Drowning in Coronavirus Research Part Diez

PriceWaterhouse’s Health Research Institute did a survey to find out what consumers were doing in regard to health care as a result of the epidemic.  (PWC Report)   About 2500 people participated.  About 22% said they already have spent or plan to spend less on medications.  31% said they have spent or plan to spend less on health care visits.  78% said they would skip at least one health care visit, while 30% said they probably would spend more on health care.  5% said they used telehealth for the first time during the epidemic and 88% of people said they would use it again.  A significant of respondents said that their employer had added or expanded a benefit during the epidemic.  Consumers also reported greater issues with social determinants of health, such as mental health, feeling lonely, spending