April 9, 2020 - 4:15pm

One popular study touts hydroxychloroquine as an effective treatment

One reason that Covid-19 is so much more concerning than superficially similar respiratory illnesses, such as the flu, is that so many more patients require ventilation. That’s a big part of why the March 16th Imperial model was so alarming: because it became clear that the virus would hugely overwhelm the NHS’s ventilator capacity.

There have been some studies, Twitter threads and so on going around suggesting that there’s something strange going on here. One much-shared thread, based on a preprint paper, suggests that the virus attacks the body’s haemoglobin in such a way as to prevent oxygen getting around the body properly. Another viral (as it were) piece shared via Pastebin suggests something similar. The claim is that it has implications for treatment: use oxygen rather than a ventilator, use hydroxychloroquine.

Other reports come from front-line doctors saying that ventilation is less effective than it should be, because it’s not like normal pneumonia. One Italian professor of intensive care writes that unlike normal acute respiratory distress syndrome (ARDS), people are capable of breathing on their own — their breath doesn’t rattle — but the oxygen isn’t getting to them.

I don’t know if these reports are compatible; to me it sounds as though they describe the same thing from opposite perspectives, but I’m no respiratory physician. I quickly DMed a couple of doctors about this and they both say that it sounds plausible and they’re not ruling it out, but that it’s far too early to say.

One, a respiratory specialist, had been chatting to a colleague, who thinks that the science in the first set I mentioned is dodgy, but that there is something strange going on with blood oxygen, possibly related to the immune response.“It will be interesting to see how those left on CPAP [continuous positive airway pressure, an alternative to ventilation] for longer fare,” she says. “Everything happens so fast it’s hard to evaluate treatments.”

The thing to remember in these situations is that science works through mutual checking. Partly that’s peer review, but more importantly, it’s replication: other scientists going back and doing the experiment again. Think of scientists as hypothesis-generating machines; they spit out possibilities, other scientists check them. A lot of these hypotheses are wrong, even the ones that get published, even in normal times (possibly most of them). There are ways of reducing the error rate, but none of them are perfect.

And at the moment, science is spitting out more hypotheses than ever, because the situation is urgent; and they’re being released as preprints. That’s great, because it means they’re available for other scientists to check. But it also means they’re out there for others to see, who might see “scientific paper” and think “thing that is true”.

The hypoxia/ventilator hypotheses mentioned above might well be true. But I wouldn’t place any sort of weight on them yet. Let the wheels of science grind a bit further before you do that.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers