Abstract
Objectives To identify COVID-19 outbreaks and border control failures associated with quarantine systems in Australia and New Zealand and to estimate the failure risks.
Design, setting, participants Observational epidemiological study of travellers transiting quarantine in Australia and New Zealand up to 31 March 2021.
Main outcome measures The incidence of COVID-19 related failures arising from quarantine, and the failure risk for those transiting quarantine, estimated both per 100,000 travellers and per 1000 SARS-CoV-2 positive cases.
Results Australia and New Zealand had 24 COVID-19 related failures arising from quarantine systems up to 31 March 2021 (14 and 10, respectively). One resultant outbreak involved an estimated 800 deaths and quarantine failures instigated nine lockdowns. The failure risk for those transiting quarantine was estimated at 6.3 failures per 100,000 travellers and 5.8 failures (95%CI: 3.5 to 8.1) per 1000 SARS-CoV-2 positive cases. The latter risk was three-fold higher in New Zealand compared with Australia. Approximately 55% of quarantine system failures could likely have been prevented with the full vaccination of frontline border workers, for 70% effectiveness at preventing transmission.
Conclusions Quarantine system failures can be costly in terms of lives and economic impacts such as lockdowns. Ongoing improvements or alternatives to hotel-based quarantine are required.
Introduction
New Zealand and Australian states have successfully eliminated community transmission of the pandemic virus SARS-CoV-2,1 albeit with occasional outbreaks from imported cases that have typically been quickly brought under control. These two countries have primarily used hotel-based quarantine for citizens returning to their countries during the pandemic period, with 14 days of quarantine combined with PCR testing and mask use in any shared spaces (eg, common exercise areas used in New Zealand, but not in most Australian states).
Converting hotels for quarantine purposes has the advantage of making use of a resource that would otherwise be underused during a pandemic, given declines in international tourism. However, the major disadvantage of hotel-based quarantine is that it is likely to be less effective than purpose-built quarantine facilities owing to shared spaces and lack of safe ventilation (as per WHO advice on air flow2). Moreover, the consequences of leakage of the virus out of quarantine (eg, through facility workers) may be more severe given higher population density in urban settings where the hotels are located. Given these issues, we aimed to estimate the failure risk of quarantine systems in New Zealand and Australia in terms of the spread of COVID-19 infection into the community.
As of 31 March 2021, the rolling 7-day average number of COVID-19 vaccine doses administrated per 100 people was 0.18 in Australia and 0.08 in New Zealand3 However, this was counted as single vaccine doses and does not equal the total number of people vaccinated (eg, the Pfizer/BioNTech vaccine which is currently used in New Zealand requires two doses).3 The majority of border workers in Australia and New Zealand have been vaccinated (eg, over 56,000 doses had been administered to border workers as of 28 March,4 and all hotel quarantine workers in Victoria who have face-to-face contact with returned travellers received their first dose of the vaccine by the first week in April5).
Methods
We defined a quarantine system failure as where a border/health worker or person in the community with a link to the quarantine/isolation system, became infected with SARS-CoV-2. This definition included people infected in hospital from cases who had been transferred from a quarantine facility (as such cases were still in the 14-day quarantine process). But this definition did not include pandemic virus transmission between returnees within the quarantine facilities (as some other commentators have included in lists of failures in the Australian context6).
We searched official websites in both countries, and for the eight states and territories in Australia, to identify outbreaks and border control failures associated with quarantine systems (searches conducted between 6 January and 29 April 2021). Where an outbreak source was uncertain (eg, the Auckland, New Zealand, August 2020 outbreak) we used the best available evidence to classify it as a quarantine failure or not. We used two denominators: a) the estimated number of travellers who went through quarantine facilities during the 2020 year up to 31 March 2021; and b) the number of SARS-CoV-2 positive people who went through these facilities in this same time period. The unit of analyses were New Zealand, the eight Australian states and territories, and both countries combined.
For New Zealand, we used official data on both travellers going through the quarantine system,7 along with official (Ministry of Health) data on SARS-CoV-2 positive cases,8 although there are some discrepancies in the information about when regular testing began in Managed Isolation and Quarantine (MIQ) facilities. For Australia we used overseas arrival data9 and health data.10,11
Results
The collated data for quarantine system failures is shown in Table 1, with specific details of each event in the Appendix (Table A1). In Australia, 14 failures were identified, one causing over 800 deaths (Victoria’s second wave) and eight out of the 14 resulting in lockdowns. In New Zealand, there were ten failures, with one causing an outbreak with three deaths, and also a lockdown.
