A Lockdown Exit Pathway for New Zealand

How will New Zealand exit the current lockdown? It’s now 2 weeks into a nation-wide lockdown to control Covid-19. This unprecedented control of civil liberties has backing across the political spectrum. Overall, the New Zealand response appears to be very well managed.  It also comes at huge social and economic cost. At the media briefing on 9 April the Prime Minister, Jacinda Ardern, announced that cabinet will be deciding on 20 April how and when New Zealand exits the Level 4 lockdown. So what is the likely lockdown exit pathway and timing?

While this article concerns the situation in New Zealand,  many countries and states around the world find themselves in a similar position just now. The concepts discussed here have wider applicability.

Lockdown Objective

To forecast how the lockdown will end, we need to understand the objective.  On April 6 the Director-General of Health, Ashley Bloomfield, clarified that the aim of the lockdown is to “stamp out” the coronavirus.

Such a policy of elimination is consistent with the National Influenza Pandemic Plan. This would have been behind the thinking of  the “go early and go hard” lockdown strategy articulated by the Prime Minister, Jacinda Ardern, earlier in March when announcing border restrictions.

On April 6 Ms Ardern explained that the policy was the course of action that would have least overall impact on the health and welfare of New Zealanders.

Going hard and going early appears though to be paying off for us. The lockdown is the best way to stop the virus and it is also the best thing for our economy. By making the pain as short as possible.

Elimination is a bold strategy that isn’t usually attempted with influenza viruses. However, the success in containing the epidemic in China and several other Asian countries shows that it may be possible for the SARS-CoV-2 coronavirus that causes the Covid-19 disease. As a well-managed and relatively wealthy island nation with a small population, and with high trust in government, we have a chance of achieving this.

So how might this play out? To get a little ahead of where this article is leading to – going into lockdown was relatively straightforward. Deciding how and when to get out will be much harder.

Going into lockdown was straightforward. Getting out will be much harder.

I explain my rationale below.

Driving the Infection Reproduction Number Down

The infection reproduction number is the number of secondary infections generated from one infected individual. If the reproduction number is high then each infection leads to many more secondary infections. If the number is less than 1 then the epidemic dies out.

Stated in epidemiological terms, the purpose of a lockdown is to drive the reproduction number down over time. This slows the epidemic down and reduces strain on the health system. For New Zealand to stamp out the coronavirus, then driving the so-called time-varying reproduction number below 1 – and keeping it there – is essential.

Research from Imperial College in London has been very influential in Covid-19 policy decisions around the globe.  A research report published on 30 March examines the effects of various stages of lockdown on a range of European countries, working backwards from the observed deaths.  The modelling is heavily influenced by Italy and Spain as the epidemics were more advanced in those two countries and presented most Covid-19 deaths.  The figure below, shows when various control measures were introduced in the UK and the effect those have had on the time-varying reproduction number Rt.

Estimates of time-varying reproduction number for the UK. Dark green 50% confidence interval, light green 95% confidence interval. Icons are interventions shown at the time they occurred. Source: Estimating the number of infections and the impact of non-pharmaceutical interventions on Covid-19 in 11 European countries, Flaxman et al, Imperial College, 30 March 2020

There were two sobering findings from this research:

1. The observed fatalities indicate that the true size of the epidemic orders of magnitude greater than the reported numbers of confirmed cases
2. Only the most stringent lockdown controls succeed in driving the reproduction number towards 1.

Given New Zealand’s relatively high levels of testing for the size of our population and low positivity rate the first finding may not apply here at this time. However, the second implication has major ramifications for how we exit the lockdown. Simply put, we can’t exit the lockdown without the epidemic reigniting, unless:

• we succeed in eliminating the disease during the lockdown, or
• a vaccine becomes widely available, or
• a treatment that reduces the effects on patients becomes widely available – so that the epidemic can be allowed to run through most of the population, or
• a new development allows us to better manage the epidemic.

While there are hopes that treatments may prove effective and many trials are underway, stocks are likely to be firstly applied in the place of manufacture. So widespread availability of treatments may be some time off. A vaccine is likely to be at least a year away, so will likely play little role in the short to medium term lockdown exit pathway.

Is the Lockdown Working?

So, what should we expect to see in the lockdown if it’s working?  Given current understanding of a typical incubation period of 6-7 days between contact and showing symptoms of the coronavirus, new cases should continue to rise for a week or so then flatten off, then decline. The ideal is that new cases fall towards zero before the lockdown is called off.

In New Zealand case counts levelled out and are now declining. That’s so far so good, but we’re also halfway through the initially declared 4-week lockdown period and businesses need to be preparing for the lockdown exit pathway.

Test, Test, Test

Declining new case counts are against a backdrop of increased testing. The government and health authorities are to be commended for the rapid increase in testing capacity.   Ten labs can now test for Covid-19 in New Zealand (up from just a couple initially) which also helps to speed up the process and to mitigate risks. 

The increase in capacity has allowed the criteria for testing to be widened. The increased testing can be seen in the positivity rate in the tests. A week ago, there was one positive test for each 20 tests conducted. On April 9 the Director General of Health, Ashley Bloomfield, reported 29 new positives and just under over 4000 daily tests which equates to less than one positive per 100 tests. As observed by Dr Bloomfield, in the daily government Covid-19 briefing on April 5, a lower positivity rate is good news. More testing means that cases are more likely to be found. That creates opportunities to break the chain of transmission.

Positivity rates are not the only metric of testing success – the other is speed. All people who are presenting with symptoms and are tested are advised to self-isolate pending the results. A colleague who was tested in the week before lockdown received his test results 3 days later. If that was indicative of when the results are available to authorities, it was too slow. However on 7 April Newshub published comments from the Ministry of Health saying that the wait time to get results back to patients was then around 24 hours on average, though admitted that it may take longer in some instances – particularly for negative results. This week another colleague got his (negative) result back in 22 hours. This appears to confirm that the pace of reporting has sped up and may be continuing to quicken. When we’re dealing with fast-moving virus that’s a good thing.

Contact Tracing is Critical

A critical area of response is contact tracing. Countries who experienced the 2002 SARS epidemic first hand were much better prepared and have had a large measure of success in controlling Covid-19. South Korea and Singapore have quickly assigned resources and have aggressively followed up contacts. In Singapore detectives and the military were drafted into the fight in order to stand it up quickly. The BBC reports that contacts in Singapore are served written quarantine orders which formally advises what is required of them and the consequences if they don’t comply.

In contrast New Zealand has a system where contact tracing is primarily the responsibility of district health boards. Central government has recognised the need to improve contact tracing, describing it as being at the forefront of the fight against Covid-19.  At the April 5 daily media conference Dr Bloomfield stated that a National Close Contact Service with 190 staff had been set up on 24 March and can be expanded as required, reportedly tracing around 700 contacts per day. At the peak of the lockdown typically 80 new cases were confirmed per day. So the capacity boost, while welcome, was equivalent to just 10 additional contact traces per case.

Although it is supplementary to local DHB resources, which have also been enhanced, the additional contact tracing capability appears to be modest. That’s especially the case when the economic harm of lockdown – measured in billions of dollars – and the social harms – as evidenced by demand for Women’s Refuge – are taken into account. Effectively New Zealand as a nation is investing billions of dollars into this lockdown. For it to work, contact tracing needs to be thorough, complete and timely so we can exit lockdown as soon as possible.

That’s very hard to do manually with phone lines for a fast moving disease. Even the Singaporeans, who have heavily resourced this approach, have declared a month-long “circuit breaker” lockdown.

The Singapore experience shows that even the most diligent manual contact tracing by phones isn’t enough. You can’t beat this 2020 outbreak using 1918 technology.

You can’t beat this 2020 outbreak using 1918 technology.

Digital Contact Tracing

A feature of the New Zealand response to date has been a high initial reliance on trust.  One example is the government paying billions of dollars within days to New Zealand businesses on the basis of self-completed declarations. Another example is at the border, initially expecting tourists to follow the self-isolation rules, through to a Police “opt-in” location checking service. A faith in high-trust has allowed the authorities to move at speed. There will undoubtedly have been some gaming of the system. However speed is essential. Moving fast and fixing later has taken courage, but it’s the right approach.

In countries such as China, Singapore, Taiwan and South Korea, digital contact tracing using mobile phone data has been at the forefront of the control efforts. In essence there are two versions of this:

• opt in – the user chooses to share their location data and is warned instantly if they were recently near a person subsequently found to have Covid-19 for more than a specified time
• unconsented – the state uses people’s location data without consent.

The general consensus appears to be that wholesale unconsented use of citizen’s location data is unlikely to be acceptable in western cultures, even in the face of a pandemic. I’m less sure about that if the lockdown alternative effectively amounts to house arrest, but fortunately there are alternatives.  A research paper, just published in the journal Science suggests that 100% uptake isn’t necessary. The research investigated the effects of a digital contact tracing method illustrated in the figure below.

Contacts of individual A (and all individuals using the app) are traced using GPS co-localisations with other app users. Individual A requests a SARS-COV-2 test (using the app) and their positive test result triggers an instant notification to individuals who have been in close contact. The app advises isolation for the case (individual A) and their contacts.

This research drew on data from China, the Diamond Princess outbreak and other sources, accounting for the following transmission pathways,

1. Symptomatic – e.g. infected person sneezes droplets which are inhaled
2. Presymptomatic – infected person sheds virus before becoming aware of symptoms
3. Asymptomatic – infected person never develops symptoms
4. Environmental – indirect transmission through contaminated surfaces

The following graph from the research paper illustrates the challenge. What is immediately apparent is that quarantining symptomatic people (dark blue band in the figure) is not enough: contacts need to be isolated before they become infectious (light blue band).  While environmental and asymptomatic sources get a lot of media attention, this research indicates that they are less important. No doubt scientific understanding of the transmission pathways will improve over time, and reducing all transmission pathways is important, but this research points to where the priority should be from a public health perspective

Model of infectiousness. The average infectiousness (rate of infecting others), β, as a function of the amount of time since infection, τ. The total coloured area found between two values of τ is the number of transmissions expected in that time window. The total coloured area over all values of τ is the number of transmissions expected over the full course of one infection, i.e., the basic reproduction number R0. The different colours indicate the contributions of the four routes of transmission (stacked on top of one another), so that the total area of one colour over all values of τ is the average number of transmissions via that route over the whole course of infection. Stopping disease spread requires reduction of R to less than 1: blocking transmission, from whatever combination of colours and values of τ we can achieve, such that the total area is halved.

The research found that speed of contact tracing is critical. A delay of just three days between a person developing symptoms and isolation of contacts makes it practically impossible to drive the reproduction number below 1, even with near perfect systems.

The good news is that, if speed of contact tracing can be reduced to a near-zero lag time, then only around 60% of the infected cases need to be identified and only 60% of potential contacts need to opt in and self-isolate. Zero time lag could be achieved by treating symptomatic individuals as presumptive cases unless testing finds otherwise and isolating all contacts of presumptive cases. This is conceptually similar to what is currently being done in New Zealand, except that the contact tracing is done manually. Manual contact tracing introduces time lags and gaps due to contacts that are not traced.

It’s difficult to conceive that New Zealand will be completely free of Covid-19 before easing of the lockdown starts. The social and economic costs of lockdown are simply too great. Deploying such a digital contact tracing tool may be a necessary precondition to easing the lockdown, but it’s not likely to be enough for us all to leave the lockdown together.

Lockdown Exit Pathway

Stuff reports that the government is already using anonymised cellphone data, aggregated from telcos at the cell-tower level, to monitor the degree of public movement. At the media briefing on April 9 the Prime Minister announced that the Ministry of Health has a contact tracing app in development locally.  Standing up an app with appropriate levels of security in a quick time is very difficult.  The initial functionally is described by the Prime Minister as “basic”. At its simplest level such an app may be limited to contact details at first.

Jacinda Ardern also referred to a call scheduled for the evening of 9 April with the Prime Minister of Singapore, Mr Lee, regarding access to the Singaporean TraceTogether app. That app uses Bluetooth technology on mobile phones to identify potential contacts. The Government of Singapore has already announced that the app code will be made open-source.

If digital contact tracing is a key part of the way out of lock down, then it becomes possible to imagine a timetable where lockdown is eased before the disease is completely stamped out. A phased approach, encompassing such an app in the frontline, may commence with:

1. announce and roll out of basic app to limited sample of early adopters
2. shake down rapidly: identify and attend to any significant issues
3. roll out basic app to the general population with strong encouragement of mass take up
4. commence easing of lock down in parts of the economy,  along with increasing app functionality to enable faster and more complete contact tracing
5. close monitoring of outcomes and either retightening of lockdown or further easing of lockdown, depending on the outcomes

It’s easy to imagine 4 weeks from initial roll out of the app to commencing lockdown exit for those who can’t work from home. If correct, this forecast suggests that we will be in full lockdown going into May.

A wide range of variations are possible but this seems likely to be the overall outline. One thing seems likely – although we all went into lockdown together we may not all leave together.

Although we all went into lockdown together we may not all leave together. 

