Study ID	Domain	Signalling question	Rating
Symptom screening
Al‐Qahtani 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	2714 individuals arriving to Bahrain airport underwent PCR testing and quarantine procedures
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The study reports on individuals travelling to Bahrain between 25 February and 14 March 2020. As it is not a repatriation study, the population is close to real world travellers. However, the transferability of the travelling population in the early phase of the pandemic (those returning at times of border closure initiation in numerous countries) could be regarded as different than regular travel
		Is there concern that the included participants do not match the review question?	Low
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Unclear
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	The measure of only testing those individuals with a positive PCR test is a hypothetical intervention. The individuals received a symptom based screening and where classified as symptomatic or non‐symptomatic. It is likely, that the status of "symptomatic" was defined prior to conducting the PCR test, but this is not clear.
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The symptom/exposure‐based screening was conducted on regular travellers travelling to Bahrain; It can be therefore assumed to be ‐ while resource intensive ‐ a regular intervention. However the travel volume is lower than would likely be expected under real world scenarios in most situations without additional travel restrictions / border closures
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	As those with a negative initial PCR test underwent quarantine and retesting prior to release from quarantine, the status of the individuals was known to the testers. But as all individuals received the test and PCR testing can be considered relatively objective, this is regarded as low risk of bias. The risk of status change is discussed in domain 4.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Reference test was Observation during 14 day quarantine‐period with retesting in the case of symptoms. and second PCR test for all asymptomatic passengers with an initial negative first PCR test and no symptoms during quarantine. While there is a risk of two negative PCR tests in a row without the development of symptoms, this risk is low.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through  how the quarantine was handled?)	Unclear
		Comments on risk of bias	All Individuals received the PCR test at the same point of time for each repatriation flight and by the same group of experts. There could be a low risk of between flights contamination; e.g. due to differences in taking the swab between experts. The risk of infection during the quarantine period is unclear, this is not well described. There is a risk of individuals having a first false negative PCR test. During the 14‐day quarantine period, those individuals cure out the infection. At the second PCR test they are correctly tested negative. There is a low risk of the time delay between the index test and the reference test leads to missing infected individuals within the population of those with two negative PCR tests.
		4. Could the passenger flow have introduced bias?	Unclear
Arima 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Unclear
		Comments	All 566 passengers repatriated were screened and followed‐up over the 14‐day quarantine period. 3 symptomatic passengers were denied boarding. Their status is unclear. This could lead to an underestimation of the effectiveness of symptom/exposure‐based screening
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers which would travel between China and Japan
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation study, real‐world generalisability should be considered. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening. "Day 1 entry screening by testing oropharyngeal swab samples collected from all 566 returnees at the hospitals to which they were initially transported for SARS‐CoV‐2 (4); all tests were based on the real‐time reverse transcription PCR developed by the National Institute of Infectious Diseases"
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	Reference test was given at both entry and 14 days later at quarantine exit; there is a slight chance given the incubation and infectiousness periods that some cases were not captured, however this risk of rather low.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Exit screening for quarantined persons who remained illness‐free (i.e. not already identified as a case) by collecting oropharyngeal swab samples on day 14, the end of the quarantine period.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All passengers received the reference test at least once at entry and once at quarantine exit
		4. Could the passenger flow have introduced bias?	Unclear
Chen J 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on the flight were included in the study. The crew was not included in the analysis.
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of travelers where an assumed outbreak on the airplane had occurred
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	Symptoms of relevance and definition for classification as "symptomatic" are provided. Screening was conducted at the airport quarantine inspection. This is a hypothetical intervention; all individuals were tested with a PCR upon arrival; and those being symptomatic were reported. It cannot be ruled out that other approaches would have been conducted; if only those with symptoms were tested
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The screening procedure was conducted based on several passengers returning from Wuhan and 2 people being found to be symptomatic. Likely, this led to higher vigilance than would take place in a normal airport screening
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	All participants received two PCR tests + symptom observation with PCR testing as reference test. it is likely that the result of this testing regime captures most infected individuals. Participants were diagnosed using PCR test kits recommended by the Chinese CDC in accordance with protocols established by the WHO. The symptom observation during quarantine was conducted by temperature measurement twice daily and report the development of symptoms.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	All participants received two PCR tests at day 0 and day 13 and a symptom observation with temperature measurement twice daily with PCR testing upon symptom development was used as reference test.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	Other than children with families, travellers were quarantined in single rooms in the hotel. There are limited descriptions on the quarantine regulation (e.g. on whether they were allowed to move freely). 3 individuals were identified at day 7 and day 9 after arrival. Two infected individuals were a couple with departure in Wuhan and belonging to the same tour group as most of the infected, including the likely two index participants. One individual with a symptom development at day 9 after arrival had a departure in Wuhan and did not belong to the tour group as most of the infected. It cannot be ruled out that those tested positive had developed symptoms during the quarantine process without additional information.
		4. Could the passenger flow have introduced bias?	Unclear
Hoehl 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All 126 passengers on board the aircraft from Wuhan received the index text. Likely low risk of selection bias
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers which would travel between China and Germany
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	The study does not provide information on the first screening in China and does not explicitly state that the measures were predefined. But this seems to be likely.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The two step screening procedure was not a regular airport screening, but rather an intervention which is similar to a symptom‐ and exposure based screening that could be conducted in the same way. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening. "Screening for symptoms and clinical signs of infection was performed before their departure from China" "During the flight, 10 passengers were isolated. “These 10 passengers were transferred to University Hospital Frankfurt immediately after arrival." "Two passengers had had contact with 1 person who had a confirmed case of SARS‐CoV‐2 infection, 6 had reported symptoms, were deemed to be clinically symptomatic, or both, and 2 passengers had accompanied family members who had been isolated on the flight because of suspected SARS‐CoV‐2 infection or because of other symptoms (i.e., symptoms related to pregnancy)." "The remaining 116 passengers [...] were sent to the medical assessment center at Frankfurt Airport, where each was evaluated by a medical team of physicians. Each passenger was asked to report current symptoms of fever, fatigue, sore throat, cough, runny nose, muscle aches, and diarrhea, and each one was screened for signs of infection in the nose and throat. The temperature of all passengers was taken."
