(1) In Dr. Seheult’s 100th update, he gives a retrospective, together with a brief discussion of the promising beta-interferon trial. At the end, we get a brief musical lagniappe. (He is apparently an amateur classical organist, seen here playing BWV 565 on the organ at Nantes Cathedral, which one of his ancestors helped build.)
(2) Wendover Productions is an educational channel that mostly centers on aviation, logistics, and engineering, with occasional geopolitics thrown into the mix. Here they discuss the logistics of COVID19 testing, as well as the concept of “pooled testing”.
In a nutshell, pooled testing saves resources as follows. If you need to test 100 people, it needs 100 “slots” in your automated RT-PCR machine. Now imagine you do the following: create 25 “pools” of 4 samples, create a mix for each pools of a fraction of each sample, then test these 25 mixes (requiring just 25 slots). If a mix comes back negative, all four patients in it are negative, end of story. If a mix comes back positive, you go back to the stored rest of the four samples in the mix and run those individually.
If you have a positive test rate of 8% in a population, for example, each batch would have a 28.36% chance (i.e., 100%*(1 – 0.92^4) of having at least one positive sample in the mix, or about 7 out of out 25 batches. For these you do individual testing on the remaining fraction of each sample, which requires 7*4=28 additional slots. So you can now handle 100 samples with 25+28=53 testing slots.
Clearly, as the percentage of positive tests runs up, this can become a mug’s game, while as it goes down, you could speed things up further and/or expand testing for the same capacity by creating larger pools. A group at UC Berkeley proposed using an artificial intelligence model of the population to guesstimate an infection rate, then automatically optimize pool size based on that.
- Vaccine trials are promising
- (Practical) herd immunity may be more quickly reached than previously thought, owing to cross-immunity with different coronaviruses in a large swath of the population
- Quoting: ” The one place where the virus did spread with horrible ease was in care homes and hospitals. Why was this? T-cell senescence is an issue, so old people’s immune systems are just not as good at coping with this kind of infection, and there were dreadful policy mistakes made, like stopping testing people, clearing patients out of hospitals to care homes without tests, and assuming no asymptomatic transmission. Healthcare and care home staff were not properly protected and were allowed to go from site to site. Many were infected and became carriers. [Now we know better.]”
- “The fourth cause for cheer is therefore that now we know about asymptomatic transmission, we have more protective equipment and we have a better, if still imperfect, capacity to test, track and isolate cases, it is likely that the hospital-acquired epidemic of the spring will not be repeated.”
- “My fifth excuse for being hopeful is that we now know better how to treat people who get seriously ill. Ventilation is not necessarily the answer, blood clotting is a real threat, making patients lie face down is helpful, dexamethasone can save lives and some antiviral drugs are showing promise.”
He also notes that due to social distancing (even short of lockdown), not only was the covid19 epidemic mitigated, but annual deaths due to seasonal influenza were significantly reduced, which is one reason overall excess mortality during the flu season was considerably less than the COVID19 death toll. Read the whole thing: Matt Ridley’s prose is lucid as always.