Pandemic modelling for regions implementing an elimination strategy =================================================================== * Amy Hurford * Maria M. Martignoni * J.C. Loredo-Osti * Franics Anokye * Julien Arino * Bilal Saleh Husain * Brian Gaas * James Watmough ## Abstract During the COVID-19 pandemic, some countries, such as Australia, China, Iceland, New Zealand, Thailand and Vietnam, successfully implemented an elimination strategy. Until June 2021, Atlantic Canada and Canada’s territories had also experienced prolonged periods with few SARS-CoV-2 community cases. Such regions had a need for epidemiological models that could assess the risk of SARS-CoV-2 outbreaks, but most existing frameworks are applicable to regions where SARS-CoV-2 is spreading in the community, and so it was necessary to adapt existing frameworks to meet this need. We distinguish between infections that are travel-related and those that occur in the community, and find that in Newfoundland and Labrador (NL), Nova Scotia, and Prince Edward Island the mean percentage of daily cases that were travel-related was 80% or greater (July 1, 2020 – May 31, 2021). We show that by December 24, 2021, the daily probability of an Omicron variant community outbreak establishing in NL was near one, and nearly twice as high as the previous high, which occurred in September 2021 when the Delta variant was dominant. We evaluate how vaccination and new variants might affect hypothetical future outbreaks in Mt. Pearl, NL. Our modelling framework can be used to evaluate alternative plans to relax public health restrictions when high levels of vaccination are achieved in regions that have implemented an elimination strategy. Keywords * Importations * COVID-zero * Elimination * SARS-CoV-2 * Atlantic Canada * COVID-19 * Pandemic preparedness * Travel-related * Reopening ## Introduction To manage SARS-CoV-2 infections, countries including Australia, China, Iceland, New Zealand, Thailand and Vietnam used an *elimination approach* (also known as a *zero-COVID* policy). This two-pronged approach combines strong border control aiming to diminish travel-related cases as much as possible, with pharmaceutical (PIs) and non-pharmaceutical interventions (NPIs) that curtail or completely end community transmissions if border measures fail (Baker et al., 2020a; Heywood and Macintyre, 2020). Elimination differs from *eradication* in that its intended region of influence is localised, typically to the jurisdiction pursuing the goal. This policy was also used in infranational jurisdictions such as Atlantic Canada and Canada’s territories (Bignami, 2021; Contandriopoulos, 2021). The elimination strategy has had mixed success. Altogether, countries that used it had less SARS-CoV-2 mortality until the end of 2021; e.g. see Baker et al. (2020b); Nam et al. (2020). However, there were also issues; for instance, Hong Kong managed very well through very strict border control, quarantine and NPIs, but did not vaccinate abundantly, which exposed the population to severe disease outcomes when the Omicron variant emerged (Ma and Parry, 2022). Further, the feasibility and success of an elimination strategy may depend on variant characteristics, and jurisdictional geographic and social characteristics (Silver, 2022; Martignoni and Hurford, 2022). In Canada, provinces that experienced prolonged periods with no SARS-CoV-2 cases in the community are the Atlantic Provinces (Newfoundland and Labrador (NL), Nova Scotia (NS), Prince Edward Island (PE), and New Brunswick (NB)), and the Territories in Northern Canada (Nunavut (NU), Northwest territories (NT), and Yukon (YT)). Other provinces all reported SARS-CoV-2 in their communities consistently from Fall 2020 until the end of 2021, when eligibility for Polymerase Chain Reaction (PCR) tests was restricted to a much smaller proportion of the provincial populations, and substantially impacted reported case numbers. This highlights some of the success stories of elimination since provinces such as NL, which implemented border restrictions (Hurford et al., 2021) and NPIs consistent with an elimination approach (Hurford and Watmough, 2022), also achieved prolonged periods with no community cases and low SARS-CoV-2 mortality: 19 deaths from the beginning of the pandemic until December 31, 2021 (Government of Newfoundland and Labrador, 2021) for a provincial population of 522,453 people (Statistics Canada, 2021). However, the elimination strategy also comes with costs, both societal and economical. Some argue that the benefit of the strategy far outweighs its cost (Philippe and Marques, 2021) and point out that it was well received by the population (Choi et al., 2021). Others note the policies impact on mental health (Aknin et al., 2022), leading some to assert that the costs are more substantial than the benefits, a view widely endorsed by political leaders in countries and regions that did not use the policy. Both these opinions are most likely valid *in specific locations and for specific periods of time* as costs and benefits likely depend on complex interactions between regional characteristics, public health policy, community behavioral responses, and variant epidemiological characteristics. As a consequence, it is important to obtain models providing a satisfactory description of the process. In the following, we call an importation an individual who arrives in a jurisdiction from another jurisdiction while infected with the virus. An importation occurs due to the traveller becoming infected at the point of origin or during travel to their destination. Travel-related infections consist of both importations and close contacts who become infected by the infected traveller. A community infection is an infection from an individual with a travel-related infection to a community member that is not a close contact of the traveller, or from a community member to another community member. We note that this terminology differs from that used in NL Public Service Advisories, which reports cases as ‘related to international (or domestic) travel’ or ‘close contacts of a known case’. Some community infections may be travel-related but are not understood to be such as no direct chain of transmission can be established from a traveller. Travel-related and community cases differ from infections in that cases are the infections that are reported. Travel-related and community infections arise through different processes, and therefore carry different risks and occur at different rates. The rate of arriving imported cases is dependent on the prevalence of infection at the travellers’ points of origin, the rates of inbound travel to the local community, and the risk of infection during travel (Russell et al., 2021). The rate that travel-related cases generate subsequent infections depends on contact rates with community members and can be curtailed through quarantine (Arino et al., 2020; Chen et al., 2021; Dickens et al., 2020), but such contact rates may be substantially higher with individuals that are defined as close contacts of the infected traveller (i.e., household members) as compared to members of the wider community. As a consequence, for regions with few community cases of SARS-CoV-2, it is necessary to distinguish between importations, close contacts who were infected by an imported infection, and community infections (see Price et al. 2020 for related comments). Here, we develop an approach to estimate the potential future impact of SARS-CoV-2 in communities that have experienced long periods with no SARS-CoV-2 community cases, to inform the relaxation of an elimination strategy once community vaccination levels are high. During the pandemic, new SARS-CoV-2 variants emerged (Otto et al., 2021), and our framework considers this evolving risk. Our approach uses two models: importations and community spread, in a pipeline rather than only a single model that couples both. Our first model describes the probability of a community outbreak, and our second model assumes that a community outbreak is occurring and considers the number of reported cases over time for a hypothetical future outbreak. The first model that describes the probability of a community outbreak is not coupled to the community spread model because for such coupling community outbreaks might hypothetically begin on any given day, and averages taken across such outbreaks obscure key information (Juul et al., 2021). Considering a pipelined uncoupled framework is useful because some decisions that public health officials make are conditional on whether a community outbreak has been detected (notably the implementation of NPIs as part of an elimination