View Article PDF

One of the challenges with a new pandemic such as COVID-19 is how best to undertake surveillance. Good-quality surveillance is needed to maximise rapid disease control, eg, with case isolation and contact tracing to identify further cases and to quarantine contacts as shown by successful control in China.1,2 This surveillance and control capacity is particularly critical for nations that decide to eliminate community transmission entirely as New Zealand aimed to3 and has succeeded with (as per mid-July 2020 and using a definition from other modelling work on elimination for New Zealand).4 Most Australian States and Territories had also eliminated community transmission of COVID-19 by mid-July 2020, the exceptions being Victoria and New South Wales. Elimination status is also relevant to the following groupings of island jurisdictions, as per WHO data on 15 July 2020:5

1. Those jurisdictions which have avoided any COVID-19 cases at the time of writing (eg, via border controls), but which are still at risk if border controls fail. These mainly include island jurisdictions in the Pacific Ocean (eg, American Samoa, Cook Islands, Federated States of Micronesia, Kiribati, Marshall Islands, Nauru, Niue, Palau, Samoa, Solomon Islands, Tokelau, Tonga, Tuvalu and Vanuatu).

2. Those jurisdictions which have only experienced sporadic cases and appear (as per mid-July 2020) to have successfully contained spread. These include some islands in the Pacific (eg, Fiji).

3. Those jurisdictions which have had larger outbreaks of COVID-19, but have instituted tight controls and have declining numbers of new cases or no new cases for many weeks (eg, Taiwan).

A recent Australian study6 suggested that timely detection and management of community transmission of COVID-19 is feasible. This modelling study concluded that “testing for infection in primary care patients presenting with cough and fever is an efficient, effective and feasible strategy for the detection and elimination of transmission chains”. For example, when testing 9,000 people per week (per million population), the authors estimated that no cases of COVID-19 would be missed in some circumstances. This type of surveillance could therefore be relevant to identifying emergent or re-emergent SARS-CoV-2, the pandemic virus causing COVID-19.

Given this background, we aimed to determine the effectiveness of surveillance using testing for the SARS-CoV-2 virus to identify an outbreak arising from a single case of border control failure in a nation that is free of community transmission: New Zealand as per mid-July 2020.

Methods

To run pandemic spread scenarios for New Zealand, we used a stochastic SEIR type model with key compartments for: susceptible [S], exposed [E], infected [I] and recovered/removed [R]. The model is a stochastic version of CovidSIM, which was developed specifically for COVID-19 by two of the authors (http://covidsim.eu; version 1.1). Work has been published from version 1.0 of the deterministic version of the model,7,8 but in the Appendix we provide updated parameters and differential equations for version 1.1. The stochastic model was built in Pascal and 100,000 simulations were run for each set of parameter values.

The parameters were based on available publications and best estimates used in the published modelling work on COVID-19. Key components were: a single undetected infected case arriving in New Zealand via a border control failure, 80% of infected COVID-19 cases being symptomatic, 39.5% of cases seeking a medical consultation in primary care settings, and 4% of symptomatic cases being hospitalised (see Appendix Table 1 for further details). We assumed that the initial undetected case could be at any stage of infection—to cover both failures of managing quarantine at the border, but also failures around the management of non-quarantined workers such as air crew and ship crew. Scenarios considered different levels of transmission with the effective reproduction number (Re) of SARS-CoV-2 to be 1.5, 2.0, 2.5 and 3.0 (Appendix Table 1). Given some evidence for superspreading phenomena with this pandemic virus,9–11 we also considered scenarios where just 10% of the cases generated 10 times the number of secondary cases as the other cases.

Other scenarios considered the impact of 75% of symptomatic people seeking a medical consultation (eg, as the result of a potential media campaign); and another considered a possible school outbreak (eg, a border control failure involving a teacher or student returning from overseas). The assumptions for the latter involved: Re = 2.0, only 5% of symptomatic cases seek medical consultation, and only 0.5% being hospitalised.

For the detection of COVID-19 cases, we assume testing of 95% of cases of symptomatic cases of respiratory illness seeking medical attention in primary care and of hospitalised cases of respiratory illness. For parameterising the size of these two groups, we used official statistics and results from the Flutracking surveillance system used in New Zealand (Appendix Table 1). The sensitivity of the PCR diagnostic test (at 89%) was based on a meta-analysis (Appendix Table 1).

Results

For what we regard as the most plausible scenario with an Re of 2.0 (ie, where people are still practicing some modest level of reduced social contact and/or increased hygiene because of the pandemic in other countries), the results suggest that 50% of outbreaks from a single imported case would be detected in the period up to day 16 and 95% in the period up to day 36 (Table 1, Figure 1). At the time of detection (to day 36), there was an estimated median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day would be required, (1,120 per million people per day) for the whole New Zealand population. The vast majority of this testing (96%) would, however, be in primary care settings and the rest in hospitals.

Table 1: Modelled impacts by the time it takes to obtain at least one positive test result for SARS-CoV-2 infection arising from a border control failure where a single case enters the island nation of New Zealand (all results adjusted for lag times in reporting and obtaining test results, using 100,000 stochastic simulations for each parameter setting).

*From the time of the border control failure to the mean day of outbreak detection. This is around 5,600 tests per day for primary care and hospital sectors combined for the first four listed scenarios.
**Includes those in the latent phase, prodromal phase, asymptomatic infections and symptomatic infections (but not recovered or deceased cases).
***These higher levels of consultation seeking result in a proportionate increase in the tests performed.
#The assumed characteristics for this school outbreak involved: Re = 2.0, only 5% of symptomatic cases seeking medical consultation, and only 0.5% being hospitalised. The level of testing was as per the first four listed scenarios.

Figure 1: Probability of COVID-19 case detection after reintroduction of the infection (the different curves represent the results of 100,000 simulations each, using Re values from 1.5 to 3.0). Dotted lines refer to a scenario where 75% of symptomatic cases seek medical help. Dashed lines refer to scenarios which allow for superspreading events. The dashed-dotted line refers to an outbreak in a school (for further details on parameter settings, see Table 1 and text).

For all scenarios except for the school scenario, 95% of outbreaks were detected in less than six weeks after introduction. A higher value (71 days) was for the simulated school outbreak where medical consultations were assumed to be much less likely (due to symptoms in young people being typically milder). Increasing the extent by which symptomatic people seek medical consultations to the 75% level (up from that reported by Flutracking at 39.5%), would reduce the time to detection (eg, from 36 to 26 days for the Re = 2.0 scenario at the 95% probability level, Table 1).

When allowing for superspreading events, introductions less frequently lead to outbreaks (Table 1) and these outbreaks have a tendency to be detected earlier (Figure 1).

Discussion

This analysis indicates the challenges for a surveillance system designed to detect the re-emergence of SARS-CoV-2 transmission in a COVID-19-free nation with border controls. A very high level of testing of symptomatic people is typically required in primary care settings and hospitals to detect an outbreak within five weeks after a single border control failure (at least at the 95% level).

This relatively ideal testing level, at 5,580 tests per day, is somewhat higher than the levels in New Zealand in early May 2020 (ie, the seven-day rolling average at this time was around 4,200 tests per day,12 although this included some screening of asymptomatic people). It is even higher than the more recent 2,240 tests per day in early July 2020 (seven-day rolling average from 3 to 9 July). This lower level in July was even in the context of publicised “escapes” from border control facilities and it may drop further in the future with enhancements in quarantine facility security. Possibly there is a need for health authorities to regularly remind health professionals to keep offering testing since there is always some (albeit low) risk of quarantine failures, as some people may still excrete virus beyond the 14-day quarantine period.13,14 Work could also be done to research any barriers for getting testing (eg, transport issues to primary care, waiting times and perceptions around cost barriers). Research could also explore why Australia has achieved a higher cumulative level of testing (112,000 tests vs 87,500 tests per million by 8 July 2020 15), although some of this will be due to the ongoing transmission of disease in states such as Victoria (as per July 2020).

Despite the high level of testing required for this type of surveillance system, there are potential ways that might improve the yield and cost-effectiveness of such testing:

• Prioritising community testing for those with relevant symptoms (as per Ministry of Health criteria updated in June 2020 16) in the cities where border control failures are most likely to become evident (ie, those operating international airports and where isolation/quarantine facilities exist: Auckland, Hamilton, Rotorua, Wellington and Christchurch). Similarly, if cargo ship crews travelling from international ports are permitted shore leave in New Zealand in the future, then testing could be focused on these port cities.

• Pooling samples for PCR testing may preserve reagents and be more efficient17 and cost-effective,18 but needs to be balanced against potential loss of sensitivity and associated diagnostic delays.

If it became difficult to maintain high levels of testing even in these priority cities, an additional safeguard might be routinely offering testing to all hospital and emergency department attendees with any respiratory symptom (ie, not just those in the Ministry guidelines16). Another safeguard would be enhancements to the contact tracing systems used in New Zealand so that they can effectively address any outbreaks that arise.

