|
While I'm not defending the entirety of Ioannidis's argument (he makes a couple of mistakes; e.g. we know social distancing has some effect), it's important to point out what his central argument actually is, and how it's different from how the CBC article that purports to describe it. He doesn't assert that the policy response has gone too far; instead, he asserts that in the absence of good data on the pandemic, it's difficult to know whether the policy response is too much or not enough. In my view, on this point he's exactly right. From the actual Stat article: "At a time when everyone needs better information, from disease modelers and governments to people quarantined or just social distancing, we lack reliable evidence on how many people have been infected with SARS-CoV-2 or who continue to become infected. Better information is needed to guide decisions and actions of monumental significance and to monitor their impact." On "we lack reliable evidence": his point is that we don't have what we need, namely an accurate denominator in calculations of morbidity and mortality. "We don’t know if we are failing to capture infections by a factor of three or 300." He's not wrong! In epidemiology, accurately estimating, say, case fatality rates is notoriously difficult - especially in the middle (or the beginning!) of an epidemic. Given the issues we're all familiar with surrounding the lack of testing, this should come as no surprise. On "better information is needed": I don't think it's that far off-base to say that in the absence of (what asserts is) good data, it's impossible to calibrate policy responses to 2019-nCov so that you're saving as many lives as possible while doing as little damage to the rest of society (e.g. economy) as possible at the same time. This suggestion is eminently reasonable: "The most valuable piece of information for answering [questions about how many people will die of the coronavirus] would be to know the current prevalence of the infection in a random sample of a population and to repeat this exercise at regular time intervals to estimate the incidence of new infections.". Ioannidis (a physician and an epidemiologist) has made a career out of shedding light on science's mistakes and blind spots, and rightly so. I think the controversy surrounding this piece is overblown: the "fiasco" he's talking about is not a policy response fiasco, but an "evidence fiasco" - without knowing (or, in the US, even attempting to know) the true prevalence and incidence of coronavirus, it's impossible to say how well we're responding. * Edit / addendum: Counterpoints to Ioannidis to the effect of "well this is all moot anyway, when faced with high uncertainty about a really scary thing, you act as cautiously as possible regardless of the quality of your data" have a lot of merit. If you're a policymaker right now and you see what's happening in Italian ERs, hell yes you implement social distancing on a mass scale, maybe full lockdown / shelter in place, etc. But these two sets of tools aren't mutually exclusive; we have a lot of smart and capable people, plus a lot of money, to sic on each. We can allow policymakers to react strongly and cautiously right now to maybe save lives, while also setting up the sorts of basic epidemiological studies Ioannidis calls for: repeated random samples of a representative chunk of a population (e.g. the US population) with the ability to test everyone for corona (preferably using multiple different tests) in order to accurately estimate the denominator. |
- incubating phase - illness phase - post-illness with immunity - fatalities linked with the pathogen
Problem is that the test(s) are currently too pricey both money- and time-wise. Limiting factor here is time, because states push incredible amounts of money to save people.
Furthermore, I could find no reference to assert exactly what the test tests, and its reliability (false negatives/positives) ; this was illustrated by cases of recontaminations that could be linked to incorrect tests.
So now you're a state actor. You've got two choices : either, you play it low regarding extreme measures to contain the pandemic (first reaction of England), you delay measures (like most of European countries, and US), or you act swiftly (China, Korea).
Challenge here is that you won't know if the measures you have taken are too harsh until it's over.
Given the number of unknowns, what will you choose ?
Well let's say that overwhelmed hospital personal, facing dilemnas everyday regarding who deserves to live and who deserves to die, dying themselves to fight the infections, will give you the answer.
It's easy to qualify the reaction as disproportionate when you don't have skin in the game.