strategy), while other decisions are better informed by the average across outbreaks with all possible hypothetical start dates (i.e., generally applicable measures, such as provincial mask mandates). We consider three categories of public health measures: border restrictions, NPIs in the local community, and vaccination. In July 2021, most Canadians had received at least one dose of a SARS-CoV-2 vaccine, and there was a need to transition to a sustainable approach for SARS-CoV-2 management should high immunity levels be maintained. At this time, continued isolation of regions that implemented elimination strategies might be unrealistic, particularly given the economic and social impacts of these strategies (Committee for the Coordination of Statistical Activities, 2021). As a consequence, there is a need to develop guidelines to advise regions with zero or low SARS-CoV-2 prevalence in exiting elimination strategies (Lokuge et al., 2021; Open Society Common Purpose Taskforce, 2021). This remains an important topic even as most Canadian provincial governments have relaxed COVID-19 control measures. Indeed, vaccine coverage still lags in a large proportion of the world, and the emergence risk of novel variants remains high (Otto et al., 2021). ## Materials and Methods ### Data Our analysis combines data from multiple sources (summarized in Table 1). Data are from the Public Health Agency of Canada or provincial governments except for travel-related case data. Some of the data for travel-related cases are from the COVID-19 Canada Open Data Working Group (CCODWG) (Berry et al., 2020; Berry, 2021), a group of volunteers who curate data from government and non-government sources. We validate the CCODWG data (Fig. S1) for NL and NB by comparing to government data from the Newfoundland and Labrador Centre for Health Information (NLCHI; Health Research Ethics Board reference number 2021.013), and with our own data curated from public releases from the Government of New Brunswick (Government of New Brunswick, 2022). We find that the CCODWG data describing travel-related cases is consistent with travel-related cases as reported by the NL and NB provincial governments (Figure S1). View this table: [Table 1:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/T1) Table 1: Data sources. The line list for CCODWG (Berry et al., 2020) was discontinued on May 31, 2021, and as such, no data on travel-related cases or cases in close contacts of travellers are available from CCODWG after this date. All modelling was performed in R (R Core Team, 2022). All publicly available data and code is archived at [https://github.com/ahurford/pandemic-COVID-zero](https://github.com/ahurford/pandemic-COVID-zero). Parameters estimates are summarized in Tables 2 and 3 with complete details provided in the Supplementary Material. View this table: [Table 2:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/T2) Table 2: Variant parameter estimates. View this table: [Table 3.](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/T3) Table 3. Parameter estimates to calculate the probability of a community outbreak, *S*(*t*) (Equation 4) ### Modelling framework Central to our approach are two quantities: *n**j,k*(*t*), the number of travel-related infections, and ![Graphic][1], the probability that a traveller or their close contacts infects a community member; both quantities depend on the date *t*. Travellers and their close contacts are indexed by their vaccination status, i.e., the number of vaccination doses completed, *j* which can be 0, 1, 2, or 3, and the infecting variant, *k*, referring to the Original (*OR*), Alpha (*α*), Delta (*δ*), or Omicron (*o*) variants (BA.1 subvariant). Border control measures applying to travellers with vaccination status *j* are described by *i**j*(*t*), and ![Graphic][2] also depends on vaccination levels and NPIs in the local community. ### Statistical model of imported cases to NL We used a Poisson regression to model the daily number of imported cases arriving to NL. Explanatory variables were time series of active cases per 10,000 population for (from east to west) Nova Scotia, Quebec, Ontario, Manitoba, Saskatchewan, Alberta and British Columbia. Provincial population sizes were based on Statistics Canada (2021) estimates for the first quarter of 2021. Fitted coefficients were constrained to be non-negative, since it is not sensible that high infection prevalence in other provinces could reduce the number of imported cases arriving in NL. No intercept was considered as we assumed no importations occurred when there were no active cases in Canada. ### The probability of a community outbreak To estimate the probability of a community outbreak arising from an infected traveller, we first obtain the number of travellers and their close contacts with vaccination status *j* and infected with variant *k* as, ![Formula][3] where *c* is the number of close contacts infected per imported infection, and *D*(*t*) is the number of imported infections reported on date *t*. We assume all infected travellers are identified and reported. Equation (1) assumes a similar frequency of vaccination statuses for travellers and their close contacts, and that the number of close contacts infected from an infected traveller does not depend on the variant. Given a travel-related infection, *T**j,k*(*t*) is the probability that the traveller has vaccination status *j* and is infected with variant *k*, ![Formula][4] where ![Graphic][5] is the fraction of travellers with vaccination status *j* at time *t, v**k*(*t*) is the frequency of the variant *k* at the origin sites of travellers, and *z**j,k* is the probability that a traveller with vaccination status *j* is infected with the variant *k*, where we assume no changes in *z**j,k* over time, i.e., as might occur due to waning of the vaccination. As data on variant frequencies is not reliably available for jurisdictions within Canada, we parameterize *v**k*(*t*) as the variant frequency in Canada. To parameterize *z**j,k*(*t*), we equate reported vaccine efficacies against symptomatic infection with the probability of infection (see Table 2). Realistically, vaccines prevent less against infection and transmission than symptomatic infection, however, data for vaccine efficacies against infection and transmission are less available. For each imported infection or their infected close contacts, the probability of transmission to a community member is, ![Formula][6] where ![Graphic][7] is the probability that a traveller with vaccination status *j* infected with the variant *k* infects an unvaccinated member of the destination community when the lowest level of NPIs are in place, *ω*(*t*) describes reduction in the probability of infection due to the level of NPIs in place at time *t*, and *i**j*(*t*) describes the border restrictions applying to a traveller with vaccination status *j* on date *t*. The parameterization of ![Graphic][8]is quite involved, and provided in Table 3) and the Supplementary Material. The fraction of the community with different vaccination statuses is ![Graphic][9]. As this methodology will be applied to NL until December 24, 2021, and because no third dose of the vaccine had been administered at the time, we do not consider members of the local community with three doses of the vaccine. A critical component of our parameter estimation was to observe that only 1 community outbreak occurred in NL due to the Alpha variant, and that when the Alpha variant was prevalent, border restrictions for travellers arriving in NL were unchanging (most travellers were to self-isolate for 14-days). As such, the estimated probabilities of infection spread from an infected traveller (or their close contact) to a member of the NL community is based on a concrete observation from the province where the analysis is to be applied. This parameter estimate is then the basis for determining the impact of different border restrictions that may involve self-isolation for less than 14 days and/or completing Rapid Antigen Test (RAT) or Polymerase Chain Reaction (PCR) tests, which are derived from known properties describing the infection probability for different days since exposure (Ferretti et al., 2020), and sensitivities of the tests (Hellewell et al. 2021, see Supplementary Material for complete details). For the model parameterization, we equate the strictness of local NPIs with Special Measures Orders in NL for mandatory masking, or NL government recommendations to wear masks. Many NPIs in the local community, beyond mask wearing, may influence the strictness of NPIs, e.g., capacity limits, but we considered only mask wearing due to a lack of data to parameterize