One concern I have about the modelling for coronavirus in Victoria is we don't know the assumptions behind the modelling. For instance, have they taken into account the different risk profiles and how that affects the model. For example healthcare workers are by the work they do, in a very risky situation. If healthcare workers and given the same weighting in the model as everyone else, the modelling will be affected.
In plain English, should the coronavirus cases for healthcare workers be included in the model. It all depends on the assumptions made by the people who create the model and no model is perfect.
So let's look at the situation that is happening in Victoria with coronavirus cases. The graph below shows the number of cases daily, the number of cases for healthcare workers and the rolling 14 day average for the daily and healthcare cases.
It is important to note the rolling averages will differ from the government's published rolling average as the I suspect the government adjusts the cases per day retrospectively as their data is updated and corrected. Otherwise the data that is published publicly can be used as a good guide.
The rolling average lines are the orange and the yellow lines. The healthcare cases is the grey line and the daily change in cases is the blue line.
As at the 14th of October the rolling average daily cases from this data is 10.1 and for Healthcare workers the figure is 3.4.
Over the past 30 days Healthcare workers have made up 36.5% of all cases, and at over a third of cases, that's a disproportionate percentage for a single group and should thus be treated differently in a model.
The reason it's important to reflect on the situation with Healthcare workers is because by their very nature these workers are known, work in known environments and are easily traced. Around 70% of healthcare workers infections have been traced to the work environment but the number could easily be higher. Healthcare workers are not a random sample of people. To weight healthcare workers the same as the general public could cause the state to remain in lockdown longer than need be.
The secondary affect of including healthcare workers with the same weighting in the model as the general public the healthcare workers will then also go on to infect their families, and perhaps to a lesser extent friends. Healthcare workers are very well educated and know the risks. Tracing of contacts of Healthcare workers you'd expect to be quick and these contacts would generally know what they need to do. This group by its nature is well informed and you would expect the larger group would work far more quickly than the general public to contain an outbreak. What we would expect to see is infected healthcare workers infecting their family but quick containment and thus low risk to the rest of the community. A model however would only see the numbers and this again may cause Victoria to remain in lockdown longer than needed since we're using a 14 day rolling average as a guide.
Healthcare workers are playing an important and valued roll in this epidemic and this post is in no way meant to disrespect healthcare workers. This post is about highlighting the importance of the model and if the model doesn't weight healthcare workers differently then all of Victoria continues to suffer in lockdown.
No comments:
Post a Comment