Now we are going to get some research papers. Oh boy, the fun part. First up, a study of actual infection rates in one Italian town. (Italy Infection Rate) Outside of China, Italy had one of the earliest and severest coronavirus epidemic waves. A small town called Vo’ was one of the epicenters. Researchers gathered extensive information on the epidemic in the town, including gathering test results on 86% and 71.5% of the entire population with two consecutive tests. The first was conducted at the start of a 14 day lockdown and the second at the end. The first survey, of 2800 residents, had a 2.6% positive test rate and the second, of 2350 residents, a 1.2% prevalence of which 8 were new infections. 43% of the detected infections across both surveys were of asymptomatic individuals. Most interestingly, there was no significant difference in viral load between the symptomatic and asymptomatic individuals. And as interestingly, no children age 10 and under tested positive, even though some lived in households with infected adults. Older persons had 3 times the infection rate. All asymptomatic persons at the first survey were still asymptomatic at the second test and most had cleared the infection. The asymptomatic rate was similar across age groups. So the obvious takeaways are 1) a very high percent of people who are infected are asymptomatic; 2) people seemed to be equally likely to be asymptomatic across age groups, although the likelihood of infection increased with age; and 3) a large number of people, especially the young, aren’t getting infected, even though they clearly have been exposed to a highly transmissible virus.
Next up a paper that examined potential undercounts of actual infections in regard to two different testing strategies. (Medrxiv Paper) The two strategies were “limited” and “inclusive”. In limited testing only people with severe symptoms or prior contact with a confirmed infected person are tested. In the inclusive scenario, tests are more freely available to anyone who may have been infected. The author created a model of how many actual cases there would be and how many would be detected in each scenario. Particularly at times when the daily number of infections is growing rapidly, the results suggest that actual infections could be 50 to 100 times more than test results indicate at the same time. Just modeling, but the results are intriguingly consistent with the increasing number of actual studies on the subject.
Okay, now another paper on a similar subject but this one on a method to estimate the fraction of unreported infections where there is a known epicenter. (Medrxiv Paper) The authors, as is often the case, seem to see a large number of asymptomatic or mild infections as a bad thing, assuming it means a lot of untested cases are out there infecting other people, but I see it as a good thing, indicating that the disease is not as severe as some worst case projections suggested, and the more people that have been infected, the faster we are getting to naturally slowing the transmission chain. In any event, the researchers’ basic premise is that if you know where an epidemic started and you know the number of travelers from that epicenter to other places and you know the transmissibility of the virus, we can estimate how many infections are caused in the places they travel to. They also assume that the reporting rates of infection are similar across the locations, which since the epicenter is in China, is a very difficult assumption to accept. They applied their method to the US and Iceland. Their model said for every case identified by a positive test result, there were 6 to 24 that were undetected. Since China has recently substantially upped its number of cases (although likely still far below the actual number) these multiples would go up a great deal.
Finally, no summary of research would be complete without one of my personal favorites, research on the transmission mechanics of the virus. (Trans. Paper) The researchers compared our lovely new virus strain to the original SARS and MERS. They found the new strain has a very high persistence in aerosols and can retain its ability to infect when suspended in aerosols for up to 16 hours. Four laboratories participated and created aerosols with the virus that were directed toward potential infection targets. All showed that the virus persisted longer than its counterparts and across greater distances. One laboratory tested how long it remained infective. That was at least 16 hours and there was a very slow decline in numbers of virus in the aerosol samples. You do have to wonder if this thing was engineered.