Study strengths and limitations

This is the first such modelling analysis for a country that has achieved an elimination goal for COVID-19 with the end of all community transmission. Nevertheless, this work could have been refined further by a focus on a narrower range of acute respiratory diseases (eg, excluding the category of hospital admissions for chronic lower respiratory diseases (ICD10 codes: J40–J47). But since hospital admissions for these often involve an acute aspect, eg, acute bronchitis on top of chronic obstructive respiratory disease, we took the parsimonious approach of considering all respiratory diseases. Another limitation is that we did not consider the relatively large seasonal fluctuations in the proportion of people consulting primary care for respiratory conditions (ie, with Flutracking data indicating a four-fold variation in cough/fever symptoms between May and October19).

This analysis also did not explore other surveillance options such as routine active surveillance of specific groups who might be considered at increased risk (eg, air-crew, ship-crew, port workers and quarantine facility workers). Similarly, not considered was the testing of town and city sewerage systems for the pandemic virus in wastewater, as is being explored in several jurisdictions internationally.20,21 Indeed, in the New Zealand setting, the Crown Research Institute ESR has reported detecting SARS-CoV-2 in wastewater22 and is continuing to develop this methodology. Such approaches could improve the speed of early detection in the community and allow for lower routine levels of testing people with respiratory symptoms.

Conclusions

In conclusion, this model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation. But further work is required to improve on this type of analysis and to evaluate other potential surveillance system components, particularly the testing of wastewater in sewerage systems.

Appendix

Mathematical description of the CovidSIM model (version 1.1) and model parameters

Appendix Table 1: Input parameters used for modelling the potential spread of the COVID-19 pandemic with the stochastic version of CovidSIM (v1.1) with New Zealand as a case study.

Summary

Abstract

Aim

We aimed to determine the effectiveness of surveillance using testing for SARS-CoV-2 to identify an outbreak arising from a single case of border control failure in a country that has eliminated community transmission of COVID-19: New Zealand.

Method

A stochastic version of the SEIR model CovidSIM v1.1 designed specifically for COVID-19 was utilised. It was seeded with New Zealand population data and relevant parameters sourced from the New Zealand and international literature.

Results

For what we regard as the most plausible scenario with an effective reproduction number of 2.0, the results suggest that 95% of outbreaks from a single imported case would be detected in the period up to day 36 after introduction. At the time point of detection, there would be a median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day (1,120 tests per million people per day) for the New Zealand population would be required. The vast majority of this testing (96%) would be of symptomatic cases in primary care settings and the rest in hospitals.

Conclusion

This model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation that had previously eliminated COVID-19. Nevertheless, there are plausible strategies to enhance testing yield and cost-effectiveness and potential supplementary surveillance systems such as the testing of town/city sewerage systems for the pandemic virus.

Author Information

Nick Wilson, BODE3 Programme, University of Otago, Wellington; HEIRU, University of Otago, Wellington; Markus Schwehm, ExploSYS GmbH, Germany; Ayesha J Verrall, Department of Pathology and Molecular Medicine, University of Otago, Wellington; Matthew Parry, Department of Mathematics & Statistics, University of Otago, Dunedin; Michael G Baker, HEIRU, University of Otago, Wellington; Martin Eichner, University of Tübingen, Germany; Epimos GmbH, Germany.

Acknowledgements

Dr Schwehm is supported by the University of Tübingen and the IMAAC-NEXT Association. Professor Wilson is supported by the New Zealand Health Research Council (Grant 16/443) and Ministry of Business Innovation and Employment (MBIE) funding of the BODE3 Programme (Grant UOOX1406).

Correspondence

Prof Nick Wilson, Public Health, University of Otago, Wellington.

Correspondence Email

nick.wilson@otago.ac.nz

Competing Interests

Dr Verrall reports this paper was written in Dr Verrall's capacity as Senior Lecturer at the University of Otago, not in her capacity as a candidate for Parliament. The views in this paper are not necessarily the views of the New Zealand Labour Party.

1. Bi Q, Wu Y, Mei S, Ye C, Zou X, Zhang Z, Liu X, Wei L, Truelove SA, Zhang T, Gao W, Cheng C, Tang X, Wu X, Wu Y, Sun B, Huang S, Sun Y, Zhang J, Ma T, Lessler J, Feng T. Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their close contacts in Shenzhen, China: a retrospective cohort study. Lancet Infect Dis 2020;(E-publication 27 April) S1473-3099(20)30287-5. doi: 10.1016/S1473-3099(20)30287-5.

2. WHO-China Joint Mission. Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). 2020;(16-24 February). http://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf

3. Baker M, Kvalsvig A, Verrall A, Telfar-Barnard L, Wilson N. New Zealand’s elimination strategy for the COVID-19 pandemic and what is required to make it work. N Z Med J 2020; 133(1512):10–14.

4. Wilson N, Parry M, Verrall A, Baker M, Schwehm M, Eichner M. When can elimination of SARS-CoV-2 infection be assumed? Simulation modelling in a case study island nation. medRxiv 2020;(20 May). http://medrxiv.org/cgi/content/short/2020.05.16.20104240v1

5. World Health Organization. Coronavirus disease (COVID-19) Situation Report – 177. 2020; (15 July). http://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports

6. Lokuge K, Banks E, Davies S, Roberts L, Street T, O’Donovan D, Caleo G, Glass K. Exit strategies: optimising feasible surveillance for detection, elimination and ongoing prevention of COVID-19 community transmission. medRxiv 2020;(23 April). doi: http://doi.org/10.1101/2020.04.19.20071217

7. Wilson N, Telfar Barnard L, Kvalsig A, Verrall A, Baker M, Schwehm M. Modelling the potential health impact of the COVID-19 pandemic on a hypothetical European country. medRxiv 2020; (23 March). http://medrxiv.org/cgi/content/short/2020.03.20.20039776v1

8. Wilson N, Telfar Barnard L, Kvalsvig A, Baker M. Potential health impacts from the COVID-19 pandemic for New Zealand if eradication fails: Report to the NZ Ministry of Health. Wellington: University of Otago Wellington. 2020; http://www.health.govt.nz/system/files/documents/publications/report_for_moh_-_covid-19_pandemic_nz_final.pdf

9. Endo A, Abbott S, Kucharski AJ, Funk S, Centre for the Mathematical Modelling of Infectious Diseases COVID-19 Working Group. Estimating the overdispersion in COVID-19 transmission using outbreak sizes outside China. Wellcome Open Res 2020, 5:67 (http://doi.org/10.12688/wellcomeopenres.15842.1).

10. Miller D, Martin M, Harel N, Kustin T, Tirosh O, Meir M, Sorek N, Gefen-Halevi S, Amit S, Vorontsov O, Wolf D, Peretz A, Shemer-Avni Y, Roif-Kaminsky D, Kopelman N, Huppert A, Koelle K, Stern A. Full genome viral sequences inform patterns of SARS-CoV-2 spread into and within Israel. medRxiv 2020; (22 May). http://doi.org/10.1101/2020.05.21.20104521

11. Hamner L, Dubbel P, Capron I, Ross A, Jordan A, Lee J, Lynn J, Ball A, Narwal S, Russell S, Patrick D, Leibrand H. High SARS-CoV-2 attack rate following exposure at a choir practice - Skagit County, Washington, March 2020. MMWR 2020; 69:606–10.

12. Ministry of Health. Lab testing and capacity. New Zealand Ministry of Health. [Updated 3 May 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases#lab

13. Lauer SA, Grantz KH, Bi Q, Jones FK, Zheng Q, Meredith HR, Azman AS, Reich NG, Lessler J. The incubation period of coronavirus disease 2019 (COVID-19) from publicly reported confirmed cases: Estimation and application. Ann Intern Med 2020; (E-publication 10 March).

14. Wilson N, Baker M, Eichner M. Estimating the impact of control measures to prevent outbreaks of COVID-19 associated with air travel into a COVID-19-free country: A simulation modelling study. medRxiv 2020; (17 June). http://www.medrxiv.org/content/10.1101/2020.06.10.20127977v3

15. Ritchie H, Ortiz-Ospina E, Hasel J, Beltekian D, Mathieu E, Hasell J, Macdonald B, Giattino C, Roser M. Coronavirus Pandemic (COVID-19) – the data (Assessed 11 July 2020). http://ourworldindata.org/coronavirus-data?country=~AUS 2020.

16. Ministry of Health. Case definition of COVID-19 infection (updated 24 June 2020). http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-resources-health-professionals/case-definition-covid-19-infection

17. Mallapaty S. The mathematical strategy that could transform coronavirus testing. Nature 2020.

18. Hogan CA, Sahoo MK, Pinsky BA. Sample pooling as a strategy to detect community transmission of SARS-CoV-2. JAMA 2020; (E-publication 7 April).