the effect of other measures, and because sometimes these other NPIs have applied only to specific regions within NL. On each date *t*, the probability that at least one traveller or their infected close contacts infects a community member is the probability that no such infections occur for any variants or traveller vaccination statuses subtracted from 1, ![Formula][10] This quantity, *S*(*t*), describes the daily probability that a community outbreak of SARS-CoV-2 may begin in a region that is not currently experiencing a community outbreak. Quantifying SARS-CoV-2 risks in regions that do not have SARS-CoV-2 community cases was an area of need during the first 18 months of the COVID-19 pandemic. Equation 4 has the form of a multivariate geometric distribution; however, related questions can be considered using analogous formulations based on a multinomial distribution, i.e., to calculate the expected number of community infections from a traveller or their close contacts, and a negative binomial distribution when considering super-spreading events. ### Infection spread in vaccinated communities A second aspect of quantifying SARS-CoV-2 risk in regions that do not have community cases, is to understand, ‘if a community outbreak is established, how will the number of cases change over time, and how many cases will occur in the outbreak?’ To illustrate this modelling for a region that had few community cases of SARS-CoV-2, we again consider NL. In NL, through to December 15, 2021, the largest community outbreak of SARS-CoV-2 occurred from February 1 to 27, 2021, with spread predominately in the Mt. Pearl region (in the greater St. John’s area), and due to the Alpha variant. We fit a stochastic Susceptible-Infected-Recovered (SIR) model (see Supplementary Material for details) by calibration to the data describing the daily new reported cases, and their symptom onset dates, for cases identified as belonging to the Mt. Pearl outbreak. These fitted parameters are the basis to explore the dynamics of a hypothetical future outbreak in Mt. Pearl, NL. In response to the Mt. Pearl outbreak, on February 10, 2021 the province implemented its strictest level of NPIs (Alert level 5) in the St. John’s region. For comparison, future scenarios retain this pattern of NPI implementation, although it is possible to explore scenarios without this assumption. We consider future scenarios where vaccination levels may have changed, and where a different variant may cause the outbreak. For simplicity in interpreting the results of vaccination scenarios, we assume that all individuals in the community are either unvaccinated or have had two doses of vaccine. ## Results From July 1, 2020 to May 31, 2021, most of the SARS-CoV-2 cases reported in Atlantic Canada and Canada’s territories were travel-related (Figure 1). The period prior to July 1, 2020 was not considered because very few cases of any type were reported during this time. Notable differences that occur between these jurisdictions are that a much lower percentage of daily reported cases were travel-related in NB (mean = 13%, median = 37%), NT (mean = 0%, median = 13%) and YT (mean = 0%, median = 30%), as compared to NL (mean = 100%, median = 77%), NS (mean = 80%, median = 62%), and PE (mean = 100%, median = 91%) (Figure 1G). The total number of imported cases (and their close contacts reported in parenthesis) from July 1, 2020 to May 31, 2021 also differed between jurisdictions with NL: 259 (159); NS: 239 (281); and NB: 204 (302), having reported at least 2.75 times more travel-related cases than PE: 112 (40); and the YT: 12 (18); and NT: 10(6) having reported very few travel-related cases at all (Figure 1H). Other Canadian provinces and Nunavut (NU) were not considered because travel-related case data was not reliably reported for these jurisdictions. We found that the number of importations to NL was predicted as 0.105 times the number of active cases per 10,000 population in NS (Figure 2). Estimated coefficients for the contribution of other provinces to the prediction of imported cases to NL were not different than zero. ![Figure 1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F1.medium.gif) [Figure 1:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F1) Figure 1: In Atlantic Canada and Canada’s territories, from July 1, 2020 to May 31, 2021, most SARS-CoV-2 cases were importations and close contacts of these travellers. A-F: Imported cases (dark shading, solid line), their close contacts (medium shading, dashed line), and community cases (light shading, no line) with the vertical axis limit as 20% more than the maximum number of reported weekly travel-related cases so that brief periods of large community outbreaks do not dominate the graphs. G: Percentage of reported daily cases that are travel-related with the median shown as a large semi-transparent circle. H: Total number of imported cases and their close contacts (July 1, 2020 – May 31, 2021). We note that Figure 4A does not consider the arrival of imported cases. This is because the Mt. Pearl data was strictly for cases known to belong to this outbreak and therefore does not include imported cases and any subsequent infections that could occur. The Omicron variant emerged only after NL had reopened to Canadian travellers, and realistically a hypothetical Omicron variant community outbreak might also consider the increased importation rate that would occur with high Omicron prevalence in Canada and no border restrictions for travellers arriving to NL, as was the case since early 2022. Using the same rationale as was used to derive Equation 4, but instead calculating the expected number of infections spread from an infected traveller (or their close contacts) to an NL community member, we estimate 0.014 such infections, on average, per day for the Alpha variant for February 1-27, 2021, and 5.33 such infections for the Omicron variant on December 24, 2021. The number of cases reported in the Mt. Pearl outbreak was 472 (Figure 4B, green dashed line). This outbreak was assumed to occur when community members were fully susceptible to infection. For other variants the total number of reported cases after 27 days (the duration of the Mt. Pearl outbreak) when the community is fully susceptible are: Original variant (161); Delta (4400); and Omicron variant (437). We assumed that community members could be either unvaccinated or have two doses of vaccine. The effect of vaccination is to substantially reduce the number of reported cases in the outbreak after 27 days for all variants except the Omicron variant (Figure 4B). As the Omicron variant has low vaccine efficacy for community members with two doses of vaccine, the number of reported cases after 27 days for an outbreak of the Omicron variant decreases only very slightly as the community is increasingly more vaccinated (Figure 4B, red line). ## Discussion In regions that have extended periods with few community cases of SARS-CoV-2, for example, regions that effectively implement an elimination strategy, travel-related cases are a high proportion of reported cases (Arino et al., 2021; Godin et al., 2021), and modelling importations is particularly important; e.g., Zhang et al. (2022). Here, we extend such importation modelling to incorporate border restrictions, community vaccination levels, and NPIs into the risk assessment frameworks for regions with few community SARS-CoV-2 infections. Atlantic Canada and Canada’s territories experienced few SARS-CoV-2 cases prior to June 2021, however, there were some differences between these jurisdictions. NT and YT reported few travel-related cases. NL, NS, and NB reported similar numbers of travel-related cases, but NB reported a much lower percentage of daily cases that were travel-related (Figure 1). Finally, while NL and NS had similar epidemiology until May 31, 2021, during this time NL had enacted strict travel restrictions (Hurford et al., 2021), while NS enacted an extensive community testing program (Johnson-León et al., 2021). The YT implemented strict travel restrictions, but experienced an outbreak of the Gamma variant that overwhelmed hospital capacity (McPhee-Knowles et al., 2022). We considered a statistical model describing the daily number of reported importations arriving in NL. During the pandemic response it was helpful to use this approach to forecast importations so that future risk could be assessed using Equation 4. That was not done in this manuscript because such an exercise would never be current, but this could be valuable to assess the threat of a new variant, or the impacts of waning immunity in the future. We applied our framework (Equation 4) to inform the potential for