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	No
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	1. Likelihood of having missed a positive finding in reference test: Unclear. ‐ Participants were isolated for 14 days; there should have been a chance of around 70% to show symptoms if a passenger was infected. ‐ All but 1 passenger received an RT‐PCR test. However, limited information is provided to judge if there is a risk of bias due to (a) inadequate taking of the sample (likely low, if conducted by medical professional) and (b) inadequate transportation, storage, or delay in testing (likely low, but unclear). ‐ It is unclear, (a) at what point of time during the quarantine the test was conducted, (b) if it was repeated (2 tests per person), and (c) if the repetition of the two tests (if there were 2) had a sufficient delay. 2. likelihood of having misclassified a positive case in reference test: low, cases received three tests (at least 2 PCR tests and 1 cell culture test)
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Unclear
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	All (n = 11) passengers with positive screening results received a PCR; For those (n = 116) remaining passengers in quarantine, 115 received the test (1 person declined; no reasons provided); Two passengers had a positive test result from RT‐PCR which was later confirmed by a second test. "RT‐PCR (cycle threshold value in the two samples, 24.39 and 30.25, respectively). Testing with a second protocol consisting of two commercial sets (LightMix Modular SARS and Wuhan CoV E‐gene, and LightMix Modular Wuhan CoV RdRP‐gene, both produced by TIB MOLBIOL) and retesting of the positive samples at the Institute of Virology, Philipps University Marburg, in Marburg, Germany, confirmed the results"
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Unclear
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Unclear
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All but one passenger received the standard test (1/126). The risk of bias resulting from this is considered low, but the reasons for rejecting the reference test are unclear. Limited information on the quarantine procedure is reported
		4. Could the passenger flow have introduced bias?	Unclear
Kim 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on the flight were included; there is no indication that there was a preselection of individuals
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of repatriated individuals in the early phase of the pandemic and does not represent regular air travel passengers,
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	Individuals were identified based on a fixed procedure regarding symptom identification; if action based on the existing criteria (e.g. hospitalisation of those who were screened positive)
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	As it was a relatively small sample, it is unclear if symptom screening with the same intensity could be conducted under real world circumstances of regular air travel. "...the pre‐filled Health Status Questionnaires (HSQs) and passengers’ body temperature were taken. Passengers who reported respiratory symptoms in the previous 14 days on the HSQ or who had a temperature of ≥ 37.5°C were immediately assessed for COVID‐19 by the quarantine doctor. Passengers had a blue or a red sticker placed on their chest, depending on whether they were asymptomatic or under investigation, respectively, so that their COVID‐19 status could be easily identified. [...] [Arrival screening:] The waiting quarantine officers checked the pre‐filled HSQs, measured passengers’ body temperatures, and asked them about symptoms they experienced during the flight. Based on this screening, 1 additional passenger was categorized as a PUI on arrival."
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The reference test consisted of 1 PCR test upon arrival and symptom observation. PCR tests have a risk of producing false negative results, in particular in the early days of infection. There is a risk of the testing procedure missing individuals who got infected around the days of repatriation and did not develop symptoms, and/or had long incubation periods
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Unclear
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	PCR tests are considered the gold standard for identifying SARS‐CoV‐2 Infections
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	The quarantine procedure is not described in detail; the risk of infections during the quarantine would be classified as unclear. But as there were no additional cases identified afterwards; it is regarded as low.
		4. Could the passenger flow have introduced bias?	Low
Lytras 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Unclear
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on board were screened; Out of all repatriation flights, it is unclear why these 7 selected flights were chosen.
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers which would travel between Spain, UK, Turkey and Greece
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Unclear
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	Participants reported symptoms and received some form of medical examination. It is unclear how the screening was handled. A swab was taken at the same time, therefore prior to knowing about the results of the reference test
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The symptom/exposure‐based screening and how individual cases were handled is not well described. We do not know much about the screening procedure (index test). Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening; "All passengers consented to screening, and were asked in‐flight to fill in a paper form with demographic, clinical and contact information. A temporary facility was set up by NPHO at a gate in Athens airport ‘Eleftherios Venizelos’, and swab samples were obtained from passengers immediately upon arrival; those not in need of medical care were subsequently requested to self‐quarantine at home for 14 days."
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	No
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	One‐Time PCR test upon arrival. No additional reference tests are reported, even among those 36 who developed symptoms in the course of the quarantine. Limited information on how the swab was taken is reported.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	High
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	All passengers received the screening test and reference test upon arrival. Only one reference test PCR was conducted, leading to the risk of false negatives. An additional 36 passengers in quarantine with negative tests had developed symptoms in the quarantine. It is possible that a proportion of them was suffering from COVID‐19.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All passengers received the reference test, there is no indication of a different treatment or of attrition bias
		4. Could the passenger flow have introduced bias?	Unclear
Ng 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	no
		Comments	Here, 3 febrile passengers were removed from the pool and we do not have information on their diagnosis; thus removing infected passengers at this point would lead to an underestimate of the screening
		1. Could the selection of participants have introduced bias?	High
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	Repatriation‐study, transferability to the real world as a screening intervention is unclear. "We followed up on 94 persons who boarded an evacuation flight from Wuhan to Singapore on January 30, 2020."