19. Flutracking. NZ participant annual report 2018. http://info.flutracking.net/wp-content/uploads/2020/02/NZ-participant-annual-report-2018.pdf

20. Mallapaty S. How sewage could reveal true scale of coronavirus outbreak. Nature 2020;(3 April). http://www.nature.com/articles/d41586-020-00973-x

21. Lesté-Lasserre C. Coronavirus found in Paris sewage points to early warning system. Science 2020; (21 April) http://www.sciencemag.org/news/2020/04/coronavirus-found-paris-sewage-points-early-warning-system

22. ESR. Update on ESR’s wastewater testing for the COVID-19 virus. (Media Release 13 May 2020). http://www.esr.cri.nz/home/about-esr/media-releases/update-on-esrs-wastewater-testing-for-the-covid-19-virus/

23. Gillespie D. A general method for numerically simulating the stochastic time evolution of coupled chemical reactions. J Comput Phys 1976; 22:403–34.

24. Statistics New Zealand. Population (December 2019). Statistics New Zealand. http://www.stats.govt.nz/topics/population http://www.stats.govt.nz/topics/population

25. Gudbjartsson DF, Helgason A, Jonsson H, Magnusson OT, Melsted P, Norddahl GL, Saemundsdottir J, Sigurdsson A, Sulem P, Agustsdottir AB, Eiriksdottir B, Fridriksdottir R, Gardarsdottir EE, Georgsson G, Gretarsdottir OS, Gudmundsson KR, Gunnarsdottir TR, Gylfason A, Holm H, Jensson BO, Jonasdottir A, Jonsson F, Josefsdottir KS, Kristjansson T, Magnusdottir DN, le Roux L, Sigmundsdottir G, Sveinbjornsson G, Sveinsdottir KE, Sveinsdottir M, Thorarensen EA, Thorbjornsson B, Love A, Masson G, Jonsdottir I, Moller AD, Gudnason T, Kristinsson KG, Thorsteinsdottir U, Stefansson K. Spread of SARS-CoV-2 in the Icelandic Population. N Engl J Med 2020.

26. Luo L, Dan L, Xin-long L, Xian-bo W, Qin-long J, Jia-zhen Z, al. e. Modes of contact and risk of transmission in COVID-19 among close contacts. medRxiv 2020;(24 March). http://www.medrxiv.org/content/10.1101/2020.03.24.20042606v1

27. Ministry of Health. COVID-19 - current cases. [Data as of 7 April, 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases

28. Ferguson N, Laydon D, Nedjati-Gilani G, Imai N, Ainslie K, Baguelin M, Bhatia S, Boonyasiri A, Cucunubá Z, Cuomo-Dannenburg G, Dighe A, Dorigatti I, Fu H, Gaythorpe K, Green W, Hamlet A, Hinsley W, Okell L, van Elsland S, Thompson H, Verity R, Volz E, Wang H, Wang Y, Walker P, Winskill P, Whittaker C, Donnelly C, Riley S, Ghani A. Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. Imperial College 2020; (16 March):1–20.

29. Fink S. Worst-case estimates for U.S. coronavirus deaths. New York Times 2020;(Updated 14 March). http://www.nytimes.com/2020/03/13/us/coronavirus-deaths-estimate.html

30. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report – 46. 2020;(6 March). http://www.who.int/docs/default-source/coronaviruse/situation-reports/20200306-sitrep-46-covid-19.pdf?sfvrsn=96b04adf_4

31. Liu Y, Gayle AA, Wilder-Smith A, Rocklov J. The reproductive number of COVID-19 is higher compared to SARS coronavirus. J Travel Med 2020.

32. Group of Eight Australia. COVID-19 Roadmap to Recovery: A Report for the Nation. The Group of Eight Ltd. 2020. http://go8.edu.au/research/roadmap-to-recovery

33. Zou L, Ruan F, Huang M, Liang L, Huang H, Hong Z, Yu J, Kang M, Song Y, Xia J, Guo Q, Song T, He J, Yen HL, Peiris M, Wu J. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med 2020.

34. Mizumoto K, Kagaya K, Zarebski A, Chowell G. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. Euro Surveill 2020;25:pii=2000180. http://doi.org/10.2807/1560-7917

35. Read J, Bridgen J, Cummings D, Ho A, Jewell C. Novel coronavirus 2019-nCoV: early estimation of epidemiological parameters and epidemic predictions. MedRxiv 2020. doi: http://doi.org/10.1101/2020.01.23.20018549

36. Li R, Pei S, Chen B, Song Y, Zhang T, Yang W, Shaman J. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science 2020.

37. Wolfel R, Corman VM, Guggemos W, Seilmaier M, Zange S, Muller MA, Niemeyer D, Jones TC, Vollmar P, Rothe C, Hoelscher M, Bleicker T, Brunink S, Schneider J, Ehmann R, Zwirglmaier K, Drosten C, Wendtner C. Virological assessment of hospitalized patients with COVID-2019. Nature 2020.

38. Ministry of Health. Publicly funded hospital discharges – 1 July 2016 to 30 June 2017. http://www.health.govt.nz/publication/publicly-funded-hospital-discharges-1-july-2016-30-june-2017

39. Kim H, Hong H, Yoon SH. Diagnostic Performance of CT and Reverse Transcriptase-Polymerase Chain Reaction for Coronavirus Disease 2019: A Meta-Analysis. Radiology 2020:201343.

40. Tian S, Hu N, Lou J, Chen K, Kang X, Xiang Z, Chen H, Wang D, Liu N, Liu D, Chen G, Zhang Y, Li D, Li J, Lian H, Niu S, Zhang L, Zhang J. Characteristics of COVID-19 infection in Beijing. J Infect 2020; 80:401–06.

41. Yang X, Yu Y, Xu J, Shu H, Xia J, Liu H, Wu Y, Zhang L, Yu Z, Fang M, Yu T, Wang Y, Pan S, Zou X, Yuan S, Shang Y. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med 2020;Published Online (21 February). http://doi.org/10.1016/S2213-2600(20)30079-5

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

One of the challenges with a new pandemic such as COVID-19 is how best to undertake surveillance. Good-quality surveillance is needed to maximise rapid disease control, eg, with case isolation and contact tracing to identify further cases and to quarantine contacts as shown by successful control in China.1,2 This surveillance and control capacity is particularly critical for nations that decide to eliminate community transmission entirely as New Zealand aimed to3 and has succeeded with (as per mid-July 2020 and using a definition from other modelling work on elimination for New Zealand).4 Most Australian States and Territories had also eliminated community transmission of COVID-19 by mid-July 2020, the exceptions being Victoria and New South Wales. Elimination status is also relevant to the following groupings of island jurisdictions, as per WHO data on 15 July 2020:5

1. Those jurisdictions which have avoided any COVID-19 cases at the time of writing (eg, via border controls), but which are still at risk if border controls fail. These mainly include island jurisdictions in the Pacific Ocean (eg, American Samoa, Cook Islands, Federated States of Micronesia, Kiribati, Marshall Islands, Nauru, Niue, Palau, Samoa, Solomon Islands, Tokelau, Tonga, Tuvalu and Vanuatu).

2. Those jurisdictions which have only experienced sporadic cases and appear (as per mid-July 2020) to have successfully contained spread. These include some islands in the Pacific (eg, Fiji).

3. Those jurisdictions which have had larger outbreaks of COVID-19, but have instituted tight controls and have declining numbers of new cases or no new cases for many weeks (eg, Taiwan).

A recent Australian study6 suggested that timely detection and management of community transmission of COVID-19 is feasible. This modelling study concluded that “testing for infection in primary care patients presenting with cough and fever is an efficient, effective and feasible strategy for the detection and elimination of transmission chains”. For example, when testing 9,000 people per week (per million population), the authors estimated that no cases of COVID-19 would be missed in some circumstances. This type of surveillance could therefore be relevant to identifying emergent or re-emergent SARS-CoV-2, the pandemic virus causing COVID-19.

Given this background, we aimed to determine the effectiveness of surveillance using testing for the SARS-CoV-2 virus to identify an outbreak arising from a single case of border control failure in a nation that is free of community transmission: New Zealand as per mid-July 2020.

Methods

To run pandemic spread scenarios for New Zealand, we used a stochastic SEIR type model with key compartments for: susceptible [S], exposed [E], infected [I] and recovered/removed [R]. The model is a stochastic version of CovidSIM, which was developed specifically for COVID-19 by two of the authors (http://covidsim.eu; version 1.1). Work has been published from version 1.0 of the deterministic version of the model,7,8 but in the Appendix we provide updated parameters and differential equations for version 1.1. The stochastic model was built in Pascal and 100,000 simulations were run for each set of parameter values.

The parameters were based on available publications and best estimates used in the published modelling work on COVID-19. Key components were: a single undetected infected case arriving in New Zealand via a border control failure, 80% of infected COVID-19 cases being symptomatic, 39.5% of cases seeking a medical consultation in primary care settings, and 4% of symptomatic cases being hospitalised (see Appendix Table 1 for further details). We assumed that the initial undetected case could be at any stage of infection—to cover both failures of managing quarantine at the border, but also failures around the management of non-quarantined workers such as air crew and ship crew. Scenarios considered different levels of transmission with the effective reproduction number (Re) of SARS-CoV-2 to be 1.5, 2.0, 2.5 and 3.0 (Appendix Table 1). Given some evidence for superspreading phenomena with this pandemic virus,9–11 we also considered scenarios where just 10% of the cases generated 10 times the number of secondary cases as the other cases.

