Identifying Locations with Possible Undetected Imported Severe Acute Respiratory Syndrome Coronavirus 2 Cases by Using Importation Predictions

Abstract

Cases of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection exported from mainland China could lead to self-sustained outbreaks in other countries. By February 2020, several countries were reporting imported SARS-CoV-2 cases. To contain the virus, early detection of imported SARS-CoV-2 cases is critical. We used air travel volume estimates from Wuhan, China, to international destinations and a generalized linear regression model to identify locations that could have undetected imported cases. Our model can be adjusted to account for exportation of cases from other locations as the virus spreads and more information on importations and transmission becomes available. Early detection and appropriate control measures can reduce the risk for transmission in all locations.

Keywords: 2019 novel coronavirus disease; COVID-19; SARS-CoV-2; coronavirus; outbreak; pneumonia; respiratory infections; severe acute respiratory syndrome coronavirus 2; travelers health; viruses; zoonoses.

Figure 3

Formula.

Figure 1

Plot showing imported-and-reported cases of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) against air travel volume (no. persons/day) from Wuhan, China. No. cases refers to possible imported-and-reported SARS-CoV-2 cases. Solid line indicates the expected imported-and-reported case counts for locations based on the model fit to high surveillance locations (slope = 3.3 cases/100 passengers; p<0.001) . Dashed lines represent for the same model the smoothed 95% prediction interval bounds. Purple dots indicate locations with high surveillance capacity according to category 2 of the Global Health Security Index. Cluster A is composed of Nepal, Sri Lanka, Finland, and Sweden, locations with 1 imported-and-reported case and air travel volume of <20 passengers per day. Cluster B is composed of 161 locations with no imported-and-reported cases and estimated air travel <10 passengers per day. GER, Germany; NZ, New Zealand; RUS, Russia; UAE, United Arab Emirates; UK, United Kingdom; USA, United States of America.

Figure 2

Analyses of imported-and-reported cases and daily air travel volume using a model to predict locations with potentially undetected cases of severe acute respiratory virus 2 (SARS-CoV-2). Air travel volume measured in number of persons/day. No. cases refers to possible undetected imported SARS-CoV-2 cases. Solid line shows the expected imported-and-reported case counts based on our model fitted to high surveillance locations, indicated by purple dots. Dashed lines indicate the 95% prediction interval bounds smoothed for all locations, including those with low surveillance capacity, indicated by light blue dots. A–C) Regressions that set the daily air travel volume for locations not listed by S. Lai et al. (unpub. data, https://doi.org/10.1101/2020.02.04.20020479): A) air travel volume set to 0.1 passenger/day; B) volume set to 1 passengers/day; C) volume set to 6 passengers/day. D) Regression removing locations not listed by Lai et al. before fitting. E) Regression defining high surveillance locations by using a more lenient Global Health Security (GHS) Index criterion (50th quantile) for category 2, Early Detection and Reporting of Epidemics of Potential International Concern, to define high surveillance locations. F) A more stringent GHS Index criterion (95th quantile) to define high surveillance locations. G) Regression using a negative binomial likelihood and estimated dispersion parameter of 1.27 (p = 0.097). H) Regression excluding Thailand from the model fit. Across all 8 regression analyses, Singapore lies above the 95% PI and Thailand and Indonesia lie below. India remained above 95% PI for all regressions, except when we used a more stringent GHS Index criterion (panel F). IDN, Indonesia; IND, India; SGP, Singapore; THA, Thailand.