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	On arrival in Singapore, the passengers underwent repeat screening for body temperature (fever was defined as a body temperature ≥38°C), and 2 persons had a fever. Co‐Intervention: 1. Airport Departure fever screening: Screening for body temperature was conducted at check‐in and before boarding, and 3 febrile persons were prevented from boarding (no additional information regarding the status of these 3 febrile persons was available) 2. Masks Surgical masks were provided to passengers on board the plane.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation‐study, transferability to the real world as a screening intervention is unclear. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening "The 2 febrile women identified in arrival screening were transferred immediately to a hospital, and they tested positive for SARS‐CoV‐2 (their clinical course is described in the Supplementary Appendix)"
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	No
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	no
		Comments on risk of bias	While the reference test is not perfect, most passengers received at least 2 PCR tests with 3 days in between, were observed intensely for 14 days, and tested symptoms developed.. There is a risk of participants being missed by this procedure, but likely this is low.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	1. "Quarantine and observation for 14 days and checking of symptoms and fever three times daily 2. PCR assessment of those showing symptoms; those with symptoms received only 1 test 3. On quarantine day 3, samples from 76 of the 86 asymptomatic persons (75 nasopharyngeal swab samples and 1 nasal swab sample) were obtained and tested by means of PCR assay. 4. On quarantine day 6, samples from all 87 quarantined asymptomatic persons (85 nasopharyngeal swab samples and 2 nasal swab samples [3 of the 6 persons who had been transferred to the hospital before February 2 had returned to the government quarantine facility]) were obtained and tested; all tested negative."
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias
		4. Could the passenger flow have introduced bias?	Unclear
Wong J 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Unclear
		Comments	The publication provides limited information on the population arriving. The population in question comprises individuals arriving between 21.03.2020 and 24.04.2020. It can be assumed, that all passengers underwent the same procedure and received the same reference test (RT‐PCR testing) but this is not clear from the publication
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The study reports on individuals travelling to Brunei between 21st of March and 24th of April 2020. As it is not a repatriation study, the population is close to real world travellers. However, the transferability of the travelling population in the early phase of the pandemic (those returning at times of border closure initiation in numerous countries) could be regarded as different from regular travel. The study provides limited information on the population, however.
		Is there concern that the included participants do not match the review question?	Unclear
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Unclear
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	The procedures of performing and interpreting the index test is unclear; the publication provides limited information This is a hypothetical intervention; all individuals were tested with a PCR upon arrival; and those being symptomatic were reported. It cannot be ruled out that other approaches would have been conducted; if only those with symptoms were tested
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The symptom/exposure‐based screening was conducted on a large number of individuals as a program implemented on a regular basis at the airport (i.e. not just once for a repatriation flight); however the travel volume is lower than would likely be expected under real world scenarios in most situations without additional travel restrictions / border closures
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Unclear
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The study population received at least one PCR test upon arrival and underwent a 14 day quarantine procedure. There is limited information provided (e.g. on whether there was a second PCR‐test prior to release or how the symptom‐observation during the quarantine procedure was conducted. It is therefore unclear whether the reference test can be considered as sufficient
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Unclear
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	The reference test for the symptom screening consisted of at least one PCR test close to the time point of arrival. As PCR testing is considered the gold standard with an assumed very high specificity, the definition of case and not case regarding SARS‐CoV‐2 infections can be assumed as given
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Unclear
		4.2 Did all passengers receive the same reference standard?	Unclear
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Unclear
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	The publication provides limited information on the population arriving. The population in question comprises individuals arriving between 21.03.2020 and 24.04.2020. It can be assumed, that all passengers underwent the same procedure and received the same reference test (RT‐PCR testing) but this is not clear from the publication. It is unclear, whether the 30 individuals who tested positive were identified by the PCR test upon arrival, or by another procedure.
		4. Could the passenger flow have introduced bias?	Unclear
Yamahata 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers and crew members received the index test
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	Not a real world screening intervention. This is a situation which can be seen as analogous to a screening intervention. "On February 3, 2020, when the ship arrived at Yokohama, a quarantine was initiated. All passengers and crew underwent medical examinations. On February 5, the RT‐PCR results from the throat swab for symptomatic people and their close contacts revealed that 10 of 31 individuals were positive for SARS‐CoV‐2. On the same day, the Japanese government decided that all passengers were to be quarantined in their cabins for 14 days [10]. Based on international guidance on infection control, the crew continued to maintain ship functions and support passengers for their food, clothing, and shelter‐related needs."
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Unclear
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	The study provides very limited information on the screening measure.
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Not a real world screening intervention. It is a situation which can be seen as analogous to a screening intervention Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening "On February 3, 2020, when the ship arrived at Yokohama, a quarantine was initiated. All passengers and crew underwent medical examinations. On February 5, the RT‐PCR results from the throat swab for symptomatic people and their close contacts revealed that 10 of 31 individuals were positive for SARS‐CoV‐2. On the same day, the Japanese government decided that all passengers were to be quarantined in their cabins for 14 days [10]. Based on international guidance on infection control, the crew continued to maintain ship functions and support passengers for their food, clothing, and shelter‐related needs."
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	No
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	All passengers received a PCR test. All passengers and crew were quarantined. Those developing symptoms received another PCR test. There is a risk of cases being missed if they had no or minor symptoms during the quarantine (about 30%) of infected
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	High
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	"During the quarantine, RT‐PCR testing of throat swabs was extended to all passengers in the following order: On February 3, 2020, when the ship arrived at Yokohama, a quarantine was initiated. All passengers and crew underwent medical examinations. On February 5, the RT‐PCR results from the throat swab for symptomatic people and their close contacts revealed that 10 of 31 individuals were positive for SARS‐CoV‐2. On the same day, the Japanese government decided that all passengers were to be quarantined in their cabins for 14 days. Based on international guidance on infection control, the crew continued to maintain ship functions and support passengers for their food, clothing, and shelter‐related needs."