Identified quarantine system failures in Australia and New Zealand during the COVID-19 pandemic, the relevant denominator populations and estimated failure risks (with numerator and denominator data up to 31 March 2021)
Given our estimates of the number of travellers processed via quarantine systems (Table 1), the overall risks for both countries combined were one failure per 15,972 travellers, and one failure per 173 SARS-CoV-2 positive cases in quarantine. The combined data can also be interpreted as one outbreak leading to a lockdown response per 42,592 travellers; and approximately one death from COVID-19 per 477 travellers (using the 800 deaths estimate from Australia and the three deaths from New Zealand – although this figure is largely driven by the second wave in Victoria and is unlikely generalizable forward in time).
At the country level, there were 13.2 failures per 1000 SARS-CoV-2 positive cases transiting quarantine in New Zealand (95% confidence interval [CI]: 5.0 to 21.4), compared to 4.1 per 1000 SARS-CoV-2 positive cases in Australia (95%CI: 2.0 to 6.3) – a three-fold difference in risk (relative risk: 3.2, 95%CI: 1.4 to 7.2). Approximately 55.4% of quarantine system failures could have been preventable with the full vaccination of frontline border workers (Table A1) due to the combined effects of vaccination lowering the risk of getting infected by 70% or more (as in the case of the Moderna vaccine13), and the lesser duration of infectivity and lower peak infectivity for those infected.14
Discussion
This analysis identified 24 failures of quarantine systems in Australia and New Zealand combined (up to 31 March 2021). The significantly higher failure risk per 1000 SARS-CoV-2 positive cases transiting quarantine in New Zealand versus Australia could reflect a lower quality approach in the former, with perhaps some of the difference due to greater detection in New Zealand from more border worker testing over a longer period.
These estimates are subject to chance variations due to the low numbers of failures. These estimates will also probably be an underestimate of all quarantine breaches, as not all of those infected will transmit the virus and start a detectable chain of transmission. Genomes of the first 649 viral isolates collected in New Zealand show that only 19% of virus introductions resulted in ongoing transmission of more than one additional case.15 Therefore, counts of border failures are sensitive to how they are identified and defined. Indeed, with increased testing (eg, testing of people after leaving quarantine on day 16 as is now common in Australia), we may be detecting breaches that previously would have been undetected.
Looking forward, the failure risks per month in New Zealand and Australia may increase, given that the proportion of travellers returning to these countries who are infected is increasing due to global intensification of the pandemic and the increasing infectivity of new SARS-CoV-2 variants.16 Indeed, there have been several clearly documented cases of spread within quarantine hotels (eg, two instances in Melbourne in February 2021, two instances in Sydney in April 2021), highlighting the increased risk and evolving situation with more highly infectious variants arriving from overseas.
However, offsetting this trend will be measures such as the vaccination of quarantine workers. In New Zealand, the vaccination of border workers began in February 2021 with the Pfizer/BioNTech vaccine. However, vaccination does not fully protect against SARS-CoV-2 transmission, although a moderate degree of protection is likely. For example, infection rates were halved for the AstraZeneca vaccine,17,18 and perhaps reduced by 70% for the Moderna mRNA vaccine, as indicated by using swab results for asymptomatic infection plus symptomatic cases.13 For vaccinated people who are infected, primate study evidence suggests (consistent with expectation) that the infectivity is decreased in peak and duration,14 further protecting the border.
Furthermore, the level of testing of quarantine workers has been increasing (eg,19; which will find some failures before they have a chance to establish as an outbreak in the community). There have been other improvements in the quarantine systems over time (eg, improved security, introduction of mask wearing within quarantine settings, reduction in shared spaces, improved PPE used by workers, and other procedures as detailed in both countries20,21).
Another risk reduction practice would be using better or purpose-built facilities in rural locations as these have less risk from close contacts in central business district hotels and within-building spread from poor ventilation systems. Limitations of our analysis include residual uncertainty around the cause of some outbreaks (eg, the Auckland August 2020), and imprecision with denominator data on traveller numbers for Australia (eg, some travellers were moved between states on domestic flights which is not captured in the official data we used). Additionally, case numbers are constantly changing, due to the number of reclassifications caused by false positives and duplications.
