefficients of the models. Note that we introduced the intercept term I(0) into the two models to ensure that the initial case number on the first day is accurate. This is important since the accuracy of the thereafter calculated daily growth rate and doubling time relies on the number of initial cases. We investigate how stay-at-home orders in each state associated with the doubling time of confirmed cases in that state. Within the time frame of our study, as shown in Table S5, there were five states (North Dakota, ed a stay-at-home order, and three states (Oklahoma, Utah, and Wyoming) which only had partial orders issued locally by cities or counties. All other states, as well as the District of Columbia, issued the order for all residents. South Carolina issued partial orders effective March 26, and then issued the full order effective April 7. We focus on the confirmed case data from March 11 to April 10, and we use the effective date of the stay-at-home order to split the confirmed case data into two parts: before the order and after the order. For the states with only partial orders, we use the earliest order effective date in that state. For South Carolina, we also use the earliest effective date (i.e., March 26). We investigate the doubling time in the states with statewide orders (except for Missouri since the effective date of the stay-at- -order data in this state given the date range of our study) and the District of Columbia. We fit the models on the data before the order and after the order respectively, and then we calculate the doubling time of the confirmed cases based on the two models. The doubling time of the confirmed cases is defined as: (12) Where d(t) stands for the doubling time of the cumulative infection cases on date t in each state; r(t) represents the growth rate of the cumulative infection cases on date t in each state, and is defined as: (13) Where I(t) represents the number of cumulative infection cases on date t. For the exponential model, the daily growth rate is constant, and thus the doubling time is constant and can be calculated as: (14) Where d is the doubling time of the exponential model, and a is the exponent parameter in the exponential model (Equation 10). Note that Equation (14) can be regarded as a special case of Equation (12). For the power- law model (Equation 11), since the doubling time is not constant, we first manually calculated the time- dependent growth rate of the predicted cases by the model in each day for each state. Specifically, for each date icted case number, thereby obtaining the time- dependent daily growth rate r(t) (Equation 13). We then use the time-dependent growth rate to calculate the corresponding time-dependent doubling time of the power-law model, and we use the median of the time- dependent doubling time to represent the doubling time in that state during the before or after order period. For the two curve fitting models, the fitting results are shown in Figures S3 and S4. The green dashed line in each subplot represents the fitted curve on the data before the stay-at-home order in the state; the blue curve represents the fitted curve on the data after the stay-at-home order in the state; the vertical black dashed line indicates the date when the stay-at-home order takes effect in each state. All the states showed a large initial growth rate and small doubling time in the number of total confirmed cases. However, the growth starts to slow down in general after the effective date of the stay-at-home order in each state, which shows the efficiency of the social distancing policy in suppressing the transmission of the novel coronavirus. For empirical observations, we manually calculated the time-dependent growth rate of the reported cumulative confirmed cases in each day for each state. The empirical growth rate is also calculated following the same process as the power-law model. For each date except for the starting date, we subtract the previous divided by the © 2020 Gao S et al. JAMA Network Open. -dependent daily empirical growth rate. We then use the empirical time-dependent growth rate to calculate the corresponding empirical time- dependent doubling time, and we use the median of the time-dependent doubling time to represent the empirical doubling time in that state. In addition, we visualize and investigate the overall probability density distribution of the median doubling time before and after the order in each state to have a better understanding of the overall changes in