Day 149: Hyperventilation

Another one of the unknown knowns of coronavirus was the advisability of mechanical ventilation for serious cases. Back in May, Kit Knightly documented the lack of science behind mechanically ventilating persons suffering from respiratory diseases, as well as the poor outcomes:

Perhaps it is not surprising, then, that according to this [Associated Press] article 66% of UK Covid19 patients put on ventilators are dying. A recent study found that, in New York, 88% of ventilated Covid patients died. In Italy it was over 81%, in Wuhan it was 86%.

He traces this unknown known back to WHO guidelines about the possibility of aerosolization of the virus during non-mechanical ventilation and oxygenation procedures, which presumably must take place more frequently per patient than intubation for mechanical ventilation.

This leaves wide open the possibility that hospitals are using treatments known to cause harm, simply to avoid the hypothetical spread of the virus.

Fortunately, the medical response has backed off of mechanical ventilation. PlagueBlog blames panic and the lack of any useful treatment for the disease for the initial choice to repeat a pattern had already failed in Wuhan, rather than the WHO's advice per se.

There's some new math out in preprint today: Predicting the Trajectory of Any COVID19 Epidemic From the Best Straight Line by Michael Levitt, Andrea Scaiewicz, and Francesco Zonta.

It is evident from our data analysis that the growth of a COVID19 epidemic does not follow an exponential growth law even in the very first days, but instead its growth is slowing down exponentially with time. While all growth functions decelerate exponentially when approaching the plateau, the Gompertz function is unique in that it is decelerating from the first day, and thus can fit the first part of the COVID-19 outbreak.

Why that's happening is another question, one they waffle a bit on answering. It could be asymptomatic spread depressing the numbers, network (superspreader) effects, or it could be that this sort of thing just happens:

The existence of invisible cases of individuals who are mildly symptomatic and, therefore, not counted as confirmed cases may explain the non-exponential behavior of COVID-19: the known cases cannot easily find people to infect as the hidden invisible cases have already infected them. We realize that other factors may limit growth. For example, the structure of the human interaction network can lead to sub-exponential growth (Moreno 2002). Still, we believe that as SARS-CoV-2 is so infectious, it does not have a problem finding people to infect early on due to the local network structure.

Initial sub-exponential growth is not a unique feature of COVID-19, but has been observed in previous viral outbreaks and needs to be taken into account to produce accurate predictions (Chowell 2016).

Massachusetts cases are up a fifth of a percentage point today.

P.S. In a published letter against intubation from earlier this month, A plea for avoiding systematic intubation in severely hypoxemic patients with COVID-19-associated respiratory failure, the authors confirm some of the suspected concerns mentioned above:

However, we reconsidered in other patients the rationale behind these early intubations and revisited the initial proposal of avoiding high flow oxygen in hypoxemic patients. In addition, we felt that the actual consequences of aerosolization posed by HFNC and NIV remain quite hypothetical as reported in H1N1 pneumonia. Our main concern was that a systematic intubation of every hypoxemic patient may prove to be untenable, facing a limitation of capacity and resources of intensive care units (ICU) to safely maintain a high number of patients on mechanical ventilation during the expected surge.