Other scenarios considered the impact of 75% of symptomatic people seeking a medical consultation (eg, as the result of a potential media campaign); and another considered a possible school outbreak (eg, a border control failure involving a teacher or student returning from overseas). The assumptions for the latter involved: Re = 2.0, only 5% of symptomatic cases seek medical consultation, and only 0.5% being hospitalised.

For the detection of COVID-19 cases, we assume testing of 95% of cases of symptomatic cases of respiratory illness seeking medical attention in primary care and of hospitalised cases of respiratory illness. For parameterising the size of these two groups, we used official statistics and results from the Flutracking surveillance system used in New Zealand (Appendix Table 1). The sensitivity of the PCR diagnostic test (at 89%) was based on a meta-analysis (Appendix Table 1).

Results

For what we regard as the most plausible scenario with an Re of 2.0 (ie, where people are still practicing some modest level of reduced social contact and/or increased hygiene because of the pandemic in other countries), the results suggest that 50% of outbreaks from a single imported case would be detected in the period up to day 16 and 95% in the period up to day 36 (Table 1, Figure 1). At the time of detection (to day 36), there was an estimated median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day would be required, (1,120 per million people per day) for the whole New Zealand population. The vast majority of this testing (96%) would, however, be in primary care settings and the rest in hospitals.

Table 1: Modelled impacts by the time it takes to obtain at least one positive test result for SARS-CoV-2 infection arising from a border control failure where a single case enters the island nation of New Zealand (all results adjusted for lag times in reporting and obtaining test results, using 100,000 stochastic simulations for each parameter setting).

*From the time of the border control failure to the mean day of outbreak detection. This is around 5,600 tests per day for primary care and hospital sectors combined for the first four listed scenarios.
**Includes those in the latent phase, prodromal phase, asymptomatic infections and symptomatic infections (but not recovered or deceased cases).
***These higher levels of consultation seeking result in a proportionate increase in the tests performed.
#The assumed characteristics for this school outbreak involved: Re = 2.0, only 5% of symptomatic cases seeking medical consultation, and only 0.5% being hospitalised. The level of testing was as per the first four listed scenarios.

Figure 1: Probability of COVID-19 case detection after reintroduction of the infection (the different curves represent the results of 100,000 simulations each, using Re values from 1.5 to 3.0). Dotted lines refer to a scenario where 75% of symptomatic cases seek medical help. Dashed lines refer to scenarios which allow for superspreading events. The dashed-dotted line refers to an outbreak in a school (for further details on parameter settings, see Table 1 and text).

For all scenarios except for the school scenario, 95% of outbreaks were detected in less than six weeks after introduction. A higher value (71 days) was for the simulated school outbreak where medical consultations were assumed to be much less likely (due to symptoms in young people being typically milder). Increasing the extent by which symptomatic people seek medical consultations to the 75% level (up from that reported by Flutracking at 39.5%), would reduce the time to detection (eg, from 36 to 26 days for the Re = 2.0 scenario at the 95% probability level, Table 1).

When allowing for superspreading events, introductions less frequently lead to outbreaks (Table 1) and these outbreaks have a tendency to be detected earlier (Figure 1).

Discussion

This analysis indicates the challenges for a surveillance system designed to detect the re-emergence of SARS-CoV-2 transmission in a COVID-19-free nation with border controls. A very high level of testing of symptomatic people is typically required in primary care settings and hospitals to detect an outbreak within five weeks after a single border control failure (at least at the 95% level).

This relatively ideal testing level, at 5,580 tests per day, is somewhat higher than the levels in New Zealand in early May 2020 (ie, the seven-day rolling average at this time was around 4,200 tests per day,12 although this included some screening of asymptomatic people). It is even higher than the more recent 2,240 tests per day in early July 2020 (seven-day rolling average from 3 to 9 July). This lower level in July was even in the context of publicised “escapes” from border control facilities and it may drop further in the future with enhancements in quarantine facility security. Possibly there is a need for health authorities to regularly remind health professionals to keep offering testing since there is always some (albeit low) risk of quarantine failures, as some people may still excrete virus beyond the 14-day quarantine period.13,14 Work could also be done to research any barriers for getting testing (eg, transport issues to primary care, waiting times and perceptions around cost barriers). Research could also explore why Australia has achieved a higher cumulative level of testing (112,000 tests vs 87,500 tests per million by 8 July 2020 15), although some of this will be due to the ongoing transmission of disease in states such as Victoria (as per July 2020).

Despite the high level of testing required for this type of surveillance system, there are potential ways that might improve the yield and cost-effectiveness of such testing:

• Prioritising community testing for those with relevant symptoms (as per Ministry of Health criteria updated in June 2020 16) in the cities where border control failures are most likely to become evident (ie, those operating international airports and where isolation/quarantine facilities exist: Auckland, Hamilton, Rotorua, Wellington and Christchurch). Similarly, if cargo ship crews travelling from international ports are permitted shore leave in New Zealand in the future, then testing could be focused on these port cities.

• Pooling samples for PCR testing may preserve reagents and be more efficient17 and cost-effective,18 but needs to be balanced against potential loss of sensitivity and associated diagnostic delays.

If it became difficult to maintain high levels of testing even in these priority cities, an additional safeguard might be routinely offering testing to all hospital and emergency department attendees with any respiratory symptom (ie, not just those in the Ministry guidelines16). Another safeguard would be enhancements to the contact tracing systems used in New Zealand so that they can effectively address any outbreaks that arise.

Study strengths and limitations

This is the first such modelling analysis for a country that has achieved an elimination goal for COVID-19 with the end of all community transmission. Nevertheless, this work could have been refined further by a focus on a narrower range of acute respiratory diseases (eg, excluding the category of hospital admissions for chronic lower respiratory diseases (ICD10 codes: J40–J47). But since hospital admissions for these often involve an acute aspect, eg, acute bronchitis on top of chronic obstructive respiratory disease, we took the parsimonious approach of considering all respiratory diseases. Another limitation is that we did not consider the relatively large seasonal fluctuations in the proportion of people consulting primary care for respiratory conditions (ie, with Flutracking data indicating a four-fold variation in cough/fever symptoms between May and October19).

This analysis also did not explore other surveillance options such as routine active surveillance of specific groups who might be considered at increased risk (eg, air-crew, ship-crew, port workers and quarantine facility workers). Similarly, not considered was the testing of town and city sewerage systems for the pandemic virus in wastewater, as is being explored in several jurisdictions internationally.20,21 Indeed, in the New Zealand setting, the Crown Research Institute ESR has reported detecting SARS-CoV-2 in wastewater22 and is continuing to develop this methodology. Such approaches could improve the speed of early detection in the community and allow for lower routine levels of testing people with respiratory symptoms.

Conclusions

In conclusion, this model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation. But further work is required to improve on this type of analysis and to evaluate other potential surveillance system components, particularly the testing of wastewater in sewerage systems.

Appendix

Mathematical description of the CovidSIM model (version 1.1) and model parameters

Appendix Table 1: Input parameters used for modelling the potential spread of the COVID-19 pandemic with the stochastic version of CovidSIM (v1.1) with New Zealand as a case study.

Summary

Abstract

Aim

We aimed to determine the effectiveness of surveillance using testing for SARS-CoV-2 to identify an outbreak arising from a single case of border control failure in a country that has eliminated community transmission of COVID-19: New Zealand.

Method

A stochastic version of the SEIR model CovidSIM v1.1 designed specifically for COVID-19 was utilised. It was seeded with New Zealand population data and relevant parameters sourced from the New Zealand and international literature.

Results

For what we regard as the most plausible scenario with an effective reproduction number of 2.0, the results suggest that 95% of outbreaks from a single imported case would be detected in the period up to day 36 after introduction. At the time point of detection, there would be a median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day (1,120 tests per million people per day) for the New Zealand population would be required. The vast majority of this testing (96%) would be of symptomatic cases in primary care settings and the rest in hospitals.

Conclusion

This model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation that had previously eliminated COVID-19. Nevertheless, there are plausible strategies to enhance testing yield and cost-effectiveness and potential supplementary surveillance systems such as the testing of town/city sewerage systems for the pandemic virus.

Author Information

Nick Wilson, BODE3 Programme, University of Otago, Wellington; HEIRU, University of Otago, Wellington; Markus Schwehm, ExploSYS GmbH, Germany; Ayesha J Verrall, Department of Pathology and Molecular Medicine, University of Otago, Wellington; Matthew Parry, Department of Mathematics & Statistics, University of Otago, Dunedin; Michael G Baker, HEIRU, University of Otago, Wellington; Martin Eichner, University of Tübingen, Germany; Epimos GmbH, Germany.

