Estimating the effect of South Africa travel restrictions in November 2021 on the SARS-CoV-2 Omicron outbreak in the Netherlands

Abstract

Background Governments used travel bans during the COVID-19 pandemic to limit the introduction of new variant of concern (VoC). In the Netherlands, direct flights from South Africa were banned from 26 November 2021 onwards to curb Omicron (B.1.1.529) importation. Objectives This study retrospectively evaluated the effect of the South African travel ban and the timing of its implementation on subsequent Omicron infections in the Netherlands and, in order to help inform future decision-making, assessed alternative scenarios in which the reproduction number (R e) and volume of indirectly imported cases were varied. Design Descriptive analysis and modelling study. Outcome measure Time (days) from 26 November 2021 to reach 10 000 cumulative Omicron infections in the Netherlands. Methods To benchmark the direct importation rate of Omicron from South Africa, we used the proportion (n/N, %) of passengers arriving on two direct flights from South Africa to the Netherlands on 26 November 2021 with a positive PCR sequencing result for Omicron VoC infection. We scaled the number of directly-imported Omicron infections before and after the travel ban to the incidence in South Africa. We assumed that 10% of all cases continued to arrive via indirect routes, a 'failure rate' of 2% (ie, incoming Dutch citizens not adhering to quarantine on arrival) and an effective reproduction number (R e) of Omicron of 1.3. In subsequent analyses, we varied, within plausible limits, the R e (1.1-2.0) and proportion of indirectly-imported cases (0-20%). Results Compared with no travel ban, the travel ban achieved a 14-day delay in reaching 10 000 Omicron cases, with an additional day of delay if initiated 2 days earlier. If all indirect importation had been prevented (eg, European-wide travel ban), a 21-day delay could have been achieved. The travel ban's effect was negligible if R e was ≥2.0 and with a greater volume of ongoing importation. Conclusions Travel bans can delay the calendar timing of an outbreak but are substantially less effective for pathogens where importation cannot be fully controlled and tracing every imported case is unfeasible. When facing future disease outbreaks, we urge policy-makers to critically weigh up benefits against the known socioeconomic drawbacks of international travel restrictions.