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	No
		Comments on risk of bias	Keeping passengers for 3 weeks on the ship could have led to the development of additional cases (close contact among infected). Thus using quarantine as part of the reference test would yield an underestimation of the effectiveness of the screening measure
		4. Could the passenger flow have introduced bias?	High
PCR test
Arima 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Unclear
		Comments	All 566 passengers repatriated were screened and followed‐up over the 14‐day quarantine period. 3 symptomatic passengers were denied boarding. Their status is unclear. This could lead to an underestimation of the effectiveness of symptom/exposure‐based screening
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers which would travel between China and Japan
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation study, real‐world generalisability should be considered. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening. "Day 1 entry screening by testing oropharyngeal swab samples collected from all 566 returnees at the hospitals to which they were initially transported for SARS‐CoV‐2 (4); all tests were based on the real‐time reverse transcription PCR developed by the National Institute of Infectious Diseases"
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	Reference test was given at both entry and 14 days later at quarantine exit; there is a slight chance given the incubation and infectiousness periods that some cases were not captured, however this risk is rather low
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Exit screening for quarantined persons who remained illness‐free (i.e. not already identified as a case) by collecting oropharyngeal swab samples on day 14, the end of the quarantine period.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Unclear
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	Three out of the 12 cases were missed in the screening and testing procedure. Of those, two developed symptoms, one was in facility and one in home quarantine. The risk of the individual to have acquired the infection from other individuals in the facility is regarded as low as it is reported "facility‐quarantined case‐patient was in a single room; no other person from this facility acquired COVID‐19 or had a positive test result at exit screening." The publication does not provide enough information to judge if there is a risk of the other two individuals could have acquired the infection after arrival in the quarantine:
		4. Could the passenger flow have introduced bias?	Unclear
Al‐Qahtani 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	2714 individuals arriving at the Bahrain airport underwent PCR testing and quarantine procedures
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The study reports on individuals travelling to Bahrain between 25th of February and 14th of March 2020. As it is not a repatriation study, the population is close to real world travellers. However, the transferability of the travelling population in the early phase of the pandemic (those returning at times of border closure initiation in numerous countries) could be regarded as different than regular travel
		Is there concern that the included participants do not match the review question?	Low
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	PCR is the only means of identifying individuals; there is no indication of additional procedures (e.g. double testing for certain individuals / case definition based on reporting of individual). Due to the PCR test; the predefinition of the PCR threshold definition is implied within the use of the method
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The PCR testing was conducted as part of a regular measure which was conducted on all travellers; however the travel volume is lower than would likely be expected under real world scenarios in most situations without additional travel restrictions / border closures
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	As those with a negative initial PCR test underwent quarantine and retesting prior to release from quarantine, the status of the individuals was known to the testers. But as all individuals received the test and PCR testing can be considered relatively objective, this is regarded as low risk of bias. The risk of status change is discussed in domain 4
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Reference test was Observation during 14 day quarantine‐period with retesting in the case of symptoms and second PCR‐based test for all asymptomatic passengers with an initial negative first PCR‐test and no symptoms during quarantine. While there is a risk of two negative PCR tests in a row without the development of symptoms, this risk is low.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All Individuals received the PCR test at the same point of time for each repatriation flight and by the same group of experts. There could be a low risk of between flights contamination; e.g. due to differences in taking the swab between experts. The risk of infection during the quarantine period is unclear, this is not well described. There is a risk of individuals having a first false negative PCR test. During the 14 day quarantine period, those individuals recover. At the second PCR test they are correctly tested negative. There is a low risk of the time delay between the index test and the reference test leads to missing infected individuals within the population of those with two negative PCR tests.
		4. Could the passenger flow have introduced bias?	Unclear
Al‐Tawfiq 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Unclear
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	1928 individuals working for a large oil company of the Kingdom of Saudi Arabia and returning in the early phase of the pandemic. Likely, all individuals working for the company underwent the procedure. A number of travellers arrived through Bahrain. The study reports that those travellers arriving through Bahrain underwent a 2 weeks quarantine procedure before coming to Saudi Arabia. There is a risk that individuals showing symptoms may have undergone different procedures e.g. individuals with respiratory symptoms not being allowed to travel to Saudi Arabia. This could be compared to a symptom/exposure‐based departure screening upon departure, with an unclear number regarding those retained.
		1. Could the selection of participants have introduced bias?	High
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is characterized by their status as employees of a large oil company of Saudi Arabia. This leads to a distortion which makes the population different from regular air travel passengers (e.g. 66% males).
		Is there concern that the included participants do not match the review question?	Unclear
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	All individuals received a nasopharyngeal and oropharyngeal swab PCR test within 24h after arrival. The test was taken by trained health professionals. A bias due to the process of performing the index test is unlikely
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The symptom screening was conducted on a large number of individuals as a program implemented on a regular basis at the airport (i.e. not just once for a repatriation flight); however the travel volume is lower than would likely be expected under real world scenarios in most situations without additional travel restrictions / border closures
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The reference test was a PCR based test on day 13, as well as symptom‐observation during a 14 day quarantine period. As a case is defined by having a positive PCR test, it can be assumed very likely, that all individuals who tested negative were truly not infected with SARS‐CoV‐2 Due to the combination of symptom observation for 14 days and PCR‐based testing prior to release, it can be assumed that there is a low risk of bias due to the reference test falsely classifying an individual as negative.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	As PCR testing is considered the gold standard with an assumed very high specificity, the definition of case and not case regarding SARS‐CoV‐2 infections can be assumed as given
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	No
		Comments on risk of bias	The study writes: "...there were strict quarantine protocols. Guests not being allowed out of the rooms except in case of emergency". There is no indication of an outbreak in the facility.
