- JoNova - https://www.joannenova.com.au -

Victorian lockdown worked: it stopped community transmission of Covid

Science before politics

Some are claiming the Victorian lockdown was too late, saying the rate of spread was already suppressed before it started. But that misses the point that slowing flight arrivals was responsible for most of the suppression up to that point, but that wasn’t going to stop the rising cases of community transmission. To judge if lockdown works, we need to look at domestic spread.

The graph that matters are the new daily cases, and even more so, the graph of daily new cases due to of community transmission (below).

Victoria started a major lockdown on March 24th when schools were largely closed (except to essential workers) and only essential services were allowed to run. We see daily new cases peaked 11 3 days lateralmost exactly as expected (We expect a 12 day delay as seen in Japan, Germany, Italy, Spain, South Korea and Norway). The 12 day average expected lag comes from a five day average incubation and then a roughly seven day lag for new cases to get into breathing trouble and get tested as such. In Victoria, the case numbers were  dominated by incoming arrivals, so the timing of border closures determined the daily new cases peak.

*POST CORRECTION!: Ignore this graph which was shown here which didn’t show the peak at all (it has the wrong date scale). It was irrelevant. Instead look at the community transmission graph below.  h/t to Chris who emailed.

Most of the early spread was from foreign arrivals

Lockdown is aimed to slow community spread, not to stop the importation of cases from interstate. International flight arrivals had been asked to self-quarantine from mid March, but from March 29th it became mandatory and enforced. So people flying in before this date would have potentially been able to spread infections, and the lockdown would have slowed that. (For readers overseas, lockdowns here were never as strict as Spain or Italy).

The rate of community transmission spread slowed decisively off the exponential curve on April 11th  — 18 days later. Was that delayed because people with infections went home and shared the virus with people they lived with for one more week? If they had been put in hospital or isolated, would that growth have stopped after 12 days and not 18? I don’t know. Only detailed contact tracing would give us the answer.

Nonetheless, it’s clearly not true that lockdown was instigated too late to help.


Coronavirus, Lockdown, Covid19, Graph, Victoria, Community transmission.

Community transmission of Coronavirus in Victoria Australia  |   Source: Covid19data

It’s worth knowing that quarantine measures were ramping up throughout March

The Citymobility graph shows that Australians gradually slid into a full lockdown from March 11th – March 25th. Presumably the early Re of 11 days (mentioned below) was slower than other nations partly because of there were already measures in place, and partly due to the hot dry weather, and high Vit D levels at the end of summer.

Mobility in Australia, Lockdown. Sydney Melbourne.

Australians gradually slowed down from March 11 to March 25th.  |  NewDaily

The new Seeman et al study looks at the changes in the viral genome and considered cases up to April 14th, so it doesn’t include cases potentially spreading in early April that were not diagnosed til after April 15. It helps to trace some unknown contacts and clusters.

The paper’s authors say the social restrictions reduced the spread around three days after the Stage 3 restrictions came into action:

 The birth-death skyline model suggested a considerable change in Re around 27th March (CI: 23rd – 31st March). Prior to 27th March, the estimated Re was 1.63 (CI: 1.45 – 1.8) with a subsequent decrease to 0.48 (CI: 0.27 – 0.69) after this time (Figure 4). Our estimated Re prior to the 27th March implied an epidemic doubling time of 11 days (CI: 8.3 – 14.4 days) The sampling proportion parameter (the probability of successfully sequencing an infected case) after the identification of the first case in Victoria was estimated at 0.88

Looking at the genetic diversity of the virus suggests the rate of spread decreased after Stage 3 was introduced.

Third, prior to the implementation of enhanced (stage 3) restrictions in Victoria, we identified a large genomic cluster (the largest in our dataset, comprising 75 cases) associated with several social venues in metropolitan Melbourne. This finding demonstrates the propensity for chains of SARS-CoV-2 transmission throughout urban areas associated with leisure activities and provides additional justification for the unprecedented population-level social restrictions in our setting. Further genomic support for the effectiveness of social restrictions is provided by our phylodynamic analysis, which demonstrates a decrease in Re, after the introduction of stage 3 restrictions (including mandatory quarantine in hotels for overseas returnees), from 1.63 to 0.48.

Lockdowns work, but that doesn’t mean all types lockdowns are worth it

The debate needs to move on. Lockdowns slow the spread of airborne disease. This is not to say that everything about lockdowns is necessary or effective or that all measures should continue. There was hypocrisy, inconsistency, and contradiction writ large. The aim is always to minimize. We now know that golf and outdoor activities like beaches are a low risk activity. We now know that children are not major spreaders like they are in nearly every other respiratory disease. We didn’t know that then.

