Evaluating the impact of international airline suspensions on the early global spread of COVID-19

Aniruddha Adiga, Srinivasan Venkatramanan, James Schlitt, Akhil Sai Peddireddy, Allan Dickerman, Andrei C. Bura, Andrew Warren, Brian Klahn, Chunhong Mao, Dawen Xie, Dustin Machi, Erin Raymond, Fanchao Meng, Golda H. Barrow, Henning Mortveit, Jiangzhuo Chen, Jim Walke, Joshua Goldstein, Mandy Wilson, Mark Orr, Przemyslaw Porebski, Pyrros A. Telionis, Richard J. Beckman, Stefan Hoops, Stephen Eubank, Young Yun Baek, Bryan Lewis, Madhav Marathe, Chris BarrettmedRxiv 2020 · DOI

Abstract

Global airline networks play a key role in the global importation of emerging infectious diseases. Detailed information on air traffic between international airports has been demonstrated to be useful in retrospectively validating and prospectively predicting case emergence in other countries. In this paper, we use a well-established metric known as effective distance on the global air traffic data from IATA to quantify risk of emergence for different countries as a consequence of direct importation from China, and compare it against arrival times for the first 24 countries. Using this model trained on official first reports from WHO, we estimate time of arrival (ToA) for all other countries. We then incorporate data on airline suspensions to recompute the effective distance and assess the effect of such cancellations in delaying the estimated arrival time for all other countries. Finally we use the infectious disease vulnerability indices to explain some of the estimated reporting delays.

Introduction

In the early weeks of 2020, as COVID-19 transitioned from a localized outbreak in Wuhan to a global pandemic, public health officials faced the urgent challenge of predicting the virus's trajectory. Traditional geographic distance is often a poor predictor of disease spread in the modern era; a city thousands of miles away with a direct flight hub is effectively "closer" than a neighboring town across a closed land border. This study addresses the challenge of quantifying "network distance" to predict the Time of Arrival (ToA) for the virus and evaluates the efficacy of the massive wave of airline suspensions implemented in early February 2020.

Methods

The researchers utilized the concept of Effective Distance, a metric that measures the probabilistic distance between two points based on passenger traffic flow rather than physical miles. Using IATA passenger data from February 2019 as a proxy for Lunar New Year travel, they developed a linear model to correlate effective distance from China with official first reports of COVID-19 from the WHO.

To assess the impact of travel restrictions, the global flow network was recomputed based on airline suspension data as of February 10, 2020. Finally, the Infectious Disease Vulnerability Index (IDVI)—a RAND Corporation metric scoring countries on political, economic, and public health factors—was used to investigate "reporting delays," defined as the gap between the model's predicted arrival and a country's official report.

Results

The linear model demonstrated high predictive power, achieving a coefficient of determination ($R^2$) of 0.78 for the first 24 countries to report cases. The analysis of airline suspensions revealed that these interventions provided a modest "buffer," delaying the estimated arrival of the virus by an average of 4 to 5 days globally. However, regional variations were significant: countries in the AFRICOM (Africa) region saw average delays of nearly 11 days, while specific hubs like Ethiopia and Qatar experienced delays exceeding 10 days.

Figure 3: (a) Correlation between effective distance and Time of Arrival (ToA) for the first 24 countries; (b) Estimated delay in ToA due to airline suspensions by country.

The study also identified a significant "reporting gap" influenced by national infrastructure. Developed economies with higher IDVI scores (indicating lower vulnerability) tended to report cases earlier. This suggests that these nations possess the surveillance infrastructure and transparency necessary to detect and report cases closer to the actual time of arrival.

Figure 5: Relationship between the Infectious Disease Vulnerability Index (IDVI) and Estimated Time of Arrival (ToA), categorized by Combatant Command.

Statistical analysis further revealed that political factors, including domestic governance and international collaboration, were the strongest predictors of reporting delays, explaining 24% and 18% of the variance, respectively.

Discussion and Limitations

While effective distance is a robust predictor, the model underestimated arrival times for China's immediate neighbors, such as Japan and Thailand. This discrepancy likely stems from the omission of multi-modal transport (land and sea) which is significant for adjacent territories. Furthermore, the reliance on official WHO reports introduces a detection bias; in countries with low surveillance capacity, the virus likely arrived much earlier than the official "Day 0" report. Future iterations of this model should incorporate real-time passenger behavior and multi-modal travel data to improve predictive accuracy for future emerging infectious diseases.

Summary updated

Abstract

Global airline networks play a key role in the global importation of emerging infectious diseases. Detailed information on air traffic between international airports has been demonstrated to be useful in retrospectively validating and prospectively predicting case emergence in other countries. In this paper, we use a well-established metric known as effective distance on the global air traffic data from IATA to quantify risk of emergence for different countries as a consequence of direct importation from China, and compare it against arrival times for the first 24 countries. Using this model trained on official first reports from WHO, we estimate time of arrival (ToA) for all other countries. We then incorporate data on airline suspensions to recompute the effective distance and assess the effect of such cancellations in delaying the estimated arrival time for all other countries. Finally we use the infectious disease vulnerability indices to explain some of the estimated reporting delays.
BibTeX
@misc{adiga2020evaluating,
  author = {Aniruddha Adiga and Srinivasan Venkatramanan and James Schlitt and Akhil Sai Peddireddy and Allan Dickerman and Andrei C. Bura and Andrew Warren and Brian Klahn and Chunhong Mao and Dawen Xie and Dustin Machi and Erin Raymond and Fanchao Meng and Golda H. Barrow and Henning Mortveit and Jiangzhuo Chen and Jim Walke and Joshua Goldstein and Mandy Wilson and Mark Orr and Przemyslaw Porebski and Pyrros A. Telionis and Richard J. Beckman and Stefan Hoops and Stephen Eubank and Young Yun Baek and Bryan Lewis and Madhav Marathe and Chris Barrett},
  title = {{Evaluating the impact of international airline suspensions on the early global spread of COVID-19}},
  year = {2020},
  doi = {10.1101/2020.02.20.20025882},
  howpublished = {medRxiv},
}
Figure