MusicMan":263vms7a said:
Andy Kev, from my very limited knowledge (I'm a materials scientist not a virologist, I just know some of those working in the field) I think you are probably right about the agreement amongst virologists.
About what though? The make-up of the virus? They've done amazing work (not least in China) on that sort of analysis, and the early stages of looking at cures and vaccines. Super-impressive from the scientific community.
And I do know that the epidemiology is by no means regarded as settled, but as work-in-progress. The methods are quite well-established and not controversial but the parameters and influences are at present best-assumptions, which are refined as more data from the UK and other countries are available.
Again, indisputably and inescapably the case. And they're doing great work, not least in testing each other's assumptions and outputs.
The ONE issue which has arisen was just that: a failure to use best-assumptions for COVID because the Imperial team did not update one of its assumptions made in its flu model to reflect the best assumptions for COVID.
The head of the team admitted this in an interview with the FT. They caught it in after the decision to publish their modelling. Neil Ferguson has been pretty open about that, and then popped up on Twitter last week or so to say by way of explanation that it was an assumption buried in 20k lines of C program he wrote 17 [from memory] years ago. That's a mea culpa in anyone's language.
The mistake is maybe understandable, but should have been checked as it is a key assumption from public health terms given the known empirical experiences in Wuhan and then Lombardy was that the demands were much higher than a flu and the health systems were crashing way beyond capacity.
The bigger issue is the decision making and policy process (which is not a matter of science, nor really of politics although it is inescapably going to be influenced by that, not least in popping up risks such as confirmation bias). As the govt framed its whole mitigation policy around aiming to manage the spread of the disease just below NHS surge capacity, the assumption as to how much demand on the NHS there would be is obviously key and one does not need to be an epidemiologist, virologist or scientist to understand that. It's really just a very basic check that the scientific evidence base on which you are choosing to rely has been prepared with due care and attention without any slip-ups of the sort Neil Ferguson has admitted to.
There is also the question of how best to handle the very large and disparate data sets that are coming out from the different countries. The epidemiological community, which is small but very strong in the UK, is working flat out to refine and improve the models in addition to spending much time advising government bodies. A call has in fact just gone out to the whole of the UK modelling community (all university research teams, all scientists with track record in the field including long-retired ones like myself) to add either specific expertise, cross-disciplinary insights (for example, how very large data sets in other fields are handled) or human and computing resource to help them in this national effort. It is coordinated by the Royal Society. The first aim is to understand the potential effects of the various options for exiting the escalating lock-down strategy in order to make more robust predictions.
All great applaudable stuff, and yes, thank god for the nerds.