Acknowledgements

Dr Schwehm is supported by the University of Tübingen and the IMAAC-NEXT Association. Professor Wilson is supported by the New Zealand Health Research Council (Grant 16/443) and Ministry of Business Innovation and Employment (MBIE) funding of the BODE3 Programme (Grant UOOX1406).

Correspondence

Prof Nick Wilson, Public Health, University of Otago, Wellington.

Correspondence Email

nick.wilson@otago.ac.nz

Competing Interests

Dr Verrall reports this paper was written in Dr Verrall's capacity as Senior Lecturer at the University of Otago, not in her capacity as a candidate for Parliament. The views in this paper are not necessarily the views of the New Zealand Labour Party.

1. Bi Q, Wu Y, Mei S, Ye C, Zou X, Zhang Z, Liu X, Wei L, Truelove SA, Zhang T, Gao W, Cheng C, Tang X, Wu X, Wu Y, Sun B, Huang S, Sun Y, Zhang J, Ma T, Lessler J, Feng T. Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their close contacts in Shenzhen, China: a retrospective cohort study. Lancet Infect Dis 2020;(E-publication 27 April) S1473-3099(20)30287-5. doi: 10.1016/S1473-3099(20)30287-5.

2. WHO-China Joint Mission. Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). 2020;(16-24 February). http://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf

3. Baker M, Kvalsvig A, Verrall A, Telfar-Barnard L, Wilson N. New Zealand’s elimination strategy for the COVID-19 pandemic and what is required to make it work. N Z Med J 2020; 133(1512):10–14.

4. Wilson N, Parry M, Verrall A, Baker M, Schwehm M, Eichner M. When can elimination of SARS-CoV-2 infection be assumed? Simulation modelling in a case study island nation. medRxiv 2020;(20 May). http://medrxiv.org/cgi/content/short/2020.05.16.20104240v1

5. World Health Organization. Coronavirus disease (COVID-19) Situation Report – 177. 2020; (15 July). http://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports

6. Lokuge K, Banks E, Davies S, Roberts L, Street T, O’Donovan D, Caleo G, Glass K. Exit strategies: optimising feasible surveillance for detection, elimination and ongoing prevention of COVID-19 community transmission. medRxiv 2020;(23 April). doi: http://doi.org/10.1101/2020.04.19.20071217

7. Wilson N, Telfar Barnard L, Kvalsig A, Verrall A, Baker M, Schwehm M. Modelling the potential health impact of the COVID-19 pandemic on a hypothetical European country. medRxiv 2020; (23 March). http://medrxiv.org/cgi/content/short/2020.03.20.20039776v1

8. Wilson N, Telfar Barnard L, Kvalsvig A, Baker M. Potential health impacts from the COVID-19 pandemic for New Zealand if eradication fails: Report to the NZ Ministry of Health. Wellington: University of Otago Wellington. 2020; http://www.health.govt.nz/system/files/documents/publications/report_for_moh_-_covid-19_pandemic_nz_final.pdf

9. Endo A, Abbott S, Kucharski AJ, Funk S, Centre for the Mathematical Modelling of Infectious Diseases COVID-19 Working Group. Estimating the overdispersion in COVID-19 transmission using outbreak sizes outside China. Wellcome Open Res 2020, 5:67 (http://doi.org/10.12688/wellcomeopenres.15842.1).

10. Miller D, Martin M, Harel N, Kustin T, Tirosh O, Meir M, Sorek N, Gefen-Halevi S, Amit S, Vorontsov O, Wolf D, Peretz A, Shemer-Avni Y, Roif-Kaminsky D, Kopelman N, Huppert A, Koelle K, Stern A. Full genome viral sequences inform patterns of SARS-CoV-2 spread into and within Israel. medRxiv 2020; (22 May). http://doi.org/10.1101/2020.05.21.20104521

11. Hamner L, Dubbel P, Capron I, Ross A, Jordan A, Lee J, Lynn J, Ball A, Narwal S, Russell S, Patrick D, Leibrand H. High SARS-CoV-2 attack rate following exposure at a choir practice - Skagit County, Washington, March 2020. MMWR 2020; 69:606–10.

12. Ministry of Health. Lab testing and capacity. New Zealand Ministry of Health. [Updated 3 May 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases#lab

13. Lauer SA, Grantz KH, Bi Q, Jones FK, Zheng Q, Meredith HR, Azman AS, Reich NG, Lessler J. The incubation period of coronavirus disease 2019 (COVID-19) from publicly reported confirmed cases: Estimation and application. Ann Intern Med 2020; (E-publication 10 March).

14. Wilson N, Baker M, Eichner M. Estimating the impact of control measures to prevent outbreaks of COVID-19 associated with air travel into a COVID-19-free country: A simulation modelling study. medRxiv 2020; (17 June). http://www.medrxiv.org/content/10.1101/2020.06.10.20127977v3

15. Ritchie H, Ortiz-Ospina E, Hasel J, Beltekian D, Mathieu E, Hasell J, Macdonald B, Giattino C, Roser M. Coronavirus Pandemic (COVID-19) – the data (Assessed 11 July 2020). http://ourworldindata.org/coronavirus-data?country=~AUS 2020.

16. Ministry of Health. Case definition of COVID-19 infection (updated 24 June 2020). http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-resources-health-professionals/case-definition-covid-19-infection

17. Mallapaty S. The mathematical strategy that could transform coronavirus testing. Nature 2020.

18. Hogan CA, Sahoo MK, Pinsky BA. Sample pooling as a strategy to detect community transmission of SARS-CoV-2. JAMA 2020; (E-publication 7 April).

19. Flutracking. NZ participant annual report 2018. http://info.flutracking.net/wp-content/uploads/2020/02/NZ-participant-annual-report-2018.pdf

20. Mallapaty S. How sewage could reveal true scale of coronavirus outbreak. Nature 2020;(3 April). http://www.nature.com/articles/d41586-020-00973-x

21. Lesté-Lasserre C. Coronavirus found in Paris sewage points to early warning system. Science 2020; (21 April) http://www.sciencemag.org/news/2020/04/coronavirus-found-paris-sewage-points-early-warning-system

22. ESR. Update on ESR’s wastewater testing for the COVID-19 virus. (Media Release 13 May 2020). http://www.esr.cri.nz/home/about-esr/media-releases/update-on-esrs-wastewater-testing-for-the-covid-19-virus/

23. Gillespie D. A general method for numerically simulating the stochastic time evolution of coupled chemical reactions. J Comput Phys 1976; 22:403–34.

24. Statistics New Zealand. Population (December 2019). Statistics New Zealand. http://www.stats.govt.nz/topics/population http://www.stats.govt.nz/topics/population

25. Gudbjartsson DF, Helgason A, Jonsson H, Magnusson OT, Melsted P, Norddahl GL, Saemundsdottir J, Sigurdsson A, Sulem P, Agustsdottir AB, Eiriksdottir B, Fridriksdottir R, Gardarsdottir EE, Georgsson G, Gretarsdottir OS, Gudmundsson KR, Gunnarsdottir TR, Gylfason A, Holm H, Jensson BO, Jonasdottir A, Jonsson F, Josefsdottir KS, Kristjansson T, Magnusdottir DN, le Roux L, Sigmundsdottir G, Sveinbjornsson G, Sveinsdottir KE, Sveinsdottir M, Thorarensen EA, Thorbjornsson B, Love A, Masson G, Jonsdottir I, Moller AD, Gudnason T, Kristinsson KG, Thorsteinsdottir U, Stefansson K. Spread of SARS-CoV-2 in the Icelandic Population. N Engl J Med 2020.

26. Luo L, Dan L, Xin-long L, Xian-bo W, Qin-long J, Jia-zhen Z, al. e. Modes of contact and risk of transmission in COVID-19 among close contacts. medRxiv 2020;(24 March). http://www.medrxiv.org/content/10.1101/2020.03.24.20042606v1

27. Ministry of Health. COVID-19 - current cases. [Data as of 7 April, 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases

28. Ferguson N, Laydon D, Nedjati-Gilani G, Imai N, Ainslie K, Baguelin M, Bhatia S, Boonyasiri A, Cucunubá Z, Cuomo-Dannenburg G, Dighe A, Dorigatti I, Fu H, Gaythorpe K, Green W, Hamlet A, Hinsley W, Okell L, van Elsland S, Thompson H, Verity R, Volz E, Wang H, Wang Y, Walker P, Winskill P, Whittaker C, Donnelly C, Riley S, Ghani A. Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. Imperial College 2020; (16 March):1–20.

29. Fink S. Worst-case estimates for U.S. coronavirus deaths. New York Times 2020;(Updated 14 March). http://www.nytimes.com/2020/03/13/us/coronavirus-deaths-estimate.html

30. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report – 46. 2020;(6 March). http://www.who.int/docs/default-source/coronaviruse/situation-reports/20200306-sitrep-46-covid-19.pdf?sfvrsn=96b04adf_4

31. Liu Y, Gayle AA, Wilder-Smith A, Rocklov J. The reproductive number of COVID-19 is higher compared to SARS coronavirus. J Travel Med 2020.