		4. Could the passenger flow have introduced bias?	Low
Lagier 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Unclear
		Comments	The study reports that the quarantine procedure began in China 2‐7 days prior to travelling back to France. Limited information is provided here. It is unclear, if individuals who showed symptoms upon entry into quarantine or developed symptoms during the quarantine phase were handled (e.g. did they receive a PCR test and were denied the boarding the flight?)
		1. Could the selection of participants have introduced bias?	High
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of repatriated individuals in the early phase of the pandemic and does not represent regular air travel passengers,
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	All repatriated individuals were offered a PCR test upon arrival (within 24h). As the PCR tests are established and sufficiently defined procedures, the risk of bias due to the approach of conducting the index test is unlikely. However, the study reports that the quarantine procedure began in China 2‐7 days prior to travelling back to France. Limited information is provided here. Therefore, the PCR‐based screening was conducted between day 3 and 8 of quarantine procedure in China with a flight in between. It is unclear, whether this could lead to a bias regarding the ability of the PCR test to detect cases. The symptom/exposure‐based screening was conducted by professional individuals at a military air force base. The procedure is not described in detail. It is likely, that a threshold for symptom‐status was established (although not described in detail), as this event triggered action regarding the symptomatic individual
		2. Could the conduct or interpretation of the index test have introduced bias?	Unclear
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Index test is that all individuals after a period in a quarantine facility in China of 2‐7 days and within 24h after arrival from China ‐ independent of symptom status ‐ receive a PCR test. This procedure is likely very similar to such a measure being conducted in real world airport settings
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Low
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The reference test consisted of a second PCR test (day 5 after arrival; day 7‐12 after quarantine in China) as well as symptom/exposure‐based observation during a 5‐day (unclear) quarantine with a follow‐up PCR test in the case of symptoms. While this procedure is likely sufficient assuming the 2‐7 day quarantine procedure prior to the flight (including not described local measures), this is likely sufficient to detect cases. However, assuming that the pre‐quarantine procedure did not taken place (this was already covered by the risk of bias regarding the population selection) the ability to detect cases is unclear.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	The reference test PCR testing upon arrival consisted of a 1 PCR tests as well as symptom observation during a 5 day quarantine period. As PCR‐testing is considered the gold standard with an assumed very high specificity, the definition of case and not case regarding SARS‐CoV‐2 infections can be assumed as given.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	No
		4.2 Did all passengers receive the same reference standard?	No
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Unclear
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	No
		Comments on risk of bias	7/337 individuals were tested only once and 1/337 individual refused both tests. It is unclear from the study reporting, if those who refused were symptomatic upon arrival or developed symptoms during the course of the observation period (leading to a higher pre‐test probability). Without additional information on the individuals who refused/were not tested; the risk of bias is unclear. Among those who were tested twice: While the quarantine measures are not described in detail, as there were in total 0 cases detected, it is unlikely that the disease status of individuals had changed due to the quarantine procedures. Due to the two PCR tests (day 1 and day 5), it can be assumed that all infections within this period (even those in decline) would have been covered.
		4. Could the passenger flow have introduced bias?	Unclear
Lio 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Unclear
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on the flight were included in the study. However, one passenger with symptoms was denied boarding. The status regarding SARS‐CoV‐2 infection of this passenger is not reported
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of repatriated individuals in the early phase of the pandemic and does not represent regular air travel passengers,
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	The individuals underwent a 2 day quarantine with symptom observation, followed by a PCR test on day 2. As the PCR tests are established and sufficiently defined procedures, the risk of bias due to the approach of conducting the index test is unlikely
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Index test is that all individuals after a period in a quarantine facility of 2 day ‐ independent of symptom status ‐ receive a PCR test. This procedure is likely very similar to such a measure being conducted in real world airport settings
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Low
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The reference test for the symptom based screening consisted of a 2 PCR tests at day 7 and 13 as well as symptom‐observation during a 14 day quarantine phase (11‐12 days after first PCR‐test). It can be assumed very likely, that all individuals who tested negative were truly not infected with SARS‐CoV‐2
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	The reference test for the symptom screening consisted of a 3 PCR tests at day 2, 7 and 13 as well as symptom observation during a 14 day quarantine period. As PCR testing is considered the gold standard with an assumed very high specificity, the definition of case and not case regarding SARS‐CoV‐2 infections can be assumed as given
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	No
		Comments on risk of bias	While the quarantine measures are not described in detail, as there were 0 cases detected, it is unlikely that the disease status of individuals changed due to the quarantine procedures. Due to the two PCR tests (day 2 and day 7) covering the period between arrival and the prerelease PCR‐test at day 14, it can be assumed that all infections within this period (even those in decline) would have been covered.
		4. Could the passenger flow have introduced bias?	Low
Ng 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	no
		Comments	Here, 3 febrile passengers were removed from the pool and we do not have information on their diagnosis; thus removing infected passengers at this point would lead to an underestimate of the screening
		1. Could the selection of participants have introduced bias?	High
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	Repatriation‐study, transferability to the real world as a screening intervention is unclear. "We followed up on 94 persons who boarded an evacuation flight from Wuhan to Singapore on January 30, 2020."
