Showing posts with label modelling. Show all posts
Showing posts with label modelling. Show all posts

Friday, May 1, 2020

Covid19 Fail: models vs reality

On April 7 the Federal government released its epidemiological modelling of the Covid19 outbreak in Australia. The models were used to justify strict quarantine and isolation measures enforced by Federal and State governments that will see Australia experience its worst economic downturn since the Great Depression. ABC has thus far failed to provide any meaningful coverage of the models and their failure to match reality.

The rationale behind the government's "flatten the curve" strategy was to enforce strict physical distancing measures to ensure enough ICU beds were available to handle the caseload.

Under the government's model peak ICU demand with strict Level 3-4 Quarantine + isolation + social distancing measures in place was predicted to occur in week 43 (see graph below) which would be well into November with nearly 5000 ICU beds required, about double our capacity.

It seems reality played out quite differently.

The first case of Covid19 was reported in Australia on 25 January and around 14 weeks later it seems the "curve" has been well and truly flattened with only a handful of cases now reported daily.  We never reached more than 5% of ICU bed capacity with maximum daily use being around 100 beds way back in week 10 prior to the effect of level 3-4 measures being felt and the release of the Government's epidemiological modelling. The pandemic for us is well and truly over and we now face the social and economic consequences of a completely disproportionate response. Handling of the Covid19 pandemic by Australian Governments at all levels has been perhaps the worst example of Risk Management in history.

The ABC costs us well over $1 billion annually, we could fund 15000 nurses for that figure, yet somehow it lacks the integrity to take the government to task over models that were effectively disproven on the day they were released.



Data sources:
ICU bed use from April 5 via Dep Health infographic series https://www.health.gov.au/resources/collections/coronavirus-covid-19-at-a-glance-infographic-collectioninfographic_3.pdf 

Total number of ICU beds: https://www.mja.com.au/journal/2020/surge-capacity-australian-intensive-care-units-associated-covid-19-admissions


More model fails:

From: https://pursuit.unimelb.edu.au/articles/flattening-the-curve-to-help-australia-s-hospitals-prepare

And...

Friday, April 9, 2010

Some models are better than others

Update 15 April, 2010: The Australian's Cut and Paste pick up the thread


COMMENT:
Tony Burke Minister for Population April 8: Coalition 'flip-flopping' on population
"To be able to give a precise projection for 2050 forwards would mean being able to accurately predict every cycle in the national and international economy," he said.
"I don't see how any modelling would be capable of providing that level of precision.
"I don't see how you can determine what the economic needs of the nation will be. You can project forward on current trends, that's what Treasury projections do, but to be able to say in 40 years' time this is precisely what the economic needs of the nation in the new global context will be, I think is a pretty long bow."

Minister for Climate Change Penny Wong Interviewed by Tony Jones on Lateline 10/3/2009  Penny Wong discusses the Govt's draft carbon reduction scheme
TONY JONES: How many jobs will your emissions trading scheme create?
PENNY WONG: Well, what we can say is looking at the Treasury modelling that renewable energy and the jobs associated with that we're likely to see a 30-fold increase over the decades to come.
PENNY WONG: Well, look, modelling is modelling, and obviously the Treasury modelling is out there.
TONY JONES: I want to know what your Treasury modelling is suggesting in terms of the numbers of jobs an emissions trading scheme might create. I mean, that's what you were suggesting in the Senate today.
PENNY WONG: Well, the point I was making - and I've given you the indication that the renewable energy sector will grow by some 30 times out to 2050. The point I'm making is ...
TONY JONES: So how many jobs are we talking about?
PENNY WONG: Well, I'd have to look at the modelling, but it's a 30-fold increase.
PENNY WONG: Well, Tony, the Treasury modelling is a whole of economy modelling over a number of decades.
TONY JONES: So no numbers on potential jobs created in the term of your Government, or even beyond that for the next three years?
PENNY WONG: Well, the modelling that was done was a whole-of-economy modelling over a number of decades.

A more appropriate headline: Government Flip Flopping over use of economic models