Global-scale modeling of early factors and country-specific trajectories of COVID-19 incidence: a cross-sectional study of the first 6 months of the pandemic

Studies examining factors responsible for COVID-19 incidence are mostly
focused at the national or sub-national level. A global-level
characterization of contributing factors and temporal trajectories of
disease incidence is lacking. Here we conducted a global-scale analysis of
COVID-19 infections to identify key factors associated with early disease
incidence. Additionally, we compared longitudinal trends of COVID-19
incidence at a per-country level and classified countries based on
COVID-19 incidence trajectories and effects of lockdown responses.
Univariate analysis identified eleven variables as independently
associated with COVID-19 infections at a global level (p<1e-05).
Multivariable analysis identified a 4-variable model as optimal for
explaining global variations in COVID-19 (p<0.01). COVID-19 case
trajectories for most countries were best captured by a log-logistic
model, as determined by AIC estimates. Six predominant country clusters
were identified when characterizing the effects of lockdown intervals on
variations in COVID-19 new cases per country. Globally, economic and
meteorological factors are important determinants of early COVID-19
incidence. Analysis of longitudinal trends and lockdown effects on
COVID-19 highlights important nuances in country-specific responses to
infections. These results provide valuable insights into disease incidence
at a per-country level, possibly allowing for more informed decision
making by individual governments in future disease outbreaks.