the epidemic growth nationwide. We applied Kernel Density Estimation (KDE) on doubling times in all the states to derive the overall probability density distribution, where the doubling time for each state is calculated by the median of the daily growth rate in that state. Note that the doubling time for the exponential model is a constant, so the median equals that constant. As is shown in Figure 4A, S5 and S6, the doubling time of confirmed cases nationwide has significantly increased after the order in each state (empirical observations: from median 2.7 days, IQR: 1.0 to median 6.0 days, IQR: 2.3; exponential model: from median 2.6 days, IQR: 0.8 to median 5.7 days, IQR: 2.2; power-law model: from median 2.7 days, IQR: 0.9 to median 6.3 days, IQR: 2.5)), which reinforce the point that stay-at-home orders, or social distancing policy, associated with the reduced spread of the virus when they are compliant. Furthermore, we measure the difference in probability density distribution between the fitting results and the real data using Jensen Shannon Divergence (JSD). The JSD value ranges from 0 to 1 for two probability distributions; small JSD values indicate high similarity between two probability distributions. Compared with the probability density distribution of real data shown in Figure 4A, both the fitting results by the exponential model and power-law model show high similarity (i.e., with small JSD values) to the empirical data before the order (the exponential model: 0.11; the power-law model: 0.07) and after the order (the exponential model: 0.17; the power-law model: 0.18), respectively. This means that, despite different models, the discoveries are consistent: stay-at-home orders did associate with the slowdown of the COVID-19 case growth and with the increase of doubling time. Mechanistic prediction models The exponential equation approach was particularly suitable during the early outbreak phase but the sub- exponential growth fitted better after the city lock-downs [8], stay-at-home orders and following the social distancing regulations. Our curve fitting results matched the outcomes of mechanistic prediction models (Figure S7), such as the models reported by [2, 9]. The daily number of cumulative confirmed cases per U.S. state and the serial interval (SI), that is the number of days between two consecutive cases, were used assuming an average SI of 7.5 days with standard deviations (SD): 3.4 days as reported by [7], who estimated the SI using Bayesian parametric estimation from data in the city of Wuhan in China. In addition, the SI average and SD were sampled from normal distributions of mean 7.5 and SD 0.2, as well as SD mean of 3.4 and SD of 0.4 to sample from more SI distributions at random. Estimates of the instantaneous basic reproduction number (R0) were simulated by moving an 8-day sliding time interval across all data points since the start of the outbreak per state respectively. The instantaneous R0 on the last day before the start of the stay-at-home orders and from April 10, 2020, that is after the order went in, were used for projections of the confirmed number of cases per day. The resulting graphs are shown in Figure S7 for all states respectively where black dots represent observed values and the red dots represent the projected values for daily total confirmed cases. It shows that projected daily total confirmed cases from before the start date of the stay-at-home social distancing orders increase much more rapidly compared to the projected number of total confirmed cases starting April 10, 2020, that is after the orders went in. In summary, the mathematical curve fitting models and the mechanistic epidemic prediction models draw the same conclusion. © 2020 Gao S et al. JAMA Network Open. eFigure 1. Curve Fitting Results of Total Number of COVID-19 Cases for the Top 5 States With the Largest Coefficients A: New York; B: New Jersey; C: California; D: Michigan; E: Massachusetts. © 2020 Gao S et al. JAMA Network Open. eFigure 2. Linear Fit for Distance vs Days (a) (b) (a) using the original .1 km in New York state from March 11 to March 31. © 2020 Gao S et al. JAMA Network Open. eFigure 3. Curve Fitting Results Using the Exponential Growth Model for Each State The green dashed line and the blue line represent the fitted curves on the data before and after the stay-at- home orders in each state, respectively; the vertical black dashed line indicates the effective date of the stay-at- home orders in each state. dt and dt represent