To substantially reduce the risk of SARS-CoV-2 incursion out of quarantine (until such time as enough of the population is vaccinated), the most obvious action is to reduce arrivals, or even suspend arrivals, from high infection locations (as New Zealand temporarily did for travel from India and other high risk countries in April 202122). Beyond this, there are a range of other potential improvements in ongoing arrangements and processes as detailed in Table 2. Furthermore, the start of quarantine-free travel between Australia and New Zealand (also known as a “green zone”) in April 2021 provides an opportunity to benchmark COVID-19 border control policies and practices, identify potential improvements in both countries, and harmonise best practices across the region. The green zone further intertwines the biosecurity status of both nations and it is therefore even more important to lower the risk of border failures that could disrupt such travel. This shift from a one-size-fits-all strategy to a risk-based approach to border management can be summarised as a ‘traffic light’ approach.23
List of potential policy and operational options for improved COVID-19 control associated with quarantine systems in Australia and New Zealand, including measur es to reduce the numbers of infected people arriving into quarantine facilities
Conclusions
In summary, Australia and New Zealand have had 24 COVID-19 identified failures arising from quarantine systems up to 31 March 2021. Quarantine system failures can be costly in terms of lives and economic impacts such as lockdowns. Ongoing improvements or alternatives to hotel-based quarantine are required.
Data Availability
The datasets analyzed during the current study were derived from public resources available from the New Zealand Ministry of Business Innovation & Employment, the New Zealand Ministry of Health. the Australia Bureau of Statistics. and the Australian Department of Health.
https://www.health.gov.au/resources/collections/coronavirus-covid-19-common-operating-picture
Competing interests
Nil.
Funding
Prof Baker and Dr Kvalsvig received funding support from the Health Research Council of New Zealand (20/1066). Dr Grout, Ms Katar, Dr Ait Ouakrim, Dr Summers, Prof Blakely, and Prof Wilson did not have external funding support.
Appendix
List of COVID-19 border control failures associated with quarantine systems in Australia and New Zealand during 2020 and up to 31 March 2021
Approaches to managing quarantine in Australia and New Zealand
Table A2 details the various approaches to quarantine used in both countries. Key features of which are summarised below:
The length of quarantine in all jurisdictions is 14 days.
Almost all jurisdictions require at least two separate PCR tests for travellers: one shortly after arrival and another closer to the end of the quarantine period.
Many jurisdictions have introduced or are considering additional tests for travellers, either earlier in (eg, Day 0/1 in New Zealand) or after (eg, Day 16 in New South Wales) the quarantine period. This is in response to concerns about new highly infectious variants in early 2021.
Australia does not currently have a national strategy for quarantine.
Detailed information on quarantine programs for each Australian state/territory is limited and in many cases we had to rely on news articles for additional details in compiling Table A2. Descriptions of PPE required for quarantine staff in different jurisdictions were particularly difficult to identify and information often was not available by staff role (eg, security guards vs cleaners).
Description of key infection control processes and surveillance at quarantine facilities in Australia and New Zealand as of 31 March 2021 (with the length of quarantine being 14 days in all jurisdictions)
Footnotes
The data in this manuscript have been updated to include two additional quarantine failures in Australia that had previously been overlooked. Some of the text has been revised to reflect updated estimates and to clarify our definition of a quarantine system failure.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.
- 32.
- 33.
- 34.
- 35.
- 36.
- 37.
- 38.
- 39.
- 40.
- 41.
- 42.
- 43.
- 44.
- 45.
- 46.
- 47.
- 48.
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.
- 61.
- 62.
- 63.
- 64.
- 65.
- 66.
- 67.
- 68.
- 69.
- 70.
- 71.
- 72.
- 73.
- 74.
- 75.
- 76.
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.
- 87.
- 88.
- 89.
- 90.
- 91.
- 92.
- 93.
- 94.
- 95.
- 96.
- 97.
- 98.
- 99.
- 100.
- 101.
- 102.
- 103.
- 104.
- 105.
- 106.
- 107.
- 108.
- 109.
- 110.
- 111.
- 112.
- 113.
- 114.
- 115.
- 116.
- 117.
- 118.
- 119.
- 120.
- 121.
- 122.
