Published 3 November 2020
UK Government’s 14-day quarantine policy is least effective at preventing Covid-19 from entering the community, compared to all forms of passenger testing.
New modelling finds a single test on arrival is twice as effective as the 14-day policy with a test three days pre-departure also 44% more effective than current policy.
If a ‘test and release’ regime is applied, three days of quarantine is significantly more effective to minimise infectious days (by 60%) compared to ‘test and release’ after five days (53%) or seven days (45%).
Innovative modelling by Oxera and Edge Health takes account of quarantine non-compliance evidence to analyse the effectiveness of passenger testing versus 14-day quarantine and introduces an improved measurement of risk of community transmission – the number of ‘infectious days screened’.
New research suggests UK Government’s current 14-day quarantine policy for international arrivals, which involves no test, is far more ineffective than previously considered. Factoring in recent evidence of individuals failing to adhere to the full period of quarantine, the current 14-day policy is less effective than all other passenger testing regimes currently under consideration. Just 25% of ‘infectious days’ are prevented through the policy, which has been in effect since July 2020.
New modelling from Edge Health and Oxera applies the latest evidence of quarantine non-compliance and test sensitivity to consider the effectiveness of the UK’s current quarantine policy against a range of passenger testing regimes, applying the metric of ‘infectious days screened’. Using a London School of Hygiene and Tropical Medicine (LSHTM) model as a starting point, the researchers applied recent real-world Covid-19 prevalence rates and air passenger volumes to produce a comprehensive picture of policy options.
When evidence of non-compliance is taken into account, a single test on arrival would reduce infectious days by 51% - more than double the current policy. An RT-PCR test, three days pre-departure reduces the infectious days by 36%. The analysis also finds a ‘test and release’ regime on day three is significantly more effective in minimising infectious days (60% reduction in infectious days with a RT-LAMP test) than the 14-day policy (25%), and also more effective than ‘test and release’ after five days (53%) or seven days (45%) currently under consideration by the UK’s Global Travel Taskforce. Applying the latest assumptions, based on real-world evidence, the three days option balances between ensuring sufficient time for Covid-19 to become detectable and a swift enough result to inform the non-compliant infected individuals in the community of their infectious status, thus ensuring they go into isolation.
The new analysis follows a Public Health England (PHE) paper which underpins the Government’s policymaking, which assumes 100% compliance to 14-day quarantine.
However a recent study, which is consistent with SAGE’s own findings, shows that just 71% of symptomatic individuals may be following the rules, falling to as low as 28% for asymptomatic individuals. Another study finds that compliance may be as low as 18%, even when individuals present Covid-19 symptoms.
By using ‘infectious days screened’ to measure effectiveness, in addition to ‘infectious travellers detected’, the authors were able to form a better picture of community transmission risk. On this basis, on arrival testing outperformed 14-day quarantine policy in a multitude of scenarios. The authors also determined that if passenger testing on arrival had been in place in August 2020, the presence of Covid-19 in arrivals from the EU would have been just 0.01% and from the USA just 0.09%, compared to ONS estimates that the UK community transmission rate was 0.57% during the months of September and October.
George Batchelor, Cofounder and Director of Edge Health, said:
“The current policy of 14-day quarantine is the least effective strategy due to human behaviour, with returning travellers often failing to comply and risking community transmission. Our modelling finds a much more effective path in passenger testing, which if it’s applied as a ‘test and release’ regime, performs the best at reducing infectious days after 3 days quarantine. This balances the detectability of Covid-19 and new evidence of quarantine non-compliance.”
“Our work also modelled the effectiveness of on arrival and pre-departure testing schemes. While on arrival testing is twice as effective as quarantine, there remains a lack of real-world data for pre-departure testing to calibrate the outputs from the modelling. Pre-departure trials at scale are crucial to accumulate data and aid the best policy decision-making to reopen the skies safely.”
Michele Granatstein, Partner at Oxera and Head of its Aviation practice, said:
“Our objective is to assess the effectiveness of different testing schemes and quarantine policies at preventing arriving individuals infected with Covid-19 from entering the community and spreading the infection in the UK population.
“For projected weekly incoming passenger volumes of 409,800 from the EU in August 2020, we estimate that just 45 infectious travellers, or 0.01% of air travellers, would be released into the local community. This is equivalent to one infectious person per 10,000 travellers, which can be compared to the 57 per 10,000 local community risk in England in September and October.
“Given that a number of passengers travelling from the EU to the UK do not have to adhere to the quarantine policy due to travel corridors, and given there is no testing scheme in place, it is likely that some infectious travellers are also being released into the community under the government’s current policy.”