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	On arrival in Singapore, the passengers underwent repeat screening for body temperature (fever was defined as a body temperature ≥38°C), and 2 persons had a fever. Co‐Intervention: 1. Airport Departure fever screening: Screening for body temperature was conducted at check‐in and before boarding, and 3 febrile persons were prevented from boarding (no additional information regarding the status of these 3 febrile persons was available) 2. Masks Surgical masks were provided to passengers on board the plane.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation‐study, transferability to the real world as a screening intervention is unclear. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening. "The 2 febrile women identified in arrival screening were transferred immediately to a hospital, and they tested positive for SARS‐CoV‐2 (their clinical course is described in the Supplementary Appendix)"
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	No
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	no
		Comments on risk of bias	While the reference test is not perfect, most passengers received at least 2 PCR tests with 3 days in between, were observed intensely for 14 days, and tested in the case of symptoms. There is a risk of participants being missed by this procedure, but likely this number is low.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	High
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	1. Quarantine and observation for 14 days and checking of symptoms and fever three times daily 2. PCR assessment of those showing symptoms; those with symptoms received only 1 test 3. On quarantine day 3, samples from 76 of the 86 asymptomatic persons (75 nasopharyngeal swab samples and 1 nasal swab sample) were obtained and tested by means of PCR assay 4. On quarantine day 6, samples from all 87 quarantined asymptomatic persons (85 nasopharyngeal swab samples and 2 nasal swab samples [3 of the 6 persons who had been transferred to the hospital before February 2 had returned to the government quarantine facility]) were obtained and tested; all tested negative.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias
		4. Could the passenger flow have introduced bias?	Unclear
Shaikh Abdul Karim 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	Repatriation study of 432 individuals to Malaysia. All individuals underwent the two‐step screening procedures
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers. With a focus on the proportion of cases detected/ PPV this is likely not a source of bias; although the generalisability may be limited. Strict procedures to limit risk of infection during the flight may give higher protection against infection on the flight. The lack of these very early infections might lead to an over‐estimation of sensitivity (which is very low for very early stages of infection); However it is unclear, whether this can be considered different from no protective measures.
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	PCR is the only means of identifying individuals; there is no indication of additional procedures to identify individuals and cases (e.g. double testing for certain individuals / case definition based on reporting of individual). Due to the PCR test; the predefinition of the PCR threshold definition is implied within the use of the method.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation study, real‐world generalisability should be considered. Day 1 entry screening for symptoms was conducted by a large and specialized team; it is questionable, whether this would be possible for real world symptom/exposure‐based screening for regular travellers. If the focus is the general PCR testing of all individuals, there should be limited concerns regarding the generalisability of the approach to the index test. "The reception team personnel consist of officials from NADMA, personnel from the Fire and Rescue Department as Ground Mission Commander and medical personnel from the MOH. Medical personnel from MOH comprise an emergency physician, a public health physician, nurses, assistant medical officers, a pathologist and laboratory technicians. The team utilises the AirDisaster Unit (ADU) as the base of the operations." Strict procedures to limit risk of infection during the flight may give higher protection against infection on the flight. The lack of these very early infections might lead to an over‐estimation of sensitivity (which is very low for very early stages of infection); However it is unclear, whether this can be considered different from no protective measures.
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	Reference test was observation during 14 day quarantine‐period and second PCR‐based test for all asymptomatic passengers with an initial negative first PCR‐test. While there is a risk of two negative PCR tests in a row without the development of symptoms, this risk is low.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Reference test was Observation during 14 day quarantine‐period and second PCR‐based test for all asymptomatic passengers with an initial negative first PCR‐test. While there is a risk of two negative PCR tests in a row without the development of symptoms, this risk is low.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All Individuals received the PCR test at the same point of time for each repatriation flight and by the same group of experts. There could be a low risk of between flights contamination; e.g. due to differences in taking the swab between experts. The risk of infection during the quarantine period is unclear, this is not well described. There is a risk of individuals having a first false negative PCR test. During the 14 day quarantine period, those individuals cure out the infection. At the second PCR test they are correctly tested negative. There is a low risk of the time delay between the index test and the reference test leads to missing infected individuals within the population of those with two negative PCR tests.
		4. Could the passenger flow have introduced bias?	Unclear
Quarantine of travellers and screening combined
Al‐Qahtani 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	2714 individuals arriving at the Bahrain airport underwent PCR testing and quarantine procedures
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The study reports on individuals travelling to Bahrain between 25 February and 14 March 2020. As it is not a repatriation study, the population is close to real world travellers. However, the transferability of the travelling population in the early phase of the pandemic (those returning at times of border closure initiation in numerous countries) could be regarded as different than regular travel
		Is there concern that the included participants do not match the review question?	Low
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	Here, the index test is the symptom observation during quarantine with PCR testing on top of 1 PCR test on the day of arrival. Insufficient measures in the symptom observation and testing is likely to reduce the assumed sensitivity of the quarantine measure. 188 individuals were tested positive, 136 of which (72.3%) within arrival screening (of those 44 were symptomatic and tested positive and 92 were asymptomatic but identified through the PCR test). 27 of the remaining cases were detected through symptom‐development and follow‐up PCR testing. The exact procedures regarding the testing in the quarantine facility is not well described, but it is reported that systematic testing for cough, fever, and sore throat was conducted to select individuals for additional PCR testing
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The index test refers to highly selective circumstances: Specialized quarantine facility with enough space for all individuals. While the transferability of these findings to similar circumstances is likely given, it is unclear, whether the results are transferable to e.g. home quarantine
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Low
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	Reference test was one additional PCR test conducted at day 14 after arrival. The risk for false negative PCR test is highest in the first days after infection with a decline in the course of the infection, likely in line with the infectiousness. It is unclear how the sensitivity and specificity of a single PCR test after 14 days of quarantine and therefore for individuals with a very long incubation period should be regarded
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Unclear
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Reference test was Observation during 14 day quarantine‐period with retesting in the case of symptoms and second PCR‐based test for all asymptomatic passengers with an initial negative first PCR‐test and no symptoms during quarantine. While there is a risk of two negative PCR tests in a row without the development of symptoms, this risk is low.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All Individuals received the PCR test at the same point of time for each repatriation flight and by the same group of experts. There could be a low risk of between flights contamination; e.g. due to differences in taking the swab between experts. The risk of infection during the quarantine period is unclear, this is not well described. As there were infections among those in quarantine, there would be a risk of infections taking place in the quarantine facility. Due to the limited information, this cannot be ruled out. There is no indication of an outbreak/infections among staff in the quarantine facility Those who were tested positive were sent to specialized quarantine facilities, it is therefore unlikely that they caused infections after isolation. There is a risk of individuals having a first false negative PCR test. During the 14 day quarantine period, those individuals cure out the infection. At the second PCR test they are correctly tested negative. There is a low risk of the time delay between the index test and the reference test leads to missing infected individuals within the population of those with two negative PCR tests.
