Summary of findings 1. Travel restrictions reducing or stopping cross‐border travel.

Disease: COVID‐19 Interventions: implementing travel restrictions reducing/stopping cross‐border travel; maintaining the measure; early implementation of the measure; implementing a highly stringent measure Comparators: no measure; relaxation of the measure; late implementation of the measure; implementing a less stringent measure
Outcome	Number of studies	Summary of findings	Certainty of evidence
Outcome category: cases avoided due to measure
Number or proportion of cases in the community	13 modelling studies	Ten out of 13 studies reported reductions in the number or proportion of cases resulting from various travel restrictions. These positive effects ranged from a 1.8% (95% CI ‐21.9% to 17.5%) reduction to a 97.8% reduction. The remaining three studies reported mixed effects, including a positive effect,no effect or even a negative effect. The variation in the magnitude of effect might be explained by the level of community transmission, implementation of community‐based interventions, and the countries restricted by the measure.	Very low a,b,c ⨁◯◯◯
Number or proportion of imported or exported cases	9 modelling studies	Eight out of nine studies reported reductions in importations or exportations. These positive effects ranged from a 18% reduction to a 99% reduction. One study reported mixed effects, observing both positive effects and no effect. The variation in the magnitude and direction of effect might be explained by differences in travel volumes, the timing of implementation, the comprehensiveness and severity of the measure implemented.	Very lowb,c,d ⨁◯◯◯
Number or proportion of deaths	3 modelling studies	All studies showed reductions in deaths. These positive effects ranged from a 4.3% (95% CI ‐39.1% to 39.1%) reduction to a 98% reduction in deaths. The variation in the magnitude of effect across studies might be explained by differences in the implementation of community‐based interventions.	Very lowb,c,e ⨁◯◯◯
Risk of importation or exportation	3 modelling studies	Two studies reported reductions in the risk of importing and/or exporting cases as a result of travel restrictions; however, no effect estimates were available. The other study reported mixed effects, including an increased risk of importation at some airports, but decreased risk at other airports as a result of lessening travel restrictions. One study suggested that connectedness to the international travel network and the level of community transmission might explain that variation in the effect direction.	Very lowc,f,g ⨁◯◯◯
Outcome category: shift in epidemic development
Probability of eliminating the epidemic	1 modelling study	The study reported mixed effects: the probability would be higher (66% probability) for border restrictions followed by strict community measures than for a delayed border closure (55% probability), and the same as early implementation of border restrictions (66% probability).	Very low h,i,j ⨁◯◯◯
Effective reproduction number	2 modelling studies	One study reported a beneficial change (i.e. break point) in Rt after the implementation of travel restrictions in European Union countries (mean duration 12.6 days). The other study reported mixed effects, suggesting that complete border closures would lead to a 0.045 reduction in Rt, partial relaxation through the opening of land borders would lead to a 0.177 increase in Rt, while further relaxation allowing for international travel followed by quarantine upon arrival would not lead to a change in Rt.	Very low c,e,i ⨁◯◯◯
Time to outbreak	6 modelling studies	Four out of six studies reported reductions in the time to outbreak. These positive effects ranged from a delay of less than one day to 85 days. Two studies reported mixed effects, suggesting both positive effects and no effect. The variation in the direction and magnitude of effect across studies might be explained by differences in the levels of community transmission, the timing of implementation, and the countries restricted by the measure.	Very lowb,c,d ⨁◯◯◯
Risk of outbreak	2 modelling studies	One study reported reductions in the risk of an outbreak resulting from travel restrictions with effects ranging from a 1% to a 37% reduction. The other study reported mixed effects, including both a positive effect and no effect. The variation in the magnitude and direction of effect might be explained by differences in the levels of community transmission, the number of cases in the country of departure, the severity of the travel restriction, co‐interventions, and the percentage of contacts being traced.	Very low c,i,j ⨁◯◯◯
Number or proportion of cases at peak	2 modelling studies	Both studies reported reductions in the number or proportion of cases at peak. These positive effects ranged from a 0.3% reduction to a 8% reduction. The variation in the magnitude of effect might be explained by differences in the implementation of community‐based interventions.	Lowk,l ⨁⨁◯◯
Epidemic growth acceleration	1 modelling study	The study reported that international travel controls would lead to a decrease in the growth acceleration of the epidemic progression across 62 countries (−6.05% change, P < 0.0001).	Low h,m ⨁⨁◯◯
Exportation growth rate	1 modelling study	The study reported that both the lockdown of Hubei, resulting in a ban of all travel, as well as travel restrictions imposed on China led to a decrease in the growth rate of cases exported from Hubei and the rest of China, to the rest of the world.	Low h,m ⨁⨁◯◯
Outcome category: cases detected due to the measure
No contributing study

^aDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the model’s structural elements, the input parameters, and the adequacy of assessment of the model’s uncertainty.
^bDowngraded ‐1 for imprecision, due to a wide range of plausible effects.
^cDowngraded ‐1 for indirectness, due to no reporting of external validation in some studies and/or concerns with reporting of external validation in others.
^dDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the model’s structural elements, the input parameters, the adequacy of assessment of the model’s uncertainty, and incomplete technical documentation.
^eDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the model’s structural elements and the adequacy of assessment of the model’s uncertainty.
^fDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the model’s structural elements, the adequacy of assessment of the model’s uncertainty and the lack of technical documentation.
^gDowngraded ‐1 for imprecision, due to effect estimates being unavailable.
^hDowngraded ‐1 for imprecision, due to only one contributing study.
ⁱDowngraded ‐1 for imprecision, due to insufficient data reported to enable assessment of precision.
^jDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the model's structural elements and input parameters.
^kDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the appropriateness of the models's structural elements.
^lDowngraded ‐1 for indirectness, due to no reporting of external validation in all included studies.
^mDowngraded ‐1 for risk of bias, due to major quality concerns in some studies related to the adequacy of assessment of the model’s uncertainty.