With life in Melbourne, Victoria, having just dramatically changed for at least the next six weeks under a stricter lockdown, there are some aspects of a related press conference held by the Victorian Premier announcing changes to Melbourne’s restrictions (and associated statements released by the Premier) which appear to deserve some brief commentary.
I have no expertise in technical public health matters so I don’t wish to specifically comment on these aspects (as the Chair in Epidemiology at Deakin University has done so), but I do have some expertise regarding the production, appraisal, and use of model-based projections having studied this in some detail as part of my doctoral research and also having done related professional work. This aspect deserves analysis. Related to this, I’m interested in how decisions get justified and the roles and dynamics of prospective knowledge practices in these processes (e.g. the appraisal and use of model-based projections).
At the press conference the Victorian Premier, Daniel Andrews, referred to related advice he had received that if the then current lockdown strategy and measures were retained “it would likely be the end of the year before we would be able to reopen”, i.e. when cases would be at the required low level. He added in a separate media statement that many more months of lockdown could be required (if a major shift in strategy was not enacted) because of the likelihood of many “months more of 300, 400, 500 cases a day”.
Based on this, the Premier characterised the then current ‘Stage 3’ lockdown – and related rules and epidemic control policies in force (such mandatory masks, isolating the infected, etc) – as a “six-month strategy” which, in his view, “is simply not going to work.”
At the next press conference the following day he went even further and asserted the following: “There is simply no choice [i.e. no alternative to the stricter lockdown and wider cessation of activities being implemented]… this is the only way to get these numbers under control and drive them low enough […]. The alternative is a six month strategy […] and then even at that point significant doubt that that would work.”
Though he didn’t explicitly refer to modelling studies, I know that they are doing epidemic modelling (and commissioning it from public health experts) which, amongst other outputs, results in a series of future projections of how the epidemic wave could evolve over the short-term and long-term with (and without) different policy interventions.
The Premier’s words in the press conference indicate that he had received advice – most likely based in part on such forms of modelling – which projected future numbers of COVID-19 cases over the next 6 months if no additional interventions were made by the Victorian government. The related forecast of a potential future of many “months more of 300, 400, 500 cases a day” would likely also be informed by model-based projections.
What I find particularly intriguing is the Premier’s central contention that it “would likely be the end of the year” before case numbers got down to a sufficiently low level (to enable us to begin to reopen) if ‘Stage 3’ lockdown measures had been retained along with related public health measures (e.g. mandatory mask rules, isolating infected people, etc).
Our current epidemic wave began in early-ish June so the Premier is essentially suggesting that it would likely have been a seven-month long epidemic wave before it was sufficiently under control to enable us to begin reopening (if the ‘Stage 3’ lockdown measures had been retained along with related public health measures that are in place).
Now, a basic idea or heuristic that often is very important in the appraisal of model-based projections is to compare a projection with relevant empirical experience and the real-world related to what is being modelled. This empirical plausibility check can provide a sense of whether the results are robust or perhaps should be questioned.
And in the case of COVID-19, as a starting point we could consider whether a single epidemic wave in a comparable city has ever lasted seven months when a lockdown was introduced, meaning that the process of reopening would have to wait that long.
A complication is that different cities and countries have slightly different ‘lockdowns’ with different allowable and essential activities, and different local epidemics. So, this complicates possible comparisons and we need to be highly cautious. Nonetheless, a basic pattern appears to be common. Consider the following:
- In Germany, major cities like Berlin were in lockdown for around three months after which case numbers were low and reopening began
- New York began reopening three months after entering lockdown (and they had a far more intense epidemic than has to-date occurred in Melbourne)
- London’s lockdown was a bit longer, being in lockdown for about three and a half months before beginning to reopen on July 8
- Singapore’s recent second wave was brought under control after about three months
- Even somewhere like Stockholm which had a so-called “lockdown lite” approach, as per Sweden’s approach (with some restrictions), had rapidly dropping cases in late June and through July – about four months or so after their cases first rose and epidemic ‘containment’ measures were progressively introduced
It appears that about three months is a common timeframe which empirical experience thus far indicates is approximately the length of a lockdown (of various forms) before cases often get low enough for long enough for reopening process to potentially begin. Stockholm indicates a longer process would be needed without a formal lockdown.
If we consider a roughly three-month timeframe, here in Melbourne we might have only had about two further months of Stage 3 lockdown and related policies (e.g. mandatory mask wearing, other policies to contain outbreaks, etc) until cases came down sufficiently. This makes me wonder about the length of lockdown the Premier forecasted.
I can’t think of a single example of a place where lockdowns have had to be in place for as long as was being suggested by the Victorian Premier (if you can please enlighten me!), but no journalist at the press conference questioned him about the forecast underpinning the policy change. The critical interlocutor function of political journalists (as seen in, say, the Canberra press gallery) appears to have substantially disappeared under crisis conditions.
A related question is: how could the Premier have known that “it would likely be the end of the year before we would be able to reopen” (if the Stage 3 lockdown measures had been retained along with related public health measures)? The answer is he couldn’t know this because he doesn’t have a crystal ball. It will be interesting to see whether any of inputs informing recent decisions will ever be made public and available to scrutiny.