community outbreaks in NL. The estimated risk is broadly consistent with the timing of actual community outbreaks in NL (Figure 3). There was an elevated risk of community outbreaks in NL after reopening began on July 1, 2021 (Figure 3); however, community outbreaks of the Delta variant were controlled due to high vaccination levels and vaccine efficacy, and stringent regional NPIs in place during that time (Figure 3E). In mid-December 2021, the NL community was vulnerable to a community outbreak of SARS-CoV-2, and this was due to the easing of public health restrictions since July 1, 2021 (3), and the low efficacy of two vaccine doses against the Omicron variant. ![Figure 2:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F2.medium.gif) [Figure 2:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F2) Figure 2: The daily number of imported cases to NL (reported by NLCHI for March 15, 2020-December 24, 2021) is reliably predicted as 0.105 times the number of active cases per 10,000 population in NS. This relationship was determined by using the publicly available CCODWG data (curation of these data ended on May 31, 2021), but still the relationship validates well with the NLCHI data which spans a longer time period and includes importation of the Omicron variant. The probability at least one infection spreads from a traveller or their close contacts to an NL community member (as calculated by Equation (4)) is near 1 by December 24, 2021 (Figure 3E). At this time, a community outbreak involving the Omicron variant was already occurring, with the first Omicron variant case reported in St. John’s on December 15, 2021. The high risk of a community outbreak from mid-December onward was due to both the high number of imported cases (Figure 3C) and the reduced efficacy of two vaccine doses in protecting the NL community from infection with the Omicron variant (Figure 3D). The probability that at least one infection spread to the NL community from a travel-related infection on December 24, 2021, was approximately 2 times higher than the previous high value, which occurred in September 2021 when the Delta variant was dominant (Figure 3A), and after the province relaxed entry requirements for Canadian travellers on July 1, 2021 (Table 3). ![Figure 3:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F3.medium.gif) [Figure 3:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F3) Figure 3: The probability of a community outbreak in NL rises sharply in mid-December 2021 due to high numbers of imported cases and low vaccine efficacy for NL community members with two doses of vaccine exposed to the Omicron BA.1 variant. A: Variant frequency, *v**k*(*t*). B: Vaccination levels for Canada, ![Graphic][11] (shaded), and NL, ![Graphic][12] (lines). C: Imported cases, *D*(*t*). D: The probability a traveller or their close contact infects an unvaccinated NL community member with the Omicron BA.1 variant, ![Graphic][13] when NPIs are at their lowest level (for other variants, see Figure S2). E: The probability that at least one infected traveller or their close contacts infects an NL community member, *S*(*t*) (black - equation 4; with the variant-specific probability shown with colours). The timing of actual community outbreaks with more than 5 cases are shown along the bottom bar. Grey dashed vertical lines show July 1, 2022, when the NL relaxed border restrictions and reopened to Canadian travellers. The stochastic SIR model (Figure 4A, green lines and shading) shows close agreement with the data from the Mt. Pearl, NL outbreak in February, 2021 (Figure 4A, green dots). When the Mt. Pearl outbreak occurred few community members were vaccinated, such that all the scenarios shown in Figure 4A assume a fully susceptible community. Our estimation of the relative transmissibility of variants found that in a fully susceptible community, the Alpha and the Omicron variants (specifically, the BA.1 subvariant) have similar net reproduction numbers (see the Supplementary Material for the derivation). The Delta variant is much more transmissible in a fully susceptible community than other variants (see Table 2), such that a hypothetical Delta variant outbreak in a fully susceptible Mt. Pearl, NL community (Figure 4A, blue line) cannot be completely shown given the y-axis limits that were set to emphasize the actual Mt. Pearl Alpha variant outbreak. ![Figure 4:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F4.medium.gif) [Figure 4:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F4) Figure 4: Epidemiological model fit and hypothetical future variant and vaccination scenarios for Mt. Pearl, NL. In A) the lines show the mean of 100 stochastic simulations and the shaded region shows the range. The Mt. Pearl, NL, outbreak was due to the Alpha variant and the y-axis limits were selected to show the Alpha variant (green line) and the Mt. Pearl data (A. green dots), which meant that large values for Delta variant scenarios are truncated from the graph. Not shown is that the peak number of reported daily new cases for the Delta variant is 479. In B) vaccination scenarios assume community members are either unvaccinated or vaccinated with 2 doses. The values are the mean of 100 simulations. Not shown due to the graph truncation is that after 27 days of a Delta outbreak in a fully unvaccinated community we estimate 4400 reported cases. For Parameter estimates see Table 2 and for model details see the Supplementary Material. Our analysis considers only known travel-related infections, such that estimates per infected traveller equate to per *known* infected traveller. In NL, for the pandemic until July 1, 2021, testing of arriving travellers was intensive (owing to few ports of entry, reduced travel volumes (Hurford et al., 2021), testing requirements for rotational workers, and requests for travellers potentially exposed during inbound flights to report for asymptomatic testing). This intensive testing, combined with few occurrence of community cases, suggests that a high proportion of imported cases were detected in NL during this time. The main limitation of our analysis is parameter estimation. It is difficult to estimate the change in relative transmissibility due to a new variant because these data are estimated in different regions (or pooled across regions), and as the susceptible population changes owing to vaccination and infection during the time period that the estimation is made. Vaccine efficacies are estimated in specific populations, and application to other regions assumes no differences in population structure with regard to age and immunity, and does not estimate protection against infection and onward transmission, which is a critical parameter for epidemiological models. Finally, the impact of NPIs on transmission is difficult to assess, and how the effectiveness of NPIs is impacted by new variant characteristics is unknown. Our work was motivated by a need for regions that successfully implemented an elimination strategy during the first 18 months of the COVID-19 pandemic to quantify the risk of SARS-CoV-2 spread in their communities, and from a need for guidelines to exit an elimination strategy when high vaccination levels have been achieved to be developed. While guidelines for reopening have been developed by many jurisdictions, those using criteria expressed as the number of observed community cases (Anderson et al., 2021; Nali et al., 2021) are not helpful for regions that are reopening when there are few community cases. There is a need to communicate reasonable expectations to the public in regions where elimination has been implemented, as relaxation of measures might have little or no impact on the epidemics when infection prevalence is already high (Russell et al., 2021; Chen et al., 2020), but might well bring risk to populations with zero or low COVID-19 prevalence (Russell et al., 2021; Chen et al., 2020; Arino et al., 2021), even when measures are carefully and reasonably relaxed, and particularly if the prevalence of variants of concern is higher outside the jurisdiction than in (Wells et al., 2020; Grépin et al., 2021). Prior to May 31, 2021, Atlantic Canada and Canada’s territories had experienced prolonged periods with few community SARS-CoV-2 cases. In this manuscript, we characterize differences within these jurisdictions, and distinguish between travel-related and community cases (Figure 1). We illustrate a type of epidemic modelling that is useful in these regions. This framework extends importation modelling such that border restrictions, variants, NPIs and vaccination in the local community are considered. Additionally, hypothetical future outbreaks are considered