32. Group of Eight Australia. COVID-19 Roadmap to Recovery: A Report for the Nation. The Group of Eight Ltd. 2020. http://go8.edu.au/research/roadmap-to-recovery

33. Zou L, Ruan F, Huang M, Liang L, Huang H, Hong Z, Yu J, Kang M, Song Y, Xia J, Guo Q, Song T, He J, Yen HL, Peiris M, Wu J. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med 2020.

34. Mizumoto K, Kagaya K, Zarebski A, Chowell G. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. Euro Surveill 2020;25:pii=2000180. http://doi.org/10.2807/1560-7917

35. Read J, Bridgen J, Cummings D, Ho A, Jewell C. Novel coronavirus 2019-nCoV: early estimation of epidemiological parameters and epidemic predictions. MedRxiv 2020. doi: http://doi.org/10.1101/2020.01.23.20018549

36. Li R, Pei S, Chen B, Song Y, Zhang T, Yang W, Shaman J. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science 2020.

37. Wolfel R, Corman VM, Guggemos W, Seilmaier M, Zange S, Muller MA, Niemeyer D, Jones TC, Vollmar P, Rothe C, Hoelscher M, Bleicker T, Brunink S, Schneider J, Ehmann R, Zwirglmaier K, Drosten C, Wendtner C. Virological assessment of hospitalized patients with COVID-2019. Nature 2020.

38. Ministry of Health. Publicly funded hospital discharges – 1 July 2016 to 30 June 2017. http://www.health.govt.nz/publication/publicly-funded-hospital-discharges-1-july-2016-30-june-2017

39. Kim H, Hong H, Yoon SH. Diagnostic Performance of CT and Reverse Transcriptase-Polymerase Chain Reaction for Coronavirus Disease 2019: A Meta-Analysis. Radiology 2020:201343.

40. Tian S, Hu N, Lou J, Chen K, Kang X, Xiang Z, Chen H, Wang D, Liu N, Liu D, Chen G, Zhang Y, Li D, Li J, Lian H, Niu S, Zhang L, Zhang J. Characteristics of COVID-19 infection in Beijing. J Infect 2020; 80:401–06.

41. Yang X, Yu Y, Xu J, Shu H, Xia J, Liu H, Wu Y, Zhang L, Yu Z, Fang M, Yu T, Wang Y, Pan S, Zou X, Yuan S, Shang Y. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med 2020;Published Online (21 February). http://doi.org/10.1016/S2213-2600(20)30079-5

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

One of the challenges with a new pandemic such as COVID-19 is how best to undertake surveillance. Good-quality surveillance is needed to maximise rapid disease control, eg, with case isolation and contact tracing to identify further cases and to quarantine contacts as shown by successful control in China.1,2 This surveillance and control capacity is particularly critical for nations that decide to eliminate community transmission entirely as New Zealand aimed to3 and has succeeded with (as per mid-July 2020 and using a definition from other modelling work on elimination for New Zealand).4 Most Australian States and Territories had also eliminated community transmission of COVID-19 by mid-July 2020, the exceptions being Victoria and New South Wales. Elimination status is also relevant to the following groupings of island jurisdictions, as per WHO data on 15 July 2020:5

1. Those jurisdictions which have avoided any COVID-19 cases at the time of writing (eg, via border controls), but which are still at risk if border controls fail. These mainly include island jurisdictions in the Pacific Ocean (eg, American Samoa, Cook Islands, Federated States of Micronesia, Kiribati, Marshall Islands, Nauru, Niue, Palau, Samoa, Solomon Islands, Tokelau, Tonga, Tuvalu and Vanuatu).

2. Those jurisdictions which have only experienced sporadic cases and appear (as per mid-July 2020) to have successfully contained spread. These include some islands in the Pacific (eg, Fiji).

3. Those jurisdictions which have had larger outbreaks of COVID-19, but have instituted tight controls and have declining numbers of new cases or no new cases for many weeks (eg, Taiwan).

A recent Australian study6 suggested that timely detection and management of community transmission of COVID-19 is feasible. This modelling study concluded that “testing for infection in primary care patients presenting with cough and fever is an efficient, effective and feasible strategy for the detection and elimination of transmission chains”. For example, when testing 9,000 people per week (per million population), the authors estimated that no cases of COVID-19 would be missed in some circumstances. This type of surveillance could therefore be relevant to identifying emergent or re-emergent SARS-CoV-2, the pandemic virus causing COVID-19.

Given this background, we aimed to determine the effectiveness of surveillance using testing for the SARS-CoV-2 virus to identify an outbreak arising from a single case of border control failure in a nation that is free of community transmission: New Zealand as per mid-July 2020.

Methods

To run pandemic spread scenarios for New Zealand, we used a stochastic SEIR type model with key compartments for: susceptible [S], exposed [E], infected [I] and recovered/removed [R]. The model is a stochastic version of CovidSIM, which was developed specifically for COVID-19 by two of the authors (http://covidsim.eu; version 1.1). Work has been published from version 1.0 of the deterministic version of the model,7,8 but in the Appendix we provide updated parameters and differential equations for version 1.1. The stochastic model was built in Pascal and 100,000 simulations were run for each set of parameter values.

The parameters were based on available publications and best estimates used in the published modelling work on COVID-19. Key components were: a single undetected infected case arriving in New Zealand via a border control failure, 80% of infected COVID-19 cases being symptomatic, 39.5% of cases seeking a medical consultation in primary care settings, and 4% of symptomatic cases being hospitalised (see Appendix Table 1 for further details). We assumed that the initial undetected case could be at any stage of infection—to cover both failures of managing quarantine at the border, but also failures around the management of non-quarantined workers such as air crew and ship crew. Scenarios considered different levels of transmission with the effective reproduction number (Re) of SARS-CoV-2 to be 1.5, 2.0, 2.5 and 3.0 (Appendix Table 1). Given some evidence for superspreading phenomena with this pandemic virus,9–11 we also considered scenarios where just 10% of the cases generated 10 times the number of secondary cases as the other cases.

Other scenarios considered the impact of 75% of symptomatic people seeking a medical consultation (eg, as the result of a potential media campaign); and another considered a possible school outbreak (eg, a border control failure involving a teacher or student returning from overseas). The assumptions for the latter involved: Re = 2.0, only 5% of symptomatic cases seek medical consultation, and only 0.5% being hospitalised.

For the detection of COVID-19 cases, we assume testing of 95% of cases of symptomatic cases of respiratory illness seeking medical attention in primary care and of hospitalised cases of respiratory illness. For parameterising the size of these two groups, we used official statistics and results from the Flutracking surveillance system used in New Zealand (Appendix Table 1). The sensitivity of the PCR diagnostic test (at 89%) was based on a meta-analysis (Appendix Table 1).

Results

For what we regard as the most plausible scenario with an Re of 2.0 (ie, where people are still practicing some modest level of reduced social contact and/or increased hygiene because of the pandemic in other countries), the results suggest that 50% of outbreaks from a single imported case would be detected in the period up to day 16 and 95% in the period up to day 36 (Table 1, Figure 1). At the time of detection (to day 36), there was an estimated median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day would be required, (1,120 per million people per day) for the whole New Zealand population. The vast majority of this testing (96%) would, however, be in primary care settings and the rest in hospitals.

Table 1: Modelled impacts by the time it takes to obtain at least one positive test result for SARS-CoV-2 infection arising from a border control failure where a single case enters the island nation of New Zealand (all results adjusted for lag times in reporting and obtaining test results, using 100,000 stochastic simulations for each parameter setting).

*From the time of the border control failure to the mean day of outbreak detection. This is around 5,600 tests per day for primary care and hospital sectors combined for the first four listed scenarios.
**Includes those in the latent phase, prodromal phase, asymptomatic infections and symptomatic infections (but not recovered or deceased cases).
***These higher levels of consultation seeking result in a proportionate increase in the tests performed.
#The assumed characteristics for this school outbreak involved: Re = 2.0, only 5% of symptomatic cases seeking medical consultation, and only 0.5% being hospitalised. The level of testing was as per the first four listed scenarios.

Figure 1: Probability of COVID-19 case detection after reintroduction of the infection (the different curves represent the results of 100,000 simulations each, using Re values from 1.5 to 3.0). Dotted lines refer to a scenario where 75% of symptomatic cases seek medical help. Dashed lines refer to scenarios which allow for superspreading events. The dashed-dotted line refers to an outbreak in a school (for further details on parameter settings, see Table 1 and text).

For all scenarios except for the school scenario, 95% of outbreaks were detected in less than six weeks after introduction. A higher value (71 days) was for the simulated school outbreak where medical consultations were assumed to be much less likely (due to symptoms in young people being typically milder). Increasing the extent by which symptomatic people seek medical consultations to the 75% level (up from that reported by Flutracking at 39.5%), would reduce the time to detection (eg, from 36 to 26 days for the Re = 2.0 scenario at the 95% probability level, Table 1).

When allowing for superspreading events, introductions less frequently lead to outbreaks (Table 1) and these outbreaks have a tendency to be detected earlier (Figure 1).

Discussion

This analysis indicates the challenges for a surveillance system designed to detect the re-emergence of SARS-CoV-2 transmission in a COVID-19-free nation with border controls. A very high level of testing of symptomatic people is typically required in primary care settings and hospitals to detect an outbreak within five weeks after a single border control failure (at least at the 95% level).

This relatively ideal testing level, at 5,580 tests per day, is somewhat higher than the levels in New Zealand in early May 2020 (ie, the seven-day rolling average at this time was around 4,200 tests per day,12 although this included some screening of asymptomatic people). It is even higher than the more recent 2,240 tests per day in early July 2020 (seven-day rolling average from 3 to 9 July). This lower level in July was even in the context of publicised “escapes” from border control facilities and it may drop further in the future with enhancements in quarantine facility security. Possibly there is a need for health authorities to regularly remind health professionals to keep offering testing since there is always some (albeit low) risk of quarantine failures, as some people may still excrete virus beyond the 14-day quarantine period.13,14 Work could also be done to research any barriers for getting testing (eg, transport issues to primary care, waiting times and perceptions around cost barriers). Research could also explore why Australia has achieved a higher cumulative level of testing (112,000 tests vs 87,500 tests per million by 8 July 2020 15), although some of this will be due to the ongoing transmission of disease in states such as Victoria (as per July 2020).

Despite the high level of testing required for this type of surveillance system, there are potential ways that might improve the yield and cost-effectiveness of such testing:

• Prioritising community testing for those with relevant symptoms (as per Ministry of Health criteria updated in June 2020 16) in the cities where border control failures are most likely to become evident (ie, those operating international airports and where isolation/quarantine facilities exist: Auckland, Hamilton, Rotorua, Wellington and Christchurch). Similarly, if cargo ship crews travelling from international ports are permitted shore leave in New Zealand in the future, then testing could be focused on these port cities.

• Pooling samples for PCR testing may preserve reagents and be more efficient17 and cost-effective,18 but needs to be balanced against potential loss of sensitivity and associated diagnostic delays.

If it became difficult to maintain high levels of testing even in these priority cities, an additional safeguard might be routinely offering testing to all hospital and emergency department attendees with any respiratory symptom (ie, not just those in the Ministry guidelines16). Another safeguard would be enhancements to the contact tracing systems used in New Zealand so that they can effectively address any outbreaks that arise.

Study strengths and limitations

This is the first such modelling analysis for a country that has achieved an elimination goal for COVID-19 with the end of all community transmission. Nevertheless, this work could have been refined further by a focus on a narrower range of acute respiratory diseases (eg, excluding the category of hospital admissions for chronic lower respiratory diseases (ICD10 codes: J40–J47). But since hospital admissions for these often involve an acute aspect, eg, acute bronchitis on top of chronic obstructive respiratory disease, we took the parsimonious approach of considering all respiratory diseases. Another limitation is that we did not consider the relatively large seasonal fluctuations in the proportion of people consulting primary care for respiratory conditions (ie, with Flutracking data indicating a four-fold variation in cough/fever symptoms between May and October19).

This analysis also did not explore other surveillance options such as routine active surveillance of specific groups who might be considered at increased risk (eg, air-crew, ship-crew, port workers and quarantine facility workers). Similarly, not considered was the testing of town and city sewerage systems for the pandemic virus in wastewater, as is being explored in several jurisdictions internationally.20,21 Indeed, in the New Zealand setting, the Crown Research Institute ESR has reported detecting SARS-CoV-2 in wastewater22 and is continuing to develop this methodology. Such approaches could improve the speed of early detection in the community and allow for lower routine levels of testing people with respiratory symptoms.

Conclusions

In conclusion, this model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation. But further work is required to improve on this type of analysis and to evaluate other potential surveillance system components, particularly the testing of wastewater in sewerage systems.

Appendix

Mathematical description of the CovidSIM model (version 1.1) and model parameters

Appendix Table 1: Input parameters used for modelling the potential spread of the COVID-19 pandemic with the stochastic version of CovidSIM (v1.1) with New Zealand as a case study.

Summary

Abstract

Aim

We aimed to determine the effectiveness of surveillance using testing for SARS-CoV-2 to identify an outbreak arising from a single case of border control failure in a country that has eliminated community transmission of COVID-19: New Zealand.

Method

A stochastic version of the SEIR model CovidSIM v1.1 designed specifically for COVID-19 was utilised. It was seeded with New Zealand population data and relevant parameters sourced from the New Zealand and international literature.

Results

For what we regard as the most plausible scenario with an effective reproduction number of 2.0, the results suggest that 95% of outbreaks from a single imported case would be detected in the period up to day 36 after introduction. At the time point of detection, there would be a median number of five infected cases in the community (95% range: 1–29). To achieve this level of detection, an ongoing programme of 5,580 tests per day (1,120 tests per million people per day) for the New Zealand population would be required. The vast majority of this testing (96%) would be of symptomatic cases in primary care settings and the rest in hospitals.

Conclusion

This model-based analysis suggests that a surveillance system with a very high level of routine testing is probably required to detect an emerging or re-emerging SARS-CoV-2 outbreak within five weeks of a border control failure in a nation that had previously eliminated COVID-19. Nevertheless, there are plausible strategies to enhance testing yield and cost-effectiveness and potential supplementary surveillance systems such as the testing of town/city sewerage systems for the pandemic virus.

Author Information

Nick Wilson, BODE3 Programme, University of Otago, Wellington; HEIRU, University of Otago, Wellington; Markus Schwehm, ExploSYS GmbH, Germany; Ayesha J Verrall, Department of Pathology and Molecular Medicine, University of Otago, Wellington; Matthew Parry, Department of Mathematics & Statistics, University of Otago, Dunedin; Michael G Baker, HEIRU, University of Otago, Wellington; Martin Eichner, University of Tübingen, Germany; Epimos GmbH, Germany.

Acknowledgements

Dr Schwehm is supported by the University of Tübingen and the IMAAC-NEXT Association. Professor Wilson is supported by the New Zealand Health Research Council (Grant 16/443) and Ministry of Business Innovation and Employment (MBIE) funding of the BODE3 Programme (Grant UOOX1406).

Correspondence

Prof Nick Wilson, Public Health, University of Otago, Wellington.

Correspondence Email

nick.wilson@otago.ac.nz

Competing Interests

Dr Verrall reports this paper was written in Dr Verrall's capacity as Senior Lecturer at the University of Otago, not in her capacity as a candidate for Parliament. The views in this paper are not necessarily the views of the New Zealand Labour Party.

1. Bi Q, Wu Y, Mei S, Ye C, Zou X, Zhang Z, Liu X, Wei L, Truelove SA, Zhang T, Gao W, Cheng C, Tang X, Wu X, Wu Y, Sun B, Huang S, Sun Y, Zhang J, Ma T, Lessler J, Feng T. Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their close contacts in Shenzhen, China: a retrospective cohort study. Lancet Infect Dis 2020;(E-publication 27 April) S1473-3099(20)30287-5. doi: 10.1016/S1473-3099(20)30287-5.

2. WHO-China Joint Mission. Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). 2020;(16-24 February). http://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf

3. Baker M, Kvalsvig A, Verrall A, Telfar-Barnard L, Wilson N. New Zealand’s elimination strategy for the COVID-19 pandemic and what is required to make it work. N Z Med J 2020; 133(1512):10–14.

4. Wilson N, Parry M, Verrall A, Baker M, Schwehm M, Eichner M. When can elimination of SARS-CoV-2 infection be assumed? Simulation modelling in a case study island nation. medRxiv 2020;(20 May). http://medrxiv.org/cgi/content/short/2020.05.16.20104240v1

5. World Health Organization. Coronavirus disease (COVID-19) Situation Report – 177. 2020; (15 July). http://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports

6. Lokuge K, Banks E, Davies S, Roberts L, Street T, O’Donovan D, Caleo G, Glass K. Exit strategies: optimising feasible surveillance for detection, elimination and ongoing prevention of COVID-19 community transmission. medRxiv 2020;(23 April). doi: http://doi.org/10.1101/2020.04.19.20071217

7. Wilson N, Telfar Barnard L, Kvalsig A, Verrall A, Baker M, Schwehm M. Modelling the potential health impact of the COVID-19 pandemic on a hypothetical European country. medRxiv 2020; (23 March). http://medrxiv.org/cgi/content/short/2020.03.20.20039776v1

8. Wilson N, Telfar Barnard L, Kvalsvig A, Baker M. Potential health impacts from the COVID-19 pandemic for New Zealand if eradication fails: Report to the NZ Ministry of Health. Wellington: University of Otago Wellington. 2020; http://www.health.govt.nz/system/files/documents/publications/report_for_moh_-_covid-19_pandemic_nz_final.pdf

9. Endo A, Abbott S, Kucharski AJ, Funk S, Centre for the Mathematical Modelling of Infectious Diseases COVID-19 Working Group. Estimating the overdispersion in COVID-19 transmission using outbreak sizes outside China. Wellcome Open Res 2020, 5:67 (http://doi.org/10.12688/wellcomeopenres.15842.1).

