A Head Full of Coronavirus Research, Part 20 (Because I lost track of where these are too)

Okay, quite a bit of interesting research today.  More analysis on the passengers of the Diamond Princess cruise ship.  This paper focused on persons who were asymptomatic.  (NEJM Paper)   Just a refresher, 3711 passengers and crew on the ship, 712 became infected (this is what really caught my eye very early on–why did so few people get infected under these circumstances) and 410 of those, over half, were asymptomatic.  This study tracked 96 of these asymptomatic persons and 32 of their cabinmates after they were taken ashore and quarantined.  Of the 96 positive but asymptomatic persons, only 11 later developed symptoms, in a range of 7-11 days after the positive infection test.  Being older was associated with more risk of developing symptoms.  Out of the 32 cabinmates, 8 eventually became positive but all remained symptomless.  The median number of days from a positive test to a negative test was 9 days, with a wide 3-21 day range.  48% of infections had resolved in 8 days and 90% in 15 days.  Older persons took longer for resolution.  Large number of uninfecteds, large number of asymptomatics.

This study addresses differences in antibody development between mildly and severely ill patients.  (Medrxiv Paper)  The researchers identified 38 patients with severe illness, all of whom developed strong antibody responses.  Among 24 hospital staff who had mild illness, 75% developed antibodies, but not in as great a concentration.

This paper tried to give people a sense of their risk of getting infected based on contacts per infection.  (Medrxiv Paper)  The researchers looked at the 100 most populous US counties as of the end of May.  The median probability of an infection was one per 3836 contacts.  For a 50-64 year old, the risk of hospitalization was one per 852,000 contacts and the risk of death was one per 19.1 million contacts.      The risks are even lower than estimated by the paper, as the researchers state.  The authors assume 100% susceptibility and they assumed that transmission was equally likely across all contacts and all residential settings, where as we know it is not.  They assumed a 90% asymptomatic rate, which would also raise the apparent risk of an infection.  As the authors note, the importance of their findings is to help give people a more accurate basis for their perceptions of individual risk, which currently are generally way too high.

This study involved an attempt to identify the number of undetected infections.  (Medrxiv Paper)   The authors were from Israel and attempted to ascertain the relationship between detected and undetected cases by various methods.  They also examined the utility of knowing the source of infection.  They applied their formulas to typical epidemic models.  They concluded that if you know the number of cases for which you can determine an infection source, you can also identify the number of undetected cases.  While I understand what they were suggesting, their method suffers from the usual data completeness and accuracy limitations.

Here is an interesting approach to understanding what may protect against infection or serious coronavirus illness.  The researchers examined a patient’s total set of antibody responses to infectious agents and certain biochemical measures.   (Medrxiv Paper)   The researchers, using data on patients from the United Kingdom, looked at demographics, self-reported health status and illnesses, a standard blood panel and vital signs, collected 10 to 14 years ago.  A total of 4510 patients with 7539 separate tests were included.  5329 were negative, 2210 were positive; out of the positive cases 996 were mild and 1214 were severe.  A subset of 80 patients had blood tests for antibodies to 20 common infectious diseases.  The general set of factors had modest at best explanatory power for who might get infected and have a serious illness.  The antibody tests, however, while in a smaller sample showed more association with infection risk and seriousness of illness.  In particular, exposure to certain viruses resulting in antibodies to those seemed to be correlated with higher infection risk and serious illness.  Which suggests that some kinds of prior infections may hinder immune response to new ones, especially in older people.