by simulating variant and vaccination scenarios. Our framework can be used to inform the risk associated with different candidate reopening plans when vaccination levels are high in regions that have experienced prolonged periods with few SARS-CoV-2 cases, and help inform plans to exit an elimination strategy. ## Data Availability All publicly available data used in this study are available at https://github.com/ahurford/pandemic-COVID-zero, but until the work completes peer-review please request these data and code by emailing ahurford@mun.ca ## Supplementary Material ### Validation of the CCODWG data The COVID-19 Canada Open Data Working Group (CCODWG) (Berry et al., 2020; Berry, 2021) is a group of volunteers who curate data from government and non-government sources. Data from the CCODWG line list are the basis of Figure 1. To validate these data we compared the reported number of imported cases in the CCODWG data for Newfoundland and Labrador (NL) and data from the Newfoundland and Labrador Centre for Health Information (NLCHI). Additionally, we compared the reported number of imported cases in the CCODWG data for New Brunswick (NB) with the same data reported from News Releases from the Government of New Brunswick (Government of New Brunswick, 2022). Each news release has a ‘new cases’ section that specifies the number of new cases and whether they are travel-related (which we define as an imported case in this manuscript), contacts of previously confirmed cases, or under investigation. For the most part, cases that are labelled ‘under investigation’ are not travel-related, since travel-related cases are quickly apparent. The news releases refer to the cases of the previous day and are recorded as occurring on the previous date. There is frequently a 1-day offset with the CCODWG data for this reason (corrected for in Figure S1). Figure S1 shows that the CCODWG data validates well with government data. Figure S1A and C show few differences between the CCOWDG data (lines) and the government data (shading). For both NL and NB over 60% of the daily reported number of imported cases are identical for both the CCOWDG data and the government’s data (Figure S1 B and D). ### Additional details for Table 2 Campbell et al. (2021) reports the Alpha variant (B.1.1.7) and the Delta variants (B.1.617.2) as having effective reproduction numbers 29% and 97% greater than non-variants of interest or concern, respectively (*b**OV,α* = 1.29 and *b**OV,δ* = 1.97) (pooled estimate for reported case data from 15 countries to July 3, 2021). It should be noted that these estimates reflect the change in the effective reproduction number for these variants of concern in populations with some immunity, and not in a fully susceptible population. To estimate the change in the net reproduction number (i.e., in fully susceptible population), we consider the amount of vaccination at a reference date corresponding to the data collection period reported in Campbell et al. (2021). We do not consider infection-derived immunity in our calculations as vaccine-derived immunity is a more substantial source of immunity, but we note that this is a simplification. The Delta variant was still relatively recently emerged when data collection ended on July 3, 2021 for the Campbell et al. (2021) analysis, and so we consider June 1, 2021, as the reference date. The relationship between the effective reproduction number (considering vaccine-derived immunity), and the net reproduction number (in a fully susceptible population) for the variant, *k*, is assumed to be, ![Figure S1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F5.medium.gif) [Figure S1:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F5) Figure S1: The daily number of imported cases reported in the CCODWG line list for NL and NB agrees with the number of imported cases reported by the NL and NB governments. A: The CCOWDG daily reported importations for NL (orange line) compared to NLCHI data (orange shading). B: Frequency histogram of reported importations for NL: CCOWDG value - NLCHI values (July 1, 2020 – May 31, 2021). C: The CCOWDG daily reported importations for NB (black line) compared to NB government reports (grey shading) D: Frequency histogram of reported importations for NB: CCOWDG value - NB government reported value (January 1, 2021 – May 31, 2021). ![Formula][14] where *x**j* is the fraction of individuals with *j* doses of vaccine, which for this calculation is taken on the reference date, and where *z**j,k* is the vaccine efficacy for the vaccination status, *j*, and the variant, *k*. Vaccine efficacies describe the prevention of symptomatic infection, which we equate with preventing infection owing to limited data for the latter. Noting that our values quoted from Campbell et al. (2021) are pooled estimates for 15 countries, we consider worldwide vaccination levels (Mathieu et al., 2021) on the reference date, June 1, 2021: ![Formula][15] where having one dose of vaccine is equated with being partially vaccinated. Noting that, given our assumptions, ![Formula][16] and substituting Equation 5, then: ![Formula][17] The World Health Organization Weekly Epidemiological update for March 22, 2022, states that Omicron has an effective reproduction number that is 85% greater than the Delta variant (*b**δ,o* = 1.85) under the assumption of equal generation times. Our analysis considers the BA.1 subvariant of Omicron, and so we consider February 1, 2022, as our reference date since the BA.1 variant was prevalent at this time. Using the same methodology as the previous calculations, ![Formula][18] Then, ![Formula][19] such that, ![Formula][20] These calculations suggest that in a fully susceptible population the Delta variant is the most transmissible (97% higher net reproduction number than the Original variant), and the Omicron BA.1 variant is the least transmissible (24% higher net reproduction number than the Original variant), however, given worldwide vaccination levels on February 1, 2021, the Omicron variant is much more transmissible owing to much higher vaccine efficacies for the Delta variant relative to the Omicron variant. ### Additional details for Table 3 To parameterize the probability that a traveller infects a community member, we note that from March 15, 2020 to July 1, 2021, border restrictions and community NPIs in NL saw limited changes, while vaccination levels in the NL community were low for most of this period. In NL during this time there was one community outbreak of the Alpha variant that occurred in Mt. Pearl in February 2021. This observation allows us to estimate the probability of a community outbreak under these reference conditions: all travellers are required to isolate for 14 days; mandatory maskwearing/high community NPIs; no community vaccination, and the Alpha variant. Next, we estimate how many travellers arrived in NL while infected with the Alpha variant. Multiplying the daily reported number of imported cases to NL by the estimated daily frequency of the Alpha variant in Canada, we estimate that 119 travellers infected with the Alpha variant arrived in NL over the entire study period (March 15, 2020 to December 24, 2021). For all travellers and variants, over the entire study period, each infected traveller generated a mean of *c* = 0.25 infections among close contacts (estimated from NLCHI data). Finally, we assumed that NPIs in place during this time decreased transmission by 50% (see below for details). Therefore, we estimate one community outbreak of the Alpha variant occurs per 119(1 +0.25)*/*0.5 = 298 travellers or their close contacts, when travellers were required to self-isolate for 14 days, assuming the NL community is unvaccinated, and when few NPIs are in place. Accordingly, we estimate ![Graphic][21] (1 outbreak)*/*298 = 0.0034. The subscript 0*j* refers to re-opening step 0, whereby travellers of all vaccination statuses, *j*, were required to self-isolate for 14 days, and the subscript *α* notes that this calculation is for the *α* variant. To consider different variants, we multiply ![Graphic][22] by a correction factor based on the values in Table 2, and as illustrated in Table 3. Next, we should consider how to adjust ![Graphic][23] for different reopening steps, which involved travellers self-isolating for periods of shorter than 14 days, and exiting self-isolation conditional on negative Rapid Antigen Test (RAT) or Polymerase Chain Reaction (PCR) test results. To determine the impact of shorter duration self-isolation on the probability of a community outbreak, we ask, ‘how much potential infectivity remains if an individual leaves selfisolation after *n* days?’ This depends on when the traveller was infected, and here we assume the number of days since infection of the arriving traveller is uniformly distributed between 0 to 10 days ago. Infectivity as a function of days since exposure is based on Ferretti et al. (2020) who reports that the generation interval for the Original variant follows a Weibull distribution with shape parameter equal to 2.83 and scale parameter equal to 5.67 (mean value of 5 days between infections). Summing this distribution for days since exposure from 0 to 10 gives a value near 1, meaning that after 10 days of self-isolation, there is a negligible chance of transmitting an infection. If an individual self-isolates for 8 days, assuming the uniform distribution of days since exposure at the beginning of self-isolation, then at exit from self-isolation the individual’s days since exposure is uniformly distributed from 8 to 18 days. Summing the generation interval over these values gives 0.11, which we interpret to mean that 8-day self-isolation increases the probability of a community infection by 11% relative to a 10 days or longer self-isolation. The impact of self-isolation periods of other durations are calculated similarly. If a PCR test is required to exit self-isolation, then we consider the probability of a false negative (the test sensitivity) as based on Figure 3B in Hellewell et al. (2021): ![Formula][24] where these values correspond to 0 to 10 days since exposure. We assume the PCR test occurs on the day before exiting self-isolation. On December 21, 2021, travellers with two doses of vaccine entering NL were required to self-isolate for 5 days and complete a RAT on each of these days. Our study period ends on December 24, 2021, since due to a rapid rise in infection, case data is less reliable after this date. Therefore, no individuals subject to these restrictions left self-isolation during the study period, and we assume the impact of these restrictions is mainly due to the self-isolation of travellers that arrived on December 21, 2021. The frequency of SARS-CoV-2 variants at the travellers’ points of origin *v**k*(*t*) is assumed to be the variant frequency in Canada. The proportion of travellers with a given vaccination status, ![Graphic][25], is assumed to be equal to the frequency of the different vaccination statuses in Canada. As variant and vaccination data were reported weekly, we used linear interpolation to infer daily values. We assume that mandatory masking in the community reduces the probability of infection by 50%. We assume that when masks are not mandatory, but are recommended by the government, that the probability of infection is reduced by 25%, each relative to the lowest level of NPIs. We equate the strictness of NPIs in the NL community with periods of mandatory mask wearing or mask recommendations, which is an oversimplification and only captures the impact of NPIs on infection spread very generally. ### Stochastic SIR model A stochastic Susceptible-Infected-Recovered (SIR) model was calibrated to fit the Mt. Pearl, NL outbreak data. The stochastic SIR model is a Markov chain for the time interval (*t, t* + Δ*t*] the transition probabilities are: ![Figure S2:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/07/18/2022.07.18.22277695/F6.medium.gif) [Figure S2:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/F6) Figure S2: Probability ![Graphic][26] of a community outbreak arising from a traveller, given the border control and community measures in place at the time, *i**t*, where the traveller has vaccination status *j* and is infected with variant *k*, considering vaccination levels in the Newfoundland and Labrador community at time, *t*. Values are not shown prior to April 1, 2021 since the only change was from no mask requirements to mandatory masking beginning on August 24, 2020 and ending on August 10, 2021. ![Formula][27] where *I**r*(*t*) and *I**u*(*t*) are the number of reported and unreported cases actively infected at time *t*, respectively. The probability that an infection is reported is *π* (assumed to be 0.6). Infections that are reported may self-isolate and this reduces overall transmission by *κ* (assumed to be 0.5). The recovery rate from infection is *γ* (assumed to be 1/10 days). The transmission rate is *β*(*t*) and is assumed to decrease from *β* to *a*1*β* after some time *t** (assumed to be 10 days after the outbreak begins). We assume the change in *β*(*t*) occurs gradually, at an exponentially decreasing rate, with coefficient *a*2, such that, ![Formula][28] Our model can be expressed as a multivariate counting process **N**(*t*) with exponential inter-arrival times and intensity, ![Formula][29] where *N**j*(*t*), *j* = 1, 2, 3, 4 are the counts up to time *t* of reported infections, unreported infections, and individuals recovered from the reported and unreported groups, respectively. We do not consider the number of susceptible individuals as a dynamic variable since for the application of this model to regions that have few community cases of SARS-CoV-2 it is likely that the outbreak infects only a very small fraction of the population. Therefore, the number of susceptible individuals is relatively unchanging. We assume that the outbreak is relatively short and that vaccination levels do not change during the outbreak. Parameter values for the calibrated fit are *β* = 0.35, *a*1 = 0.4 and *a*2 = 0.2. The effect of different variants was considered by multiplying *β* by the appropriate values given in Table 2. The effect of vaccination was considered by multiplying *β* by *x* + *x*2*z*2,*k* where *x* is the fraction of the population that is unvaccinated, *x*2 is the fraction of the population that is fully vaccinated (assumed to be 1 − *x*), and *z*2,*k* is the efficacy of two doses of vaccine against symptomatic infection (assumed to be equivalent to infection; see Table 2). Only the *β* parameter was changed for the different hypothetical outbreaks. ### NL reopening plan: Together.Again View this table: [Table S1:](http://medrxiv.org/content/early/2022/07/18/2022.07.18.22277695/T4) Table S1: NL’s reopening plan: Together.Again. Table 3 describes the actual course of reopening in NL, which deviated from the reopening plan. ## Acknowledgements AH, MM, FA, JW, BG and JA are supported by the Emerging Infectious Disease Modelling Consortium (the Canadian Network for Infectious Disease Modelling, Mathematics for Public Health, and the One Health Modelling Network for Emerging Infectious Diseases). AH was supported by an National Sciences and Engineering Research Council of Canada Discovery Grant (RGPIN 2014-05413) and funding from the Newfoundland and Labrador Department of Health and Community Services. AH acknowledges that conversations with Proton Rahman, Sanjeev Sahara, and responding to modelling requests from the NL Department of the Health and Community Services helped in the development of the modelling framework presented. NL COVID-19 data was provided by the NLCHI (Health Research Ethics Board reference number 2021.013). * Received July 18, 2022. * Revision received July 18, 2022. * Accepted July 18, 2022. * © 2022, Posted by Cold Spring Harbor Laboratory This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/) ## References 1. Michael G Baker, Nick Wilson, and Tony Blakely. Elimination could be the optimal response strategy for covid-19 and other emerging pandemic diseases. BMJ, 371, 2020. doi: 10.1136/bmj.m4907. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE3OiIzNzEvZGVjMjJfMy9tNDkwNyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIyLzA3LzE4LzIwMjIuMDcuMTguMjIyNzc2OTUuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 2. Anita E Heywood and C Raina Macintyre. Elimination of COVID-19: what would it look like and is it possible? The Lancet. Infectious Diseases, 20(9):1005, 2020. doi: 10.1016/s1473-3099(20)30633-2. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S1473-3099(20)30633-2&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=32771079&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 3. Simona Bignami. The burden of COVID-19 in Canada. Canadian Studies in Population, 48(2):123–129, 2021. doi: 10.1007/s42650-021-00056-w. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s42650-021-00056-w&link_type=DOI) 4. Damien Contandriopoulos. The year public health lost its soul: a critical view of the COVID-19 response. Canadian Journal of Public Health, 112(6):970–972, 2021. doi: 10.17269/s41997-021-00583-8. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.17269/s41997-021-00583-8&link_type=DOI) 5. Michael G. Baker, Nick Wilson, and Andrew Anglemyer. Successful elimination of Covid-19 transmission in New Zealand. The New England Journal of Medicine, 383: e56, August 2020b. ISSN 1533-4406. doi: 10.1056/NEJMc2025203. URL [https://www.semanticscholar.org/paper/c279c1aff07a1adb4af655b8d0bf47c98cacc1f2](https://www.semanticscholar.org/paper/c279c1aff07a1adb4af655b8d0bf47c98cacc1f2). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJMc2025203&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=32767891&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 6. Nguyen Hai Nam, Pham Nguyen Quy, Truong-Minh Pham, and Joel Branch. No new community COVID-19 infection in four consecutive weeks: what lesson can be learned from Vietnam. Journal of infection in developing countries, 14(10):1125–1127, October 2020. ISSN 1972-2680. doi: 10.3855/jidc.13080. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3855/jidc.13080&link_type=DOI) 7. Arisina Ma and Jane Parry. When Hong Kong’s “dynamic zero” covid-19 strategy met omicron, low vaccination rates sent deaths soaring. BMJ (Clinical research ed.), 377: o980, April 2022. ISSN 1756-1833. doi: 10.1136/bmj.o980. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE3OiIzNzcvYXByMTNfMTMvbzk4MCI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIyLzA3LzE4LzIwMjIuMDcuMTguMjIyNzc2OTUuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 8. Andrew Silver. Covid-19: What went wrong after initial success in Laos? British Medical Journal, 377, 2022. URL doi:10.1136/bmj.o994. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE2OiIzNzcvYXByMjBfMi9vOTk0IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMDcvMTgvMjAyMi4wNy4xOC4yMjI3NzY5NS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 9. Maria M. Martignoni and Amy Hurford. It’s not realistic to eliminate COVID-19 in Newfoundland and Labrador. Here’s why. Canadian Broadcasting Corportation, Newfoundland and Labrador, 2022. URL [https://www.cbc.ca/news/canada/newfoundland-labrador/covid-19-here-to-stay-1.6384033](https://www.cbc.ca/news/canada/newfoundland-labrador/covid-19-here-to-stay-1.6384033). 10. Amy Hurford, Proton Rahman, and J. Concepción Loredo-Osti. Modelling the impact of travel restrictions on COVID-19 cases in Newfoundland and Labrador. Royal Society Open Science, 8:202266, June 2021. ISSN 2054-5703. doi: 10.1098/rsos.202266. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1098/rsos.202266&link_type=DOI) 11. 1. Murty, V. Kumar and 2. Wu, Jianhong, editor Amy Hurford and James Watmough. Don’t Wait, Re-escalate: Delayed Action Results in Longer Duration of COVID-19 Restrictions. In Murty, V. Kumar and Wu, Jianhong, editor, Mathematics for Public Health, pages 201–213. Springer, 2022. 12. Government of Newfoundland and Labrador. News releases. public advisory: Update on covid-19 in newfoundland and labrador, december 31. Department of Health and Community Services, 2021. URL [https://www.gov.nl.ca/releases/2021/health/1231n06/](https://www.gov.nl.ca/releases/2021/health/1231n06/). 13. Statistics Canada. Population estimates, Quarterly. Government of Canada, 2021. URL [https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000901](https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000901). 14. Cécile Philippe and Nicolas Marques. The Zero Covid strategy protects people and economies more effectively. Technical report, Paris, Bruxelles: Institut Molinari, 2021. 15. Kimberley Choi, Namratha Giridharan, Abigail Cartmell, Dominique Lum, Louise Signal, Viliami Puloka, Rose Crossin, Lesley Gray, Cheryl Davies, Michael Baker, and Amanda Kvalsvig. Life during lockdown: a qualitative study of low-income new zealanders’ experience during the covid-19 pandemic. The New Zealand medical journal, 134:52–67, July 2021. ISSN 1175-8716. 16. Lara B. Aknin, Bernardo Andretti, Rafael Goldszmidt, John F. Helliwell, Anna Petherick, Jan-Emmanuel De Neve, Elizabeth W. Dunn, Daisy Fancourt, Elkhonon Goldberg, Sarah P. Jones, Ozge Karadag, Elie Karam, Richard Layard, Shekhar Saxena, Emily Thornton, Ashley Whillans, and Jamil Zaki. Policy stringency and mental health during the COVID-19 pandemic: a longitudinal analysis of data from 15 countries. The Lancet. Public Health, 7:e417–e426, May 2022. ISSN 2468-2667. doi: 10.1016/S2468-2667(22)00060-3. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S2468-2667(22)00060-3&link_type=DOI) 17. Timothy W Russell, Joseph T Wu, Sam Clifford, W John Edmunds, Adam J Kucharski, Mark Jit, et al. Effect of internationally imported cases on internal spread of COVID-19: a mathematical modelling study. The Lancet Public Health, 6(1):e12–e20, 2021. doi: 10.1016/s2468-2667(20)30263-2. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/s2468-2667(20)30263-2&link_type=DOI) 18. Julien Arino, Nicolas Bajeux, Stephanie Portet, and James Watmough. Quarantine and the risk of COVID-19 importation. Epidemiology & Infection, 148:e298, 2020. doi: 10.1017/S0950268820002988. URL [https://dx.doi.org/10.1017/S0950268820002988](https://dx.doi.org/10.1017/S0950268820002988). See [https://github.com/julien-arino/covid-19-importation-riskforcode](https://github.com/julien-arino/covid-19-importation-riskforcode). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1017/S0950268820002988&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 19. Tiange Chen, Siwan Huang, Guanqiao Li, Yuan Zhang, Ye Li, Jinyi Zhu, Xuanling Shi, Xiang Li, Guotong Xie, and Linqi Zhang. An integrated framework for modelling quantitative effects of entry restrictions and travel quarantine on importation risk of COVID-19. Journal of biomedical informatics, 118:103800, June 2021. ISSN 1532-0480. doi: 10.1016/j.jbi.2021.103800. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jbi.2021.103800&link_type=DOI) 20. Borame L. Dickens, Joel R. Koo, Jue Tao Lim, Haoyang Sun, Hannah E. Clapham, Annelies Wilder-Smith, and Alex R. Cook. Strategies at points of entry to reduce importation risk of COVID-19 cases and reopen travel. Journal of Travel Medicine, 27, December 2020. ISSN 1708-8305. doi: 10.1093/jtm/taaa141. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/jtm/taaa141&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=32841354&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 21. David J Price, Freya M Shearer, and Michael T Meehan. Early analysis of the australian covid-19 epidemic. eLife, 9:e58785, 2020. URL [https://doi.org/10.7554/eLife.58785](https://doi.org/10.7554/eLife.58785). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.7554/eLife.58785&link_type=DOI) 22. Sarah P. Otto, Troy Day, Julien Arino, Caroline Colijn, Jonathan Dushoff, Michael Li, Samir Mechai, Gary Van Domselaar, Jianhong Wu, David J. D. Earn, and Nicholas H. Ogden. The origins and potential future of SARS-CoV-2 variants of concern in the evolving COVID-19 pandemic. Current biology : CB, 31:R918–R929, July 2021. ISSN 1879-0445. doi: 10.1016/j.cub.2021.06.049. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cub.2021.06.049&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=34314723&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 23. J.L. Juul, K. Græsbøll, and L.E. Christiansen. Fixed-time descriptive statistics underestimate extremes of epidemic curve ensembles. Nat. Phys., 17:5–8, 2021. URL [https://doi.org/10.1038/s41567-020-01121-y](https://doi.org/10.1038/s41567-020-01121-y). 24. Committee for the Coordination of Statistical Activities. How COVID-19 is changing the world: a statistical perspective. Technical report, United Nations, March 2021. Volume III. New York, 2021. [https://unstats.un.org/unsd/ccsa/documents/](https://unstats.un.org/unsd/ccsa/documents/). 25. Kamalini Lokuge, Emily Banks, Stephanie Davis, Leslee Roberts, Tatum Street, Declan O’Donovan, Grazia Caleo, and Kathryn Glass. Exit strategies: optimising feasible surveillance for detection, elimination, and ongoing prevention of COVID-19 community transmission. BMC medicine, 19(1):1–14, 2021. doi: 10.1186/s12916-021-01934-5. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12916-021-01934-5&link_type=DOI) 26. Open Society Common Purpose Taskforce. A roadmap to reopening. Technical report, University of Sydney, 2021. 27. Isha Berry, Jean-Paul R. Soucy, Ashleigh Tuite, David Fisman, and C. O. V. I. D.-19 Canada Open Data Working Group. Open access epidemiologic data and an interactive dashboard to monitor the COVID-19 outbreak in Canada. Canadian Medical Association Journal, 192:E420, April 2020. ISSN 1488-2329. doi: 10.1503/cmaj.75262. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiY21haiI7czo1OiJyZXNpZCI7czoxMToiMTkyLzE1L0U0MjAiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMi8wNy8xOC8yMDIyLjA3LjE4LjIyMjc3Njk1LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 28. O’Neill M. Sturrock S.L. et al. Berry, I. A sub-national real-time epidemiological and vaccination database for the covid-19 pandemic in canada. Sci Data, 2021. 