		4. Could the passenger flow have introduced bias?	Unclear
Arima 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Unclear
		Comments	All 566 passengers repatriated were screened and followed‐up over the 14‐day quarantine period. 3 symptomatic passengers were denied boarding. Their status is unclear. This could lead to an underestimation of the effectiveness of symptom/exposure‐based screening
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers which would travel between China and Japan
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation study, real‐world generalisability should be considered. Likely, the awareness of the risk status of the individuals led to led to higher vigilance than would take place in a normal airport screening. "Day 1 entry screening by testing oropharyngeal swab samples collected from all 566 returnees at the hospitals to which they were initially transported for SARS‐CoV‐2 (4); all tests were based on the real‐time reverse transcription PCR developed by the National Institute of Infectious Diseases."
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	Reference test was given at both entry and 14 days later at quarantine exit; there is a slight risk given the incubation and infectiousness periods that some cases were not captured, however this risk of rather low
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Exit screening for quarantined persons who remained illness‐free (i.e. not already identified as a case) by collecting oropharyngeal swab samples on day 14, the end of the quarantine period.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Unclear
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	Three out of the 12 cases were missed in the screening and testing procedure. Of those, two developed symptoms, one was in facility and one in home quarantine. The risk of the individual to have acquired the infection from other individuals in the facility is regarded as low as it is reported "facility‐quarantined case‐patient was in a single room; no other person from this facility acquired COVID‐19 or had a positive test result at exit screening." The publication does not provide enough information to judge if there is a risk of the other two individuals could have acquired the infection after arrival in the quarantine:
		4. Could the passenger flow have introduced bias?	Unclear
Chen J 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on the flight were included in the study. The crew was not included in the analysis.
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of travellers where an assumed outbreak during the flight occurred
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	Index test refers to the combination of general PCR test upon arrival and symptom‐focused quarantine observation with PCR testing upon symptom development. All participants underwent symptom observation with PCR testing upon symptom development for 14 days, with passengers from Wuhan an additional 7 days. All participants received a second (reference) PCR test independent of symptom status at day 13. Participants were diagnosed using PCR test kits recommended by the Chinese CDC in accordance with protocols established by the WHO. The symptom observation during quarantine was conducted by temperature measurement twice daily and report the development of symptoms.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The index test refers to highly selective circumstances: quarantine in single rooms in a hotel with strict symptom observation and reporting of symptoms. While the transferability of these findings to similar circumstances is likely given, it is unclear, whether the results are transferable to e.g. home quarantine. All participants underwent symptom observation with PCR testing upon symptom development for 14 days, with passengers from Wuhan an additional 7 days. All participants received a second (reference) PCR test independent of symptom status at day 13. Participants were diagnosed using PCR test kits recommended by the Chinese CDC in accordance with protocols established by the WHO. The symptom observation during quarantine was conducted by temperature measurement twice daily and report the development of symptoms.
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Low
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	The reference test is a PCR test prior to release (day 13 after arrival). Due to the high specificity of PCR tests, the risk of false positive PCR tests can be regarded as low. While there is some risk of false‐negative results, this probability can be regarded as low due to the very low pre‐test probability (this would require 2 negative PCR tests and no symptom development during the observation period).
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	All participants received two PCR tests at day 0 and day 13 and symptom observation with temperature measurement twice daily with PCR testing upon symptom development was used as reference test.
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	Other than children with families, travellers were quarantined in single rooms in the hotel. There are limited descriptions on the quarantine regulation (e.g. on whether they were allowed to move freely). 3 individuals were identified at day 7 and day 9 after arrival. Two infected individuals were a couple with departure in Wuhan and belonging to the same tour group as most of the infected, including the likely two index participants. One individual with a symptom development at day 9 after arrival had a departure in Wuhan and did not belong to the tour group as most of the infected. It cannot be ruled out that those tested positive had developed symptoms during the quarantine process without additional information
		4. Could the passenger flow have introduced bias?	Unclear
Lio 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	All passengers on the flight were included in the study. However, one passenger with symptoms was denied boarding. the status regarding SARS‐CoV‐2 infection of this passenger is not reported
		1. Could the selection of participants have introduced bias?	Unclear
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample of repatriated individuals in the early phase of the pandemic and does not represent regular air travel passengers,
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	N/A
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	N/A
		2.3 If a threshold was used, was it prespecified?	Unclear
		Comments on risk of bias	The index test (2 delayed PCR tests on day 2 and 7 as well as symptom observation during quarantine) and the reference test (1 PCR tests on day 13) did not identify any infected individuals. The process for symptom‐observation during quarantine is not well described; but no individuals were identified as cases.