Hindsight is a wonderful thing but only the psychic, not the scientific, knew the answers in March.

The evidence that lockdowns slow transmission of disease is now overwhelming. In nearly every country, the effect of increased lockdown became evident in the daily new cases curve about 12 days after it was started. The rate of influenza and other respiratory disease has ALSO been reduced by an extraordinary 85% across 17 different countries. But countries with half-baked lockdowns, or lockdowns sabotaged by open borders, haven’t controlled the virus and are suffering longer outbreaks with more deaths and longer disruption — that’s Sweden, UK, US, Canada, and Brazil.

The question that matters is: which parts of lockdown are worth it?

In Victoria 60% of cases came from overseas, so quarantining flight arrivals made the biggest difference to the rate of growth. But contact with a confirmed case, and unknown (community spread) cases were starting to grow exponentially too, and were only a few weeks behind.

Because the flights were not quarantined til too late, the expensive Stage 2 restrictions were needed, and possibly the Stage 3 as well. The graph below is unusual — it’s marked from “first day of symptoms,” not from detection. There is an average of five days of asymptomatic incubation. So cases in this graph were seeded, on average, five days earlier. What matters is the drop in infections five days or so after new restrictions are brought in. There is a drop visible five days after Stage two was introduced, and another drop again five days after Stage 3 came in, but the numbers are statistically small. (Bad for maths, but good for Victoria).

Stage 1 included early school holidays — effectively a school closure, so it was quite a high level of restriction.

 From Seemann et al: The graph of “first day of symptoms”

Victoria, graph, covid infections, lockdown stages.

Epidemic curve of COVID-19 cases, by putative mode of acquisition, and implementation of key public health interventions, Victoria, Australia, 06 January – 14 April 2020. Cases were categorized as (i) travel overseas if reporting travel in the 14 days prior to symptom onset or (ii) contact with a confirmed case if no overseas travel reported and case contact occurred within the same time period. Cases are plotted by reported date of symptom onset, or if unknown, date of initial specimen collection.

Without the Stage 2 restrictions would community spread have continued at the rate of the orange arrow (like Sweden)? Or would they have continued to decline from the peak and reached zero but weeks later than they are currently headed for?

Definitions from the paper:

If the government didn’t bring in Stage 3, would the 3 -5 cases of community spread keep running, dragging on the Stage 2 lockdown for month after month, causing sporadic cluster outbreaks like the Cedar Meatworks outbreak where nearly 100 people were infected?

If Stage 2 restrictions reduced the rate of spread below 1.0 the virus would still track to zero, but it’s a slower decay. If Stage 3 reduced the Re (or Ro)  further, it makes the whole lockdown shorter.

There’s a cost both ways. Would it be cheaper to have avoided the Stage 3 restrictions, but paid the price with low level constant running outbreaks, with schools and factories being closed periodically, and with 20% of the population self-restricting, avoiding pubs, clubs and restaurants for months? Not to mention the burden on health workers living under constant PPE stressful conditions, some separated from their families, and facing a significant morbidity rate on an ongoing basis?

With the rest of the nation clearing out the virus, would it mean that other states would keep the borders shut and fly right over Victoria on the way to New Zealand? The high cost of Stage 3 may end up looking like a bargain if Australia does indeed clear the virus and free itself from all lockdowns (apart from quarantining international visitors, of course). Or perhaps with hindsight we could have avoided Stage 3, but we ran the risk (still unknown) of missing the chance to eliminate the virus before mid-winter.

And of course, if Vitamin D, Zinc and HCQ or Ivermectin improve the survival odds and the burden of disease, then the lockdown bought us the time to put our armor on.

The exponential growth meant more deaths for every day delayed, but decisions had to be made in the fog of unknown unknowns: How severe would the virus be; would there long term damage; were many cases asymptomatic; would immunity be lasting; how long would it take; how long would businesses survive; how fast will customers return; how much would customers change their spending even without mandatory quarantine; how soon would we find a treatment; does winter matter; do children spread it; how much is transmitted by air or surface or fecal, or via carriers with no symptoms?

The only thing we know for sure is that quarantining of international visitors should have started in mid February. As was suggested here at the time. If that had been done, none of these domestic restrictions would have been necessary. Not one.


Things worth knowing about Coronavirus:



Seemann et al (2020) Tracking the COVID-19 pandemic in Australia using genomics, doi: https://doi.org/10.1101/2020.05.12.20099929

Abbbreviations: for some reason the reproductive number is Re in this paper not Ro.

Deleted Graph Source: Covid19data

6.2 out of 10 based on 51 ratings