29. Government of New Brunswick. News Releases. 2022, 2022. URL [https://www2.gnb.ca/content/gnb/en/corporate/promo/covid-19/news.html](https://www2.gnb.ca/content/gnb/en/corporate/promo/covid-19/news.html). 30. Nick Andrews, Julia Stowe, Freja Kirsebom, Samuel Toffa, Tim Rickeard, Eileen Gallagher, Charlotte Gower, Meaghan Kall, Natalie Groves, Anne-Marie O’Connell, David Simons, Paula B. Blomquist, Asad Zaidi, Sophie Nash, Nurin Iwani Binti Abdul Aziz, Simon Thelwall, Gavin Dabrera, Richard Myers, Gayatri Amirthalingam, Saheer Gharbia, Jeffrey C. Barrett, Richard Elson, Shamez N. Ladhani, Neil Ferguson, Maria Zambon, Colin N.J. Campbell, Kevin Brown, Susan Hopkins, Meera Chand, Mary Ramsay, and Jamie Lopez Bernal. Covid-19 vaccine effectiveness against the Omicron (B.1.1.529) variant. New England Journal of Medicine, 386(16):1532–1546, 2022. doi: 10.1056/NEJMoa2119451. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJMoa2119451&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=35249272&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 31. Jasmin Khateeb, Yuchong Li, and Haibo Zhang. Emerging SARS-CoV-2 variants of concern and potential intervention approaches. Critical Care, 25(1):1–8, 2021. doi: 10.1186/s13054-021-03662-x. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s13054-021-03469-w&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=33608030&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 32. Jamie Lopez Bernal, Nick Andrews, Charlotte Gower, Eileen Gallagher, Ruth Simmons, Simon Thelwall, Elise Tessier, Natalie Groves, Gavin Dabrera, Richard Myers, et al. Effectiveness of COVID-19 vaccines against the B. 1.617. 2 variant. medRxiv, 2021. doi: 10.1101/2021.05.22.21257658. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoibWVkcnhpdiI7czo1OiJyZXNpZCI7czoyMToiMjAyMS4wNS4yMi4yMTI1NzY1OHYxIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMDcvMTgvMjAyMi4wNy4xOC4yMjI3NzY5NS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 33. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 2022. URL [https://www.R-project.org/](https://www.R-project.org/). 34. Luca Ferretti, Chris Wymant, Michelle Kendall, Lele Zhao, Anel Nurtay, Lucie AbelerDörner, Michael Parker, David Bonsall, and Christophe Fraser. Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science, 368 (6491), 2020. doi: 10.1101/2020.03.08.20032946. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1101/2020.03.08.20032946&link_type=DOI) 35. Hellewell et al. Estimating the effectiveness of routine asymptomatic PCR testing at different frequencies for the detection of SARS-CoV-2 infections. BMC Medicine, 19, 2021. doi: 10.1186/s12916-021-01982-x. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12916-021-01982-x&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=33902581&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F07%2F18%2F2022.07.18.22277695.atom) 36. Julien Arino, Pierre-Yves Böelle, Evan M Milliken, and Stephanie Portet. Risk of COVID-19 variant importation – How useful are travel control measures? Infectious Disease Modelling, 6:875–897, 2021. doi: 10.1016/j.idm.2021.06.006. URL [https://doi.org/10.1016/j.idm.2021.06.006](https://doi.org/10.1016/j.idm.2021.06.006). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.idm.2021.06.006&link_type=DOI) 37. Arnaud Godin, Yiqing Xia, David L. Buckeridge, Sharmistha Mishra, Dirk Douwes-Schultz, Yannan Shen, Maxime Lavigne, Mélanie Drolet, Alexandra M. Schmidt, Marc Brisson, and Mathieu Maheu-Giroux. The role of case importation in explaining differences in early SARS-CoV-2 transmission dynamics in Canada - a mathematical modeling study of surveillance data. International Journal of Infectious Diseases, 102:254–259, January 2021. ISSN 1878-3511. doi: 10.1016/j.ijid.2020.10.046. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ijid.2020.10.046&link_type=DOI) 38. Xianghong Zhang, Yunna Song, Sanyi Tang, Haifeng Xue, Wanchun Chen, Lingling Qin, Shoushi Jia, Ying Shen, Shusen Zhao, and Huaiping Zhu. Models to assess imported cases on the rebound of COVID-19 and design a long-term border control strategy in Heilongjiang Province, China. Mathematical Biosciences and Engineering, 19:1–33, January 2022. ISSN 1551-0018. doi: 10.3934/mbe.2022001. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3934/mbe.2022001&link_type=DOI) 39. Maureen Johnson-León, Arthur L Caplan, Louise Kenny, Iain Buchan, Leah Fesi, Phoebe Olhava, Desmond Nsobila Alugnoa, Mara G Aspinall, Emily Costanza, Brianna Desharnais, et al. Executive summary: It’s wrong not to test: The case for universal, frequent rapid COVID-19 testing. EClinicalMedicine, 33, 2021. 40. S McPhee-Knowles, B Hoffman, and L. Kanary. The Yukon’s experience with COVID-19: Travel restrictions, variants and spread among the unvaccinated. Can Commun Dis Rep, 48:17–21, 2022. URL [https://doi.org/10.14745/ccdr.v48i01a03](https://doi.org/10.14745/ccdr.v48i01a03). 41. Sean C Anderson, Nicola Mulberry, Andrew M Edwards, Jessica E Stockdale, Sarafa A Iyaniwura, Rebeca C Falcao, Michael C Otterstatter, Naveed Z Janjua, Daniel Coombs, and Caroline Colijn. How much leeway is there to relax COVID-19 control measures? Epidemics, 35:100453, 2021. doi: 10.1016/j.epidem.2021.100453. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.epidem.2021.100453&link_type=DOI) 42. Luiz Henrique da Silva Nali, Felipe Scassi Salvador, Graciela dos Santos Soares Bonani, Heitor Franco de Andrade, Expedito José de Albuquerque Luna, and Dennis Minoru Fujita. Reopening borders: protocols for resuming travel during the COVID-19 pandemic, 2021. 43. Shi Chen, Qin Li, Song Gao, Yuhao Kang, and Xun Shi. State-specific projection of COVID-19 infection in the United States and evaluation of three major control measures. Scientific Reports, 10(1):1–9, 2020. doi: 10.1101/2020.04.03.20052720. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoibWVkcnhpdiI7czo1OiJyZXNpZCI7czoyMToiMjAyMC4wNC4wMy4yMDA1MjcyMHYyIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMDcvMTgvMjAyMi4wNy4xOC4yMjI3NzY5NS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 44. Chad R. Wells, Pratha Sah, Seyed M. Moghadas, Abhishek Pandey, Affan Shoukat, Yaning Wang, Zheng Wang, Lauren A. Meyers, Burton H. Singer, and Alison P. Galvani. Impact of international travel and border control measures on the global spread of the novel 2019 coronavirus outbreak. Proceedings of the National Academy of Sciences of the United States of America, 117:7504–7509, March 2020. ISSN 1091-6490. doi: 10.1073/pnas.2002616117. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMToiMTE3LzEzLzc1MDQiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMi8wNy8xOC8yMDIyLjA3LjE4LjIyMjc3Njk1LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 45. Karen Ann Grépin, Tsi-Lok Ho, Zhihan Liu, Summer Marion, Julianne Piper, Catherine Z Worsnop, and Kelley Lee. Evidence of the effectiveness of travel-related measures during the early phase of the COVID-19 pandemic: a rapid systematic review. BMJ Global Health, 6(3):e004537, 2021. doi: 10.1136/bmjgh-2020-004537. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NToiYm1qZ2giO3M6NToicmVzaWQiO3M6MTE6IjYvMy9lMDA0NTM3IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMDcvMTgvMjAyMi4wNy4xOC4yMjI3NzY5NS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 46. Finlay Campbell, Brett Archer, and Henry Laurenson-Schafer. Increased transmissibility and global spread of sars-cov-2 variants of concern as at june 2021. Eurosurveillance, 2021. URL [https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2021.26.24.2100509](https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2021.26.24.2100509). 47. E. Mathieu, H. Ritchie, and E. Ortiz-Ospina. A global database of covid-19 vaccinations. Nature Human Behaviour, 2021. [1]: /embed/inline-graphic-1.gif [2]: /embed/inline-graphic-2.gif [3]: /embed/graphic-4.gif [4]: /embed/graphic-5.gif [5]: /embed/inline-graphic-3.gif [6]: /embed/graphic-6.gif [7]: /embed/inline-graphic-4.gif [8]: /embed/inline-graphic-5.gif [9]: /embed/inline-graphic-6.gif [10]: /embed/graphic-7.gif [11]: F3/embed/inline-graphic-7.gif [12]: F3/embed/inline-graphic-8.gif [13]: F3/embed/inline-graphic-9.gif [14]: /embed/graphic-13.gif [15]: /embed/graphic-14.gif [16]: /embed/graphic-15.gif [17]: /embed/graphic-16.gif [18]: /embed/graphic-17.gif [19]: /embed/graphic-18.gif [20]: /embed/graphic-19.gif [21]: /embed/inline-graphic-10.gif [22]: /embed/inline-graphic-11.gif [23]: /embed/inline-graphic-12.gif [24]: /embed/graphic-20.gif [25]: /embed/inline-graphic-13.gif [26]: F6/embed/inline-graphic-14.gif [27]: /embed/graphic-22.gif [28]: /embed/graphic-23.gif [29]: /embed/graphic-24.gif