the median doubling time before and after the order in before after each state. © 2020 Gao S et al. JAMA Network Open. eFigure 4. Curve Fitting Results Using the Power-Law Growth Model for Each State The green dashed line and the blue line represent the fitted curves on the data before and after the stay-at- home order in each state, respectively; the vertical black dashed line indicates the effective date of the stay-at- home order in each state. dt and dt represent the median doubling time before and after the order in before after each state. © 2020 Gao S et al. JAMA Network Open. eFigure 5. Overall Changes of the Doubling Time Nationwide Before and After Stay-at-Home Orders Using the Exponential Fitting Model Figure S5: The overall changes of the doubling time nationwide before and after the order using the exponential fitting model. © 2020 Gao S et al. JAMA Network Open. eFigure 6. Overall Changes of the Doubling Time Nationwide Before and After the Stay-at-Home Orders Using the Power-Law Fitting Model © 2020 Gao S et al. JAMA Network Open. eFigure 7. Empirical Observations of Confirmed Cases in the 45 States and the District of Columbia and the Projection of Cases Using the Mechanistic Prediction Model Before and After the Stay-at-Home Orders © 2020 Gao S et al. JAMA Network Open. eTable 1. Coefficient and Mean Squared Error for Models Fitting the Confirmed Cases From March 11 to March 31 The fitting method Coefficient b (mean) MSE yc = t + k 2.284 2.694E+05 yc = t 2.259 3.779E+05 y = at + k 3.210 4.336E+04 bt yc = ae 0.191 2.960E+05 © 2020 Gao S et al. JAMA Network Open. eTable 2. Top 10 States with Highest Coefficients for the Confirmed Cases Across 4 Models From March 11 to March 31 b b b bt y = t + k yc = t y = at + k yc = ae c c New York New York Rhode Island Rhode Island New Jersey New Jersey Idaho Idaho California California Indiana Indiana Michigan Michigan Pennsylvania Pennsylvania Massachusetts Florida Massachusetts New Jersey Florida Massachusetts New Jersey Missouri Illinois Washington Maryland Massachusetts Louisiana Illinois Iowa Texas Washington Louisiana Missouri Maryland Pennsylvania Pennsylvania Texas Michigan © 2020 Gao S et al. JAMA Network Open. eTable 3. Top 5 States With the Largest Absolute Distance Coefficients From March 11 to March 31 State Distance coefficient Michigan -0.494 Washington, D.C. -0.485 New York -0.478 Missouri -0.452 Delaware -0.450 © 2020 Gao S et al. JAMA Network Open. eTable 4. Coefficient and Mean Squared Error for Models Fitting the Dwell Time at Home The fitting method Coefficient b MSE y = t + k 1.677 4850.671 bt y = ae 0.767 4791.662 y = bt + k 8.608 4864.841 © 2020 Gao S et al. JAMA Network Open. eTable 5. Order Type and Effective Date of Stay-At-Home Orders in Each State and the District of Columbia State Name Order Type Earliest Effective Date California Full 3/19/2020 Illinois Full 3/21/2020 New Jersey Full 3/21/2020 New York Full 3/22/2020 Connecticut Full 3/23/2020 Louisiana Full 3/23/2020 Ohio Full 3/23/2020 Oregon Full 3/23/2020 Washington Full 3/23/2020 Delaware Full 3/24/2020 Indiana Full 3/24/2020 Massachusetts Full 3/24/2020 Michigan Full 3/24/2020 New Mexico Full 3/24/2020 West Virginia Full 3/24/2020 Hawaii Full 3/25/2020 Idaho Full 3/25/2020 Oklahoma Partial 3/25/2020 Vermont Full 3/25/2020 Wisconsin Full 3/25/2020 Colorado Full 3/26/2020 Kentucky Full 3/26/2020 South Carolina* Partial, Full 3/26/2020, 4/7/2020 Minnesota Full 3/27/2020 New Hampshire Full 3/27/2020 Utah Partial 3/27/2020 Alaska Full 3/28/2020 Montana Full 3/28/2020 Rhode Island Full 3/28/2020 Wyoming Partial 3/28/2020 Kansas Full 3/30/2020 Maryland Full 3/30/2020 North Carolina Full 3/30/2020 Virginia Full 3/30/2020 Arizona Full 3/31/2020 Tennessee Full 3/31/2020 Washington, D.C. Full 4/1/2020 Nevada Full 4/1/2020 Pennsylvania Full 4/1/2020 Maine Full 4/2/2020 Texas Full 4/2/2020 Florida Full 4/3/2020 Georgia Full 4/3/2020 Mississippi Full 4/3/2020 Alabama Full 4/4/2020 Missouri Full 4/6/2020 North Dakota No - South Dakota No - Nebraska No - Iowa No - © 2020 Gao S et al. JAMA Network Open. Arkansas No - -at-home order in this state. *South Carolina issued partial orders effective March 26, and then issued the full order effective April 7. We use the earliest effective date for South Carolina. © 2020 Gao S et al. JAMA Network Open. eTable 6. Regression Results of Travel Distance Changes at States Variables Coefficients Standard Error Standardized Coefficients Intercept 1.116 2.513 0.444 0.660 -0.364 Population 0.000 0.000 -1.688 0.100 -0.022 Median Age -0.025 0.029 -0.853 0.399 -0.053 Population Density 0.000 0.000 -1.177 0.247 -0.026 Household Income 0.000 0.000 1.042 0.304 0.014 Proportion of Population under Age 18 -2.438 2.801 -0.870 0.390 -0.051 Proportion of Population between Age 18 to 44 -1.680 2.133 -0.788 0.436 -0.041 Proportion of Population between Age 45 to 64 -3.147 1.670 -1.884 0.067 -0.057 Proportion of Population over Age 65 0.000 0.000 Proportion of Whites 1.029 0.623 1.652 0.107 0.140 Proportion of Blacks 0.676 0.583 1.160 0.253 0.073 Proportion of Asians 2.786 1.399 1.991 0.054 0.153 Proportion of Natives 2.473 0.954 2.593 0.014 0.066 Proportion of Hawaiian -3.191 2.250 -1.419 0.164 -0.046 Proportion of Other Race Groups 0.000 0.000 1.024 0.961 1.065 0.294 0.023 R : 0.59, p-value < 0.05 © 2020 Gao S et al. JAMA Network Open. eTable 7. Regression Results of Dwell Time at Home at States Variables Coefficients Standard Error Standardized Coefficients Intercept 61.218 70.426 0.869 0.390 8.814 Population 0.000 0.000 -1.476 0.148 -0.544 Median Age -0.716 0.824 -0.868 0.391 -1.508 Population Density 0.003 0.002 1.327 0.193 0.825 Household Income -0.007 0.008 -0.907 0.370 -0.351 Proportion of Population under Age 18 -39.540 78.514 -0.504 0.618 -0.825 Proportion of Population between Age 18 to 44 -79.376 59.786 -1.328 0.192 -1.925 Proportion of Population between Age 45 to 64 -7.773 46.809 -0.166 0.869 -0.142 Proportion of Population over Age 65 0.000 0.000 Proportion of Whites 26.084 17.457 1.494 0.144 3.551 Proportion of Blacks 25.046 16.326 1.534 0.134 2.715 Proportion of Asians 126.581 39.213 3.228 0.003 6.959 Proportion of Natives 25.069 26.732 0.938 0.354 0.672 Proportion of Hawaiian -200.726 63.055 -3.183 0.003 -2.863 Proportion of Other Race Groups 0.000 0.000 -20.269 26.941 -0.752 0.457 -0.449 R : 0.66, p-value < 0.05 © 2020 Gao S et al. JAMA Network Open. eAppendix 2. Mapping Videos (1) Mapping the COVID-19 infected areas in the U.S., available at: "https://geods.geography.wisc.edu/wp-content/uploads/2020/04/US_cases_animation_March.mov" (2) Mapping human mobility changes at the U.S. county level, available at: "https://geods.geography.wisc.edu/wp-content/uploads/2020/04/US_mobilitychanges_animation. mp4" © 2020 Gao S et al. JAMA Network Open. eReferences [1] SafeGraph Inc. Accessed April 12, 2020, https://www.safegraph.com/. [2] A. Cori, N. M. Ferguson, C. Fraser, and S. Cauchemez. A new framework and software to estimate timevarying reproduction numbers during epidemics. American Journal of Epidemiology, 178(9):1505 1512, 2013. [3] S. Gao, J. Rao, Y. Kang, Y. Liang, and J. Kruse. Mapping county-level mobility pattern changes in the united states in response to covid-19. SIGSPATIAL Special, 12(1):16 26, 2020. [4] S. Ghader, J. Zhao, M. Lee, W. Zhou, G. Zhao, and L. Zhang. Observed mobility behavior data reveal social distancing inertia. arXiv preprint arXiv:2004.14748, 2020. [5] W.-j. Guan, Z.-y. Ni, Y. Hu, W.-h. Liang, C.-q. Ou, J.-x. He, L. Liu, H. Shan, C.-l. Lei, D. S. Hui, et al. Clinical characteristics of coronavirus disease 2019 in China. New England Journal of Medicine, 2020. [6] J. R. Hipp and A. Boessen. The shape of mobility: Measuring the distance decay function of household mobility. The Professional Geographer, 69(1):32 44, 2017. [7] R. Li, S. Pei, B. Chen, Y. Song, T. Zhang, W. Yang, and J. Shaman. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science, 2020. [8] B. F. Maier and D. Brockmann. Effective containment explains sub-exponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China. Science, 2020. [9] R. Thompson, J. Stockwin, R. van Gaalen, J. Polonsky, Z. Kamvar, P. Demarsh, E. Dahlqwist, S. Li, E. Miguel, T. Jombart, et al. Improved inference of time-varying reproduction numbers during infectious disease outbreaks. Epidemics, 29:100356, 2019. [10] M. S. Warren and S. W. Skillman. Mobility changes in response to covid-19. arXiv preprint arXiv:2003.14228, 2020. [11] L. Zhang, S. Ghader, M. L. Pack, C. Xiong, A. Darzi, M. Yang, Q. Sun, A. Kabiri, and S. Hu. An interactive covid- 19 mobility impact and social distancing analysis platform. medRxiv, 2020. © 2020 Gao S et al. JAMA Network Open.

Journal

JAMA Network Open – American Medical Association

Published: Sep 8, 2020

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

Association of Mobile Phone Location Data Indications of Travel and Stay-at-Home Mandates With COVID-19 Infection Rates in the US

References (102)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies