We are more than one hundred days into lockdown here in metro Melbourne — and it’s been going on even longer in the unlucky postcodes where the virus landed after the hotel quarantine breaches. That makes it a good time to look back at what we learned from the first wave to guide us through the second, and what more we’ve learned from the second lockdown, one of the longest and strictest anywhere in the world.
It’s important to bear in mind one striking characteristic of this virus: it spreads in clusters. Between 10 and 20 per cent of cases are believed to be responsible for 80 per cent of infections via “superspreader” events, which happen when a symptomatic person with a high viral load is coughing, sneezing or talking loudly in a group of people in a poorly ventilated environment. An estimated two-thirds or more of infected people, on the other hand, don’t pass on the virus at all.
Of the key metrics that epidemiologists employ to describe an epidemic, the one most commonly cited is the effective reproduction number, or R. As we all know by now, this number captures how many new people each case infects on average, and roughly corresponds to the relationship between the number of new cases today and the number we had one incubation period ago (the five to six days it takes for a person to become infected following exposure). A reproduction number of one means that each case infects one other person on average; above one, the outbreak will take off exponentially.
Uncontrolled, the R value for this coronavirus is somewhere between two and three, depending on the setting, which means the caseload more than doubles each five to six days. Once you have 200 cases, you will quickly get to 400 and then 800 within the space of just ten days. Get R below one, and the outbreak peters out.
We reduce R by putting in place interventions including the individual precautions of distancing, hygiene and masks, as well as population-level restrictions such as closing hospitality venues, retail outlets and other services deemed non-essential. Generally the advice on lockdowns is to go early and go hard; four to six weeks covers up to eight average incubation periods, and is seen as sufficient to close down community transmission while testing facilities are moved to surge capacity and workplaces prepare to operate in Covid-safe ways. Shorter, stricter lockdowns can work as circuit-breakers in established outbreaks when transmission rates have started to creep higher.
As a measure of the epidemic, the R number has the drawback of being an average. We now know that the number of cases one person infects can vary from zero to one hundred or more, but we also know that an average doesn’t tell the full story. So we use another metric as well: the dispersion factor, or k. This figure has not been routinely reported in Australia, but it is important for Covid-19 because it describes how much a disease clusters. The lower the k value, the smaller the proportion of cases responsible for transmissions. Flu, for example, has a higher k value than SARS-CoV-2 (the virus responsible for Covid-19) and spreads more evenly through the population if people aren’t immune. SARS and MERS outbreaks, also caused by coronaviruses, featured cluster transmission and in fact had even lower k numbers — and hence greater clustering — than the virus responsible for the current pandemic.
Like the reproduction number, k depends on local population characteristics (demographics, population density, number and size of high-risk transmission settings) and the interventions made by health authorities and governments. We didn’t know all this at the start of the pandemic, and the clustering transmission only became dramatically evident in wave two.
The beginning of Victoria’s second wave was buried in the tail of the first wave. Some of the new cases came as no surprise — authorities were still wrapping up the Cedar Meats outbreak, after all, and small numbers were continuing to appear, with the odd, refreshing zero-case day.
Health authorities didn’t test for community transmission in the first wave because they were anxious to focus testing capacity on high-risk returned travellers and their close contacts. But the serious cases that did appear in hospitals — the tip of the iceberg that was community transmission — made it clear there was a problem.
We may never know the true size of the first wave. Mercifully, though, with the big testing blitz in April failing to detect many more cases, we knew the lockdown worked. With measured steps, we managed to open up, having successfully eliminated all the community transmission caused by the first wave.
What’s critical about the hotel breach that sparked Victoria’s second wave is that a significant number of staff and guards were infected in quick succession, taking the virus home and into well-connected communities just as the government was easing the first-wave restrictions. In the third week of June, five additional hotel staff cases were reported in one day, a figure that doubled to ten and then to twenty in a single week, just as the second step out of restrictions commenced. The following week, total cases hit sixty a day; late the week after, 288 were recorded on one day.
Because authorities were increasingly testing symptomatic people (surveys suggest about half of Victorians with symptoms seek testing), this resurgence provided the data to build a more complete picture of the outbreak. We also gained a better understanding of transmission, with the sum of global evidence to date confirming that it mostly occurs via the respiratory route from droplets or, less frequently, aerosols. The evidence was also confirming that transmission usually occurs in clusters, most often when people are in close proximity indoors, with poor ventilation, for extended periods.
Melbourne’s second wave took off quickly because of the early superspreading, which also rapidly introduced the virus into a series of aged care facilities. The devastating acceleration of transmission in workplaces spread back into the workers’ homes and across workplaces or residential facilities. It’s likely that the early lockdown in the first wave prevented this, but it happened so quickly in the second wave — probably because of the demographics of those first local cases among hotel workers — that by the time the second lockdown took effect in wave two, it was too late.
Some commentators began arguing that Australia had opened too soon after wave one and that the resurgence in Victoria was evidence of a “failed suppression strategy.” This no doubt helped erode the public’s faith in the state government’s stage three lockdowns. This was one of the important lessons Victoria could have taken from wave one: that stage three for six weeks is effective in our setting, especially with the addition of masks. This insight was obscured by the unfolding story of what went wrong in hotel quarantine.
Despite the stage three restrictions having flattened the curve for a second time, and the addition of mandatory masks helping push the effective reproduction number below one, stage four was introduced just sixteen days later. Melbourne had already recorded 723 new cases in a single day, and a lack of confidence in stage three and an increasing concern about compliance pushed the setting to the next level of strict restrictions.
With lockdown tightening, and a night-time curfew added to very strict rules about how far and for how long individuals could travel from home, attention focused on the healthcare system’s capacity, and particularly the effectiveness of the contact-tracing process. Reassurances that it had sufficient capacity were countered by regular reports of clear failures. The concern was rising that lockdown would go on for much longer if the public health response was not up to the challenge.
Another sign that the authorities were losing control was the reversion to the most basic views about viral spread. Rather than a nuanced account drawn from the state’s experience and evidence from around the world, the premier’s daily press conferences increasingly focused on the “deadly,” “wildly infectious” virus rather than on the improving response.
Stage four added a series of time and distance constraints on leaving home, targeting the “aggregate movement” the government was starting to cite as the main indicator of risk. Rather than relying on people to follow straightforward rules — work from home, be conscious of the number of close contacts, don’t have visitors at home, take personal precautions and so on, all of which worked in pre-mask wave one — the second-wave restrictions were detailed and complicated.
In the process, an increasing number of the restrictions couldn’t clearly be explained in terms of risk or Covid-19 epidemiology. The curfew was one famous example, but other decisions also sought to micromanage behaviour in seemingly conflicting ways. Cricket nets could be used, for instance, but not tennis courts; a clinical Pilates class couldn’t be run with one instructor and three participants, but in the same space up to five clients could have sessions at the one time, each with his or her own instructor.
By step two of the second-wave restrictions, Melburnians could meet in groups of up to five people from two households in public, as long as they met within five kilometres of home. Many households had to split and take only a subgroup, but of course they then returned to their homes. Epidemiologists would be concerned about the number of households in a potential transmission network, not the number of people meeting in the park. On top of that, premier Daniel Andrews added that the rule was not restricted to the same household over time, effectively placing no limit on the households with which you could connect day to day, though this is where the potential for community transmission lay. Rules that could easily be enforced seemed to trump epidemiology.
When authorities start being prescriptive rather than providing high-level advice and broad principles, they confuse people and are then pressured into making even more specific rules to try to clarify the situation. Rules beget rules. People wait to be told exactly what they can and can’t do. They lose agency, and the government loses engagement.
Strict rules and large fines tied to enforcement also reinforce fear. People get the message that the second wave is somehow different, more “wildly infectious” than the first. They see neighbouring New South Wales managing to combat community transmission without restrictions and wonder how Victorians could be in this situation.
The coronavirus found the state’s capacities to be wanting in many respects. The health department had been pared back so severely that its starting point was behind those of other states. More importantly, though, the health authorities seemed to lack the capacity for the detailed analysis needed to inform and evaluate the components of the lockdown as it was unfolding.
Modelling was a feature of the “science” behind Victoria’s response, but it was communicated in a way that failed to instil confidence. The questions asked of the modellers focused on the implications of opening up further than the government was planning, and the high risk of a resurgence was then used to argue for stricter or longer restrictions. The time modelling was used to good effect, in my view, was when it evaluated risk among schoolchildren and led to an earlier-than-planned staged return to the classroom.
The aged care sector was not only woefully unprepared for the second wave, but authorities also took a long time to face up to fundamental challenges of workforce training, movement between facilities, and resident management. Lockdown played out in a different way in these settings, with many elderly residents confined to rooms and denied family access for extended periods. Aged care shortcomings were highlighted early in the second wave by both the unacceptable death toll and testimony to the royal commission into the sector. The case numbers, and especially the deaths, fed into the pressure for lockdown.
With part of the population now extremely concerned about the virus and/or our ability to contain it, and others unconvinced about the rationale behind the rules, public discussion — as opposed to assertion — largely closed down. The lockdown was politicised to a point where questioning any of the decisions, or the “science” behind them, could be met with threats of violence against the expert or journalist. In fact, divisiveness featured in the rules themselves, and was dominant in Premier Andrews’s commentary at his daily press conferences.
The five-kilometre rule and the “ring of steel” around Melbourne divided families, and divided customers and tourists from businesses and accommodation. Some services were unable to reopen because too many of their clients lived beyond the “ring” or outside a five-kilometre radius. Those on Melbourne’s fringes had fewer services within their five kilometres; a lucky few had a beach. Support services could open in step two, with indoor groups of up to twenty, but in regional Victoria twenty people could dine indoors at a restaurant while religious meeting places were closed and only ten could attend a prayer service outdoors. The former figure has now increased to forty, but the indoor–outdoor divide remains.
Importantly, some decisions ultimately determined which businesses and industries would survive. The loudest voices from different sectors can have more influence when the grounds for deciding what is and is not allowed become more fine-grained and arbitrary.
Regional Victorians were praised for their good behaviour and promised rewards. They were “fiercely protective” of their low-risk status, Melburnians were told, and didn’t want the metropolitan area opened because of the virus risk, even when case numbers fell to very low figures. As Melbourne was poised for step three, regional Victoria — which had already been granted additional leniency under step three — opened further; instead of returning to one state, the ring of steel was reinforced.
It will be a long time before we can see all the consequences of these extraordinary restrictions, much less measure them. The data collected on all 20,000 cases will provide rich evidence of transmission patterns, the effectiveness of different components of lockdown, and those mystery cases whose risk profiles we may be able to glean from the data.
Most of the information has been gathered via lengthy interviews with cases, which only moved from hard copy to direct computer entry surprisingly late in the piece. Artificial intelligence is now being employed to extract information from the data, but ideally this would be done at the time of interview for key exposures so that the epidemic dynamics could be monitored in real time and public health responses tailored appropriately.
Blanket restrictions were right for the first wave when we were still learning about this virus, its transmission routes and the consequences of infection. Somehow Victoria took a path that led to an even blunter set of restrictions in the second wave, and they didn’t work nearly as fast or as well. Public health benefits must always be weighed against the wider costs to health, economy and society.
In the end, the very failings of the second lockdown were used to justify more of the same. My main concern with these latest restrictions is that by September, with case numbers in low double digits, this lockdown was essentially being used to suppress transmission risk in the wider metro community so that, if public health responses failed, the collateral damage would be minimised.
By October, the health department had adopted the “contacts of contacts” approach to outbreak control that I had been advocating for months. Along with renewed efforts to find sources, which are critical in superspreader epidemics, rapid and comprehensive contact tracing promised to limit cases and end outbreaks sooner. In essence, it brings lockdown to the virus itself by asking all contacts of cases, and all of their contacts, to isolate until tests are completed and the extent of spread is determined. The second ring of contacts, if infected, are thus likely to be in isolation before they are even infectious, a game changer that removes the need for heavy state- or city-wide restrictions.
The biggest lesson from Victoria’s second lockdown is that everything must be done to prevent the need for another. We will no doubt see more cases, even if we get down to zero now. But the upgraded public health response we are told is in place — early-warning systems, workplace screening, sentinel surveillance, increased testing capacity, and a ramped-up, localised public health response — should prevent further waves and restrictions. Nonetheless, we must interrogate what happened, what we did well, and where there was more cost than benefit.
The final lessons of lockdown will be in the opening up. So far, the government’s extreme caution suggests a great fear of relying on the new system in the absence of strict lockdown. I trust this is a story of too much caution from our leaders rather than of concerns about the state’s capacity. •