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	The screening procedure is not well described. But it was conducted not as part of a regular screening of travellers, but of a small set of repatriated individuals. However, an institutionalised quarantine measure with PCR testing would likely be close to what was conducted in this study. The transferability of this particular measure to other similar approaches needs to be considered
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	Unclear
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	yes
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	Unclear
		Comments on risk of bias	The reference test is a PCR test prior to release. Due to the high specificity of PCR tests, the risk of false positive PCR tests can be regarded as low. Due to the very low pre‐test probability (2 negative PCR tests and no symptom development during the observation period) the risk of a false negative test result can be regarded as low.
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Low
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	The reference test for the symptom screening consisted of a 3 PCR tests at day 2, 7 and 13 as well as symptom observation during a 14 day quarantine phase. As PCR testing is considered the gold standard with an assumed very high specificity, the definition of case and not case regarding SARS‐CoV‐2 infections can be assumed as given
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	No
		Comments on risk of bias	While the quarantine measures are not described in detail, as there were in total 0 cases detected, it is unlikely that the disease status of individuals changed due to the quarantine procedures. Due to the two PCR tests (day 2 and day 7) covering the period between arrival and the prerelease PCR‐test at day 14, it can be assumed that all infections within this period (even those in decline) would have been covered.
		4. Could the passenger flow have introduced bias?	Low
Shaikh Abdul Karim 2020	Domain 1: Patient selection ‐ A. Risk of bias	1.1 Was a consecutive or random sample of participants enrolled?	Yes
		1.2 Was a case‐control design avoided?	Yes
		1.3 Did the study avoid inappropriate exclusions?	Yes
		Comments	Repatriation study of 432 individuals to Malaysia. All individuals underwent the two‐step screening procedures
		1. Could the selection of participants have introduced bias?	Low
	Domain 1: Patient selection ‐ B. Concerns regarding applicability	Describe included participants (prior testing, presentation, intended use of index test and setting):	The population is a highly selected sample in the early phase of the pandemic and does not represent regular air travel passengers. With a focus on the proportion of cases/ PPV this is likely not a source of bias; although the generalisability may be limited. Strict procedures to limit risk of infection during the flight may give higher protection against infection on the flight. The lack of these very early infections might lead to an overestimation of proportion of cases detected (which is very low for very early stages of infection). However it is unclear, whether this can be considered different from no protective measures.
		Is there concern that the included participants do not match the review question?	High
	Domain 2: Index test(s) A. Risk of bias	2.1 Can we be sure that those identified in index test (true and false positive screening results) were identified by the index test (e.g. automated fever scanner) rather than any other means (e.g. self‐reporting)?	Yes
		2.2 Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
		2.3 If a threshold was used, was it prespecified?	Yes
		Comments on risk of bias	Here the index test is the combination of mass PCR testing + follow up quarantine. No symptom/exposure‐based observation was conducted. Therefore, all individuals identified in the index test result from the PCR test. The risk of bias here is therefore equivalent to the assessment for the sensitivity of the mass RT‐PCR testing
		2. Could the conduct or interpretation of the index test have introduced bias?	Low
	Domain 2: Index test(s) B. Concerns regarding applicability	Describe the index test and how it was conducted and interpreted:	Repatriation study, real‐world generalisability should be considered. However, an institutionalised quarantine measure with PCR testing would likely be close to what was conducted in this study. The transferability of this particular measure to other similar approaches needs to be considered. Strict procedures to limit risk of infection during the flight may give higher protection against infection on the flight. The lack of these very early infections might lead to an overestimation of sensitivity (which is very low for very early stages of infection); however it is unclear, whether this can be considered different from no protective measures.
		Is there concern that the index test, its conduct, or interpretation differ from the review question?	High
	Domain 3: Reference standard A. Risk of bias	3.1 Is the reference standard (the approach to identify and classify ‘cases’) likely to correctly classify the target condition (here active infection with SARS‐CoV‐2)?	Unclear
		3.2. Were the reference standard results interpreted without knowledge of the results of the index test?	No
		Comments on risk of bias	Reference test was one additional PCR test conducted at day 14 after arrival. The risk for false negative PCR tests is highest in the first days after infection with a decline in the course of the infection, likely in line with the infectiousness. It is unclear how the sensitivity and specificity of a single PCR test after 14 days of quarantine and therefore for individuals with a very long incubation period should be regarded
		3. Could the reference standard, its conduct, or its interpretation have introduced bias? (Leading to an under‐determination or over‐determination of true findings in the reference test)	Unclear
	Domain 3: Reference standard & timing B. Concerns regarding applicability	Describe the reference standard and how it was conducted and interpreted:	Reference test was one additional PCR test conducted at day 14 after arrival. As a case is defined by having a positive PCR test and the risk for false positive PCR tests is relatively low; the reference test is in line with the review question
		Is there concern that the target condition as defined by the reference standard does not match the review question?	Low
	Domain 4: Flow and timing ‐ Risk of bias	4.1 Did all passengers receive the reference standard?	Yes
		4.2 Did all passengers receive the same reference standard?	Yes
		4.3 Were all passengers included in the analysis? Is there likely no or a very low risk of attrition bias?	Yes
		4.4. Is it possible that the true disease status could have changed between the application of the index test and the reference standard? (e.g. additional infections through how the quarantine was handled?)	Unclear
		Comments on risk of bias	All Individuals received the PCR test at the same point of time for each repatriation flight and by the same group of experts. There could be a low risk of between flights contamination; e.g. due to differences in taking the swab between experts. The risk of infection during the quarantine period is unclear, this is not well described. There is a risk of individuals having a first false negative PCR test. During the 14 day quarantine period, those individuals cure out the infection. At the second PCR test they are correctly tested negative. There is a low risk of the time delay between the index test and the reference test leads to missing infected individuals within the population of those with two negative PCR tests.
		4. Could the passenger flow have introduced bias?	Unclear