Back

Cryptic spread of SARS-CoV-2 undermined targeted travel restrictions

Gangavarapu, P.; Abdo, J. M.; Parker, E.; Zeller, M.; Robles-Sikisaka, R.; Kurzban, E.; Perkins, S.; Schwab, M.; Nguyen, K.; McGraw, M.; Ramesh, K.; Anderson, C.; Rasheed, N. A.; Isa, R. H.; Taher, S. M.; Gangavarapu, K.; Andersen, K. G.

2025-09-08 epidemiology
10.1101/2025.08.29.25334701 medRxiv
Show abstract

The effectiveness of travel restrictions in mitigating the spread of novel viral variants remains understudied, particularly in regions with sparse genomic surveillance. We evaluated the efficacy of these restrictions in Western Asia, where countries sequenced less than one percent of SARS-CoV-2 cases in 2021, focusing on Iraqs travel ban on India and the tourist travel restriction. We generated 535 SARS-CoV-2 whole-genome sequences from clinical samples collected in Duhok, Iraq, between June 2021 and March 2022, encompassing the Delta and Omicron variant waves. Using discrete Bayesian phylogeographic modeling, we reconstructed spatiotemporal viral spread patterns to determine whether travel restrictions delayed the introduction of the variants into the country, postponed community transmission onset, or altered the geographic sources of viral introductions over time. We found that the Delta variant was introduced into Iraq three months before the targeted travel ban was enacted, predominantly from Western and Southern Asia, whereas the Omicron variant was introduced despite the tourist travel restriction already being in place. We also show that both variants had established community transmission within the country despite travel restrictions, rendering these measures ineffective. Analysis of travel restrictions on a global scale reveals similar patterns of limited efficacy in mitigating the introductions of novel SARS-CoV-2 variants, challenging the usefulness of such measures.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.