The Effect of GDP and Distance on Timing of COVID-19 Spread in Chinese Provinces in 2020

The geographical spread of COVID-19 across Chinas provinces provides the opportunity for retrospective analysis on contributors to the timing of the spread. Highly contagious diseases need to be seeded into populations and we hypothesized that greater distance from the epicenter in Wuhan, as well as higher province-level GDP per capita, would delay the time until a province detected COVID-19 cases. To test this hypothesis, we used province-level socioeconomic data such as GDP per capita and percentage of the population aged over 65, distance from the Wuhan epicenter, and health systems capacity in a Cox proportional hazards analysis of the determinants of each provinces time until epidemic start. The start was defined by the number of days it took for each province to reach thresholds of 3, 5, 10, or 100 cases. We controlled for the number of hospital beds and physicians as these could influence the speed of case detection. Surprisingly, none of the explanatory variables had a statistically significant effect on the time it took for each province to get its first cases; the timing of COVID-19 spread appears to have been random with respect to distance, GDP, demography, and the strength of the health system. Looking to other factors, such as travel, policy, and lockdown measures, could provide additional insights on realizing most critical factors in the timing of spread.