Insights into population behavior during the COVID-19 pandemic from cell phone mobility data and manifold learning

Abstract

Understanding the complex interplay between human behavior, disease transmission and non-pharmaceutical interventions during the COVID-19 pandemic could provide valuable insights with which to focus future public health efforts. Cell phone mobility data offer a modern measurement instrument to investigate human mobility and behavior at an unprecedented scale. We investigate aggregated and anonymized mobility data, which measure how populations at the census-block-group geographic scale stayed at home in California, Georgia, Texas and Washington from the beginning of the pandemic. Using manifold learning techniques, we show that a low-dimensional embedding enables the identification of patterns of mobility behavior that align with stay-at-home orders, correlate with socioeconomic factors, cluster geographically, reveal subpopulations that probably migrated out of urban areas and, importantly, link to COVID-19 case counts. The analysis and approach provide local epidemiologists a framework for interpreting mobility data and behavior to inform policy makers’ decision-making aimed at curbing the spread of COVID-19.

Examining Social Vulnerability and the Association With COVID-19 Incidence in Harris County, Texas

“…The stay-at-home data is approximated from smartphone location data and is most likely missing not at random. Though missingness may not be random, the data provider corrects for this ( 18 ). Additionally, there may be limitations in the accuracy and precision of the measurements used.…”

“…The percent of individuals fully staying-at-home was acquired from SafeGraph, a company that aggregates anonymized smartphone location data in the United States. SafeGraph compiles individuals' cell phone data into aggregate measures on the census tract-level over time and then calculates the proportion of smartphone users who spent all day at home for each date based on inferring the user's overnight location during the previous six weeks ( 18 ). Proportion staying-at-home was calculated by cumulatively averaging the measures over the study time by census tract.…”

Viral Viruses and Modified Mobility: Cyberspace Disease Salience Predicts Human Movement Patterns

“…In the current research, we develop a framework to examine how cyberspace disease salience is related to macro-and micro-level mobility patterns. Most existing studies on related topics focus primarily on the relationship between mobility patterns, typically measured using cell phone data, and SARS-CoV-2 transmission and adherence to preventive measures to slow the spread of the virus (Levin et al, 2021;Nouvellet et al, 2021;Van Bavel et al, 2022). These data, often aggregated at the country level, are insufficient to examine people's mobility behavior at the micro-level.…”

Evaluation of non-pharmaceutical interventions and optimal strategies for containing the COVID-19 pandemic

“…Utilizing compartmental models, scholars simulated and predicted the effects of NPIs on multiple epidemic indicators, including infections 3,4,[6][7][8][9][10][11] , deaths 3,[8][9][10]12 , the reproduction number 3,4,9,13 , and demand for hospital services 6,10 . Touching on similar themes, another line of research focused more on quantifying the effects of NPIs on mobility [14][15][16] and further explored the relationship between human movements and COVID-19 transmission 7,[17][18][19] . Apart from characterizing the unfolding of the COVID-19 pandemic from an epidemiological perspective, there are also studies investigating the impact of NPIs from various angles covering economic contraction 20,[20][21][22] , social issues 23,24 , and mental health [25][26][27][28] .…”