Leaving Lockdown Two-by-two

To date public discussions have been about moving between the Covid-19 alert levels as if we’re all going to do that together.  That is a simple and easily understood message and easy to police, and therefore the ideal outcome. But I think it’s unlikely.

My starting point is to observe that we’re not all in Level 4 lockdown, only most of us are. Those industries deemed to be essential have been allowed to carry on, albeit with workplace restrictions. That is a sensible and pragmatic response.  That the list of essential industries includes agriculture is is also pragmatic as the produce will rot if not harvested and people still need to eat. (I declare an interest in aquaculture which is part of the agriculture sector).

A recent paper by consultants at McKinsey, who advise governments around the world, gives an insight into the economic strategy advice that governments are receiving at this time. The McKinsey paper recommends a graduated reopening of the economy looking at factors such as how important the industry is economically, and how easy it is to keep workers safe from Covid-19. The generic diagram below illustrates the concept. In the New Zealand context we’re already seeing different thinking on which industries are considered critical, but the McKinsey conceptual model remains useful.

Sectors categorised by criticality and ability to safeguard. Source: Winning the (local) Covid-19 war, McKinsey 2020

Priority industries for restarting will be those deemed to be important to the national and regional economies. The practical ability to maintain a physical separation of 2 metres between workers will also important. The timber industry ticks all these boxes so expect that to be an early starter.

As suggested by a group of University of Otago epidemiologists, writing in a recent editorial for the New Zealand Medical Journal, all regions may not leave lockdown together. Building on this suggestion, it may be that regions free of recent cases leave the lockdown first, with movement between zones restricted to essential needs only.

An important factor will be how to maintain an easily understood framework so that interventions don’t become less effective as a whole. In comments recently reported in the New Zealand Herald, Professor Shaun Hendy, director of Te Pūnaha Matatini, New Zealand’s Centre of Research Excellence in Complex Systems and Data Analytics, said caution needed to be exercised because we are dealing with a complex system with a tipping point – “will we contain the outbreak or won’t we?”.

Given that epidemic health outcomes are worst for deprived segments of our society, regional deprivation will also be an important consideration in New Zealand. This suggests the we may see relaxation first in easily-defined rural regions that have a high reliance on primary industries such as forestry, and have been free of known cases of community transmission for a period of time.  Laying the  groundwork for such an approach have been hinted at in recent government media conferences – where a goal of gaining a regional understanding of the disease has been outlined by the Director General of Health.

However it won’t simply be a return to business as usual. The social, economic and political costs of failure are simply too great. Industries will be required to put in place and maintain physical distancing. Workplace Health and Safety legislation requires that employers take a precautionary approach. Specialist risk managers who understand health risks and safety will likely be called on to provide enterprises with the assurance they need.

When retail stores reopen we should initially expect social distancing requirements like those currently in place in supermarkets. The constant reminder of contagion risk will likely have a chilling effect on in-store consumer spending. 

In all of this, those who can apply social distancing with minimal disruption will be encouraged to do so. For those businesses who can work from home, calls to continue doing so are likely to continue for many months in all regions, possibly extending into 2021.  Some businesses will get used to the advantages of this mode of working and will never return completely. This acceleration of an already emerging trend will have long term impacts on demand for public transport, commercial offices and urban construction.

Social distancing to contain contagion risks is a numbers game. Mass gatherings in close contact such as church services, sports arenas, gyms, and cinemas might be able partially reopen with initially strict controls on numbers and separation.

The last controls to be lifted are likely to be local controls at the facility level on visitors where there are vulnerable patients, such as hospitals and residential aged care facilities. Such restrictions could run well into 2021.

Many variations are possible, and technologies will play an important role. Travellers should expect a return to needing medical passports showing our antibody and vaccination status, like our pets currently need to stay in kennels and catteries. At times this might apply between regions within New Zealand if there is a resurgence of localised outbreaks.

Successful elimination, combined with tight border controls, would allow a quicker and more collective lockdown exit. That’s why success of the current elimination effort is critical.

Your Views

Thanks to all of you who have posted comments from around the world on my earlier Covid-19 risk management articles.  It would be great to hear your comments and views on how this will likely play out. Just click on the button and comment below.

Limitations

These are my personal views as a safety and risk manager, drawing on information in the public domain.

Kevin Oldham is a director of Navigatus with a depth of experience in helping clients to achieve success. A brief profile for Kevin can be found here.

This article was updated on 9 April to incorporate the announcements made by the Prime Minister regarding developing a contact tracing app. Several links were updated overnight on 10 April and a new link to latest reporting on use of telco data added.