I restarted the analyses from early July 2020 with the onset of the new Melbourne outbreak. The logic behind these charts is that they fill an information gap. Official data sources only give historic data series, and mainstream media typically only give near term predictions based on opinion.
Chart update 2 August 2020
What’s new?
Today’s new number of cases, 687 as per the Australian Government Department of Health 2/8/2020 update is higher than the projected estimate. It should be noted that case counts have been “lumpy” when considered at a daily level, bouncing up and down. However, when considered as a moving average and trend, it has been largely consistent with the model.
We are now over three weeks into the stage 3 restrictions of Melbourne, which started on 9 July 2020. Stage 4 restrictions were announced by Premier Andrews today, 2 August 2020. These restrictions will place much greater limits on social movement, including an overnight curfew. The move to stage 4 can be considered an acknowledgement that at a whole of system level, “stage 3” was insufficiently reducing transmission. This would be consistent with what we have seen with the model projections so far. If stage 4 restrictions reduce transmission, what we should see in the coming week is the model progressively over-estimating case counts with the “bending” of the curve.
My experience with this model from the March 2020 was that projections from early on in the epidemic tend to underestimate slightly in the short-term (days), and overestimate in the longer-term (weeks). This bias is something to keep in mind.
Projection of new daily cases of COVID-19 with data up to 2 August 2020
What is this?
The image is a chart of the confirmed daily new cases of COVID-19 in Australia, with a projection for the next 2 weeks. The projection is made using a model by fitting the data since 1 June 2020 to a Gompertz equation using non-linear regression. The dark blue dashed line is the model estimate. The grey dashed lines are the 95% prediction intervals, with the values given at 7 and 14 days into the future. The blue gradations can be understood as the degree of uncertainty in the model projections.
“Gompertz” equation?
The Gompertz function is a type of sigmoid, or “S”-shaped curve. It’s been around since the early 19th century and was initially used to describe and model demographic mortality curves, and hence, well known to actuaries. The Gompertz function can also be used to accurately model biological growth (e.g., epidemics, tumour size, enzymatic reactions). I have chosen to use this model to help with creating insights as earlier in the pandemic, it was found to be useful in modelling cumulative cases of COVID-19 from the Chinese outbreaks (Jia et al. arXiv:2003.05447v2 [q-bio.PE]). My experience from the initial outbreak from earlier in the year was that this equation gave reasonable descriptions of Australian and New Zealand data (for instance, NZ data below).
How have the model projections changed over the month?
The video demonstrates how the projections have evolved over time as new daily data have become available. This can give a better sense of where we are headed, given that the model cannot account for changes in context (e.g., policy changes, changes in testing rates, etc.)
My interpretation
According to the model, the peak in new cases is about two weeks away. I am concerned, and have noted for some days, at how “wide” the peak appears. An implication of this is that if transmission suppression is not improved, new case counts may take a long time to lower even after growth has plateaued. On a cumulative case chart, this would appear as a period of relatively linear growth.
This was the rationale given by Premier Andrews in his press release today.
“As they [health experts] tell us, based on the current numbers, cases might begin to drop off not in days or weeks – but in months. Months more of lockdown restrictions. Months more of 300, 400, 500 cases a day.”
This statement is compatible with the projections of the model. Based on data up to today, if the underlying pattern remains unchanged, we’ll still be having average daily new cases in the 300s in two months.
Note: I have little confidence in any projections beyond about two weeks, and medium- to longer-term projections must be taken as nothing more than an indicative glimpse of a possible future, and be interpreted with extreme caution.
I remain concerned with the daily number of cases in NSW at multiple locations across Sydney and the state. New daily numbers haven’t climbed in the past fortnight, but also haven’t whittled away. I think the phrase “on a knife’s edge” has been used to describe the situation in NSW. Widespread testing and contact tracing is taking place, along with improving physical distancing adherence and use of masks. Hopefully, it will be enough.
More information about the “peak” in new cases
What does it mean to have reached the peak in new cases? Assuming that our suppression of transmission doesn’t become MORE effective after the peak, it’s important to recognise that it is not the “halfway point”, which might be the intuition. The peak in the “new cases” curve corresponds to the “inflexion point” on the S-shaped cumulative cases curve (e.g., the first chart of the NZ cases in the brief description on the “Gompertz equation”. Roughly, the peak in new cases occurs at 40% of the total cumulative cases in an outbreak. That means that at the time we hit the peak, we can expect another one-and-a-half times the number of cases so far in the outbreak, before it ends. The insight is that we must resist the psychological temptation to relax transmission control mechanisms simply because we “crossed the peak”.
Want to know more?
Primary data source is the Australian Government Department of Health COVID-19 website for daily new cases. Analysis done using RStudio Cloud using R version 4.0.2.