There are four passenger testing regimes considered in the modelling paper; ‘test and isolate’ three days pre-departure, testing on departure, testing on arrival and ‘test and release’ one to eight days after arrival. The modelling also considers two test types; RT-PCR (Reverse Transcription Polymerase Chain Reaction) and RT-LAMP (Reverse Transcription Loop-mediated Isothermal Amplification), both of which can be used in all regimes with the exception of testing on departure, where RT-LAMP is the most appropriate due to its c.60 minute result turnaround time. The paper finds RT-LAMP “comparable in sensitivity” to RT-PCR, having “significant practical and logistical benefits.”
RT-LAMP tests are found to be more effective at screening infectious days than RT-PCR testing in on-arrival or post-arrival regimes, despite its marginally lower sensitivity. This is due to a shorter delay between test administration and test results for this technology, resulting in higher compliance if travellers receive a positive test.
Even when 100% compliance with testing regimes is assumed, the new modelling suggests that a single test on arrival would catch 58% of infectious travellers, far higher than PHE’s 7% estimate, and above LSHTM’s (45%) and the Animal and Plant Health Agency’s (APHA) (40%) estimates of effectiveness. PHE’s 7% figure significantly underestimates the effectiveness of passenger testing, as identified in recent work. The result is also in line with real-world evidence from the Jersey and Iceland case studies recently considered.
The latest paper has been submitted to the Global Travel Taskforce.
Notes to editors
What are ‘infectious days’?
Every traveller with Covid-19 is infectious for a different number of days, depending on when they first contracted the virus infectious days are the average number of days that an infected individual (symptomatic or asymptomatic) spends in the community risking transmission of Covid-19 to others. For each scenario the authors calculated infectious days that would be released into the community by summing up the number of days an individual was not in quarantine while being infectious.
By using this metric and applying a new measure - % of infectious days screened - Oxera and Edge Health can capture a better view of the different policy scenarios’ effectiveness at reducing infection spread. That’s because, unlike infectious individuals screened, once non-compliance is introduced, infectious days screened account for changing compliance levels upon receipt of a positive test or upon developing symptoms.
The authors of the paper are:
Dr Kit Yates, Senior Lecturer and co-director of the Centre for Mathematical Biology in the Department of Mathematical Sciences, University of Bath
George Batchelor, a Cofounder and Director of Edge Health, a leading analytics firm that works with the NHS
Michele Granatstein, Partner at Oxera, leading its Aviation team and advising airports, airlines, regulators and policymakers across Europe and internationally
Dr Eva Deuchert, Senior Adviser at Oxera
Jennifer Connolly, an experienced health economist from Edge Health
Mona Sachter, an Analyst at Oxera
Maria Starovoitova, a data scientist from Edge Health
About Edge Health
Edge Health is a firm that specialises in helping organisations to deliver better health and care outcomes by creating better analytics and insights from their data. This work ranges from consultancy through to the development of bespoke data products. Established in 2017, Edge Health is a firm composed of economists and data scientists who have worked extensively with the NHS as well as local government, private and independent organisations.
Oxera is an economics consultancy that advises companies, policymakers, regulators and lawyers on any economic issue connected with competition, finance or regulation. We have been doing this for more than 35 years, gathering deep and wide-ranging knowledge in the sectors we specialise in, such as Aviation. We have a reputation for credibility and integrity among those we advise, and among key decision-makers, such as policymakers, regulators and courts.
The independent review was commissioned by a consortium of airlines, airports and industry organisations: Virgin Atlantic, IAG, TUI, Heathrow, MAG, Collinson, Airlines UK and IATA.
Contact: George Batchelor
---------  This is available at the following link: https://github.com/cmmid/travel_screening_strategies
 Steens Anneke, Freiesleben de Blasio Birgitte , Veneti RT-LAMPrini , Gimma Amy , Edmunds W John , Van Zandvoort Kevin , Jarvis Christopher I , Forland Frode , Robberstad Bjarne . Poor self-reported adherence to COVID-19-related quarantine/isolation requests, Norway, April to July 2020. Euro Surveill. 2020;25(37):pii=2001607. https://doi.org/10.2807/1560-7917.ES.2020.25.37.2001607
 SAGE, Multidisciplinary Task and Finish Group on Mass Testing, Behavioural Considerations. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/916896/tf ms-mass-testing-behavioural-considerations-s0724-200827.pdf
 Smith, L.E. et al. (2020) Adherence to the test, trace and isolate system: results from a time series of 21 nationally representative surveys in the UK (the COVID-19 Rapid Survey of Adherence to Interventions and Responses [CORSAIR] study) Preprint https://www.medrxiv.org/content/10.1101/2020.09.15.20191957v1.full.pdf