10. Miller D, Martin M, Harel N, Kustin T, Tirosh O, Meir M, Sorek N, Gefen-Halevi S, Amit S, Vorontsov O, Wolf D, Peretz A, Shemer-Avni Y, Roif-Kaminsky D, Kopelman N, Huppert A, Koelle K, Stern A. Full genome viral sequences inform patterns of SARS-CoV-2 spread into and within Israel. medRxiv 2020; (22 May). http://doi.org/10.1101/2020.05.21.20104521

11. Hamner L, Dubbel P, Capron I, Ross A, Jordan A, Lee J, Lynn J, Ball A, Narwal S, Russell S, Patrick D, Leibrand H. High SARS-CoV-2 attack rate following exposure at a choir practice - Skagit County, Washington, March 2020. MMWR 2020; 69:606–10.

12. Ministry of Health. Lab testing and capacity. New Zealand Ministry of Health. [Updated 3 May 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases#lab

13. Lauer SA, Grantz KH, Bi Q, Jones FK, Zheng Q, Meredith HR, Azman AS, Reich NG, Lessler J. The incubation period of coronavirus disease 2019 (COVID-19) from publicly reported confirmed cases: Estimation and application. Ann Intern Med 2020; (E-publication 10 March).

14. Wilson N, Baker M, Eichner M. Estimating the impact of control measures to prevent outbreaks of COVID-19 associated with air travel into a COVID-19-free country: A simulation modelling study. medRxiv 2020; (17 June). http://www.medrxiv.org/content/10.1101/2020.06.10.20127977v3

15. Ritchie H, Ortiz-Ospina E, Hasel J, Beltekian D, Mathieu E, Hasell J, Macdonald B, Giattino C, Roser M. Coronavirus Pandemic (COVID-19) – the data (Assessed 11 July 2020). http://ourworldindata.org/coronavirus-data?country=~AUS 2020.

16. Ministry of Health. Case definition of COVID-19 infection (updated 24 June 2020). http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-resources-health-professionals/case-definition-covid-19-infection

17. Mallapaty S. The mathematical strategy that could transform coronavirus testing. Nature 2020.

18. Hogan CA, Sahoo MK, Pinsky BA. Sample pooling as a strategy to detect community transmission of SARS-CoV-2. JAMA 2020; (E-publication 7 April).

19. Flutracking. NZ participant annual report 2018. http://info.flutracking.net/wp-content/uploads/2020/02/NZ-participant-annual-report-2018.pdf

20. Mallapaty S. How sewage could reveal true scale of coronavirus outbreak. Nature 2020;(3 April). http://www.nature.com/articles/d41586-020-00973-x

21. Lesté-Lasserre C. Coronavirus found in Paris sewage points to early warning system. Science 2020; (21 April) http://www.sciencemag.org/news/2020/04/coronavirus-found-paris-sewage-points-early-warning-system

22. ESR. Update on ESR’s wastewater testing for the COVID-19 virus. (Media Release 13 May 2020). http://www.esr.cri.nz/home/about-esr/media-releases/update-on-esrs-wastewater-testing-for-the-covid-19-virus/

23. Gillespie D. A general method for numerically simulating the stochastic time evolution of coupled chemical reactions. J Comput Phys 1976; 22:403–34.

24. Statistics New Zealand. Population (December 2019). Statistics New Zealand. http://www.stats.govt.nz/topics/population http://www.stats.govt.nz/topics/population

25. Gudbjartsson DF, Helgason A, Jonsson H, Magnusson OT, Melsted P, Norddahl GL, Saemundsdottir J, Sigurdsson A, Sulem P, Agustsdottir AB, Eiriksdottir B, Fridriksdottir R, Gardarsdottir EE, Georgsson G, Gretarsdottir OS, Gudmundsson KR, Gunnarsdottir TR, Gylfason A, Holm H, Jensson BO, Jonasdottir A, Jonsson F, Josefsdottir KS, Kristjansson T, Magnusdottir DN, le Roux L, Sigmundsdottir G, Sveinbjornsson G, Sveinsdottir KE, Sveinsdottir M, Thorarensen EA, Thorbjornsson B, Love A, Masson G, Jonsdottir I, Moller AD, Gudnason T, Kristinsson KG, Thorsteinsdottir U, Stefansson K. Spread of SARS-CoV-2 in the Icelandic Population. N Engl J Med 2020.

26. Luo L, Dan L, Xin-long L, Xian-bo W, Qin-long J, Jia-zhen Z, al. e. Modes of contact and risk of transmission in COVID-19 among close contacts. medRxiv 2020;(24 March). http://www.medrxiv.org/content/10.1101/2020.03.24.20042606v1

27. Ministry of Health. COVID-19 - current cases. [Data as of 7 April, 2020]. http://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases

28. Ferguson N, Laydon D, Nedjati-Gilani G, Imai N, Ainslie K, Baguelin M, Bhatia S, Boonyasiri A, Cucunubá Z, Cuomo-Dannenburg G, Dighe A, Dorigatti I, Fu H, Gaythorpe K, Green W, Hamlet A, Hinsley W, Okell L, van Elsland S, Thompson H, Verity R, Volz E, Wang H, Wang Y, Walker P, Winskill P, Whittaker C, Donnelly C, Riley S, Ghani A. Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. Imperial College 2020; (16 March):1–20.

29. Fink S. Worst-case estimates for U.S. coronavirus deaths. New York Times 2020;(Updated 14 March). http://www.nytimes.com/2020/03/13/us/coronavirus-deaths-estimate.html

30. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report – 46. 2020;(6 March). http://www.who.int/docs/default-source/coronaviruse/situation-reports/20200306-sitrep-46-covid-19.pdf?sfvrsn=96b04adf_4

31. Liu Y, Gayle AA, Wilder-Smith A, Rocklov J. The reproductive number of COVID-19 is higher compared to SARS coronavirus. J Travel Med 2020.

32. Group of Eight Australia. COVID-19 Roadmap to Recovery: A Report for the Nation. The Group of Eight Ltd. 2020. http://go8.edu.au/research/roadmap-to-recovery

33. Zou L, Ruan F, Huang M, Liang L, Huang H, Hong Z, Yu J, Kang M, Song Y, Xia J, Guo Q, Song T, He J, Yen HL, Peiris M, Wu J. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med 2020.

34. Mizumoto K, Kagaya K, Zarebski A, Chowell G. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. Euro Surveill 2020;25:pii=2000180. http://doi.org/10.2807/1560-7917

35. Read J, Bridgen J, Cummings D, Ho A, Jewell C. Novel coronavirus 2019-nCoV: early estimation of epidemiological parameters and epidemic predictions. MedRxiv 2020. doi: http://doi.org/10.1101/2020.01.23.20018549

36. Li R, Pei S, Chen B, Song Y, Zhang T, Yang W, Shaman J. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science 2020.

37. Wolfel R, Corman VM, Guggemos W, Seilmaier M, Zange S, Muller MA, Niemeyer D, Jones TC, Vollmar P, Rothe C, Hoelscher M, Bleicker T, Brunink S, Schneider J, Ehmann R, Zwirglmaier K, Drosten C, Wendtner C. Virological assessment of hospitalized patients with COVID-2019. Nature 2020.

38. Ministry of Health. Publicly funded hospital discharges – 1 July 2016 to 30 June 2017. http://www.health.govt.nz/publication/publicly-funded-hospital-discharges-1-july-2016-30-june-2017

39. Kim H, Hong H, Yoon SH. Diagnostic Performance of CT and Reverse Transcriptase-Polymerase Chain Reaction for Coronavirus Disease 2019: A Meta-Analysis. Radiology 2020:201343.

40. Tian S, Hu N, Lou J, Chen K, Kang X, Xiang Z, Chen H, Wang D, Liu N, Liu D, Chen G, Zhang Y, Li D, Li J, Lian H, Niu S, Zhang L, Zhang J. Characteristics of COVID-19 infection in Beijing. J Infect 2020; 80:401–06.

41. Yang X, Yu Y, Xu J, Shu H, Xia J, Liu H, Wu Y, Zhang L, Yu Z, Fang M, Yu T, Wang Y, Pan S, Zou X, Yuan S, Shang Y. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med 2020;Published Online (21 February). http://doi.org/10.1016/S2213-2600(20)30079-5

Contact diana@nzma.org.nz
for the PDF of this article

Subscriber Content

The full contents of this pages only available to subscribers.

LOGINSUBSCRIBE