Another Viewpoint.

[Captain23]

https://www.medrxiv.org/content/10.1101/2020.04.24.20078717v1.full.pdf

[woofgang]

Iam not sure that I would belive an oceaographer over virologists, epidemiologists and medics.

[Canary42]

Medrxiv, the publisher of this paper, issues the following warning in respect of such papers :-

Preprints are preliminary reports of work that have not been certified by peer review. They should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.

I think I'd rather listen to Karen on Facebook LOL.

[jim360]

I haven't gone over it thoroughly but it seems to me that there are more than a few flaws. Not least, the focus on the worst-affected countries seems a little difficult to justify: would it not be appropriate to consider countries where lock-down was implemented earlier in the disease's growth curve? Is there a point where the intervention measures are, perhaps, "too late" to have a major effect? What happens if the intervention was brought in earlier?

The focus on death statistics is also somewhat problematic because these are still being updated. Today's round of NHS England updates, for example, added a few deaths from towards the end of March. Even leaving those problems aside, the author's "pre-lockdown projections" in eg Figure 4 seem wildly optimistic, both in the choice of function and in the parameters. Consider, for example, the claim that without interventions the Covid-related death number in Italy would have dropped to zero by about two weeks ago. Presumably this is meant to represent an optimistic scenario but must come with a huge uncertainty that appears to have been not taken into account, in the presentation of the data if not in the processing. Otherwise it is simply unsustainable to make this claim.

I'm sure I could think of other criticisms. One that is more than suggestive is the number of typos, which in itself doesn't undermine the work but does show that the work has been rushed a little and not already examined by somebody else prior to preprint. Coupled with the fact that it's not been subject to peer review, it could seriously benefit from an outside voice asking questions.

Oh, yes, and it's a paper from someone outside the specialism. Again, in itself that doesn't undermine the paper, but it gives pause for thought: does the author have the experience needed to analyse these data in this context in a sensible way? I'm not an expert in epidemiology or anything close to it, and I wouldn't presume to write a paper questioning the efficacy of public policy with so little background in the field.

[TheChair]

As jim360 says // would it not be appropriate to consider countries where lock-down was implemented earlier in the disease's growth curve? Is there a point where the intervention measures are, perhaps, "too late" to have a major effect? //
The 2 countries that acted quickest, Greece and Croatia have considerably better results. Greece started planning before there was a single known case, and started moves toward a lockdown the very next day after the first confirmed death.

[Peter Pedant]

// I am not sure that I would believe an oceanographer over virologists, epidemiologists and medics.//

hi evrybardy ! - see above - yeah but they could be right. Bearing in mind that there is a fallacy post hoc ergo propter hoc.....
you have a situation where Ro was above 1 ( up to 4 they now say) = bad really bad
and now we have Ro < 1
and it seems relevant to ask - IF an interaction led to this - and which one or ones

yes I am writing this in a deliberate technical scientific style as I see Jim The chair and a few uvvuz have contributed

this morning I was thinking of Pere Le Maire ( or Le maitre) who was right over the beeg bang 1929 and of whom Einstein said - "You av boodifoll ideas but offool maff" OR - "you have beautiful maths but your ideas are awful"

and um Le Maitre was right and Einstein wrong

so I will give it a read - it is not so that I am strapped for time....

Moral story - one of you suggested that population density should be factored into infection models - it would reduce Ro anyway. And I was rather rude about that. And in the times three weeks later - an article about northern isles and er Wiltshire ( yes wiltshire) and how they have seen nothing.....It was all to do, the hacks said, with lots of sheep and not many people.
On the same day - - - the outbreak in the Skye nursing home. oops and no mention of Panum and the Faroe Islands ( 1860 isolated island that got measles and they all died - see wiki)

Hey did anyone notice today - ? Beeb Araby uncovered a story where an Iran airline ( not arab see?) ran consciously infected flights - to everywhere ! and by implication was the real culprit in the world wide spread - and were told by the workers and said they were gonna do it anyway

[fiction-factory]

Yes, the fact that Ro has fallen from around 3.5 to around 0.7 suggests something has changed. Of course it could have been the sunny weather or just random chance but it seems pretty obvious that restricting the number of interactions and introducing social distancing will have reduced the spread. To take an extreme case, if everyone had locked themselves in a room by themselves for 2 -3 weeks with food and water then some would have died , some would have recovered and most would have ended up not having it, and the disease would have pretty much disappeared for now here.
Scientists now seem to agree the peak here for deaths was around 8th April and the peak for infections was around the time of the lockdown (although the distribution curve is a skewed normal one). The Ro was it seems already falling before lockdown from other measures such as closing schools, shielding, home working and stopping large gatherings, so it's not clear whether the decisive factor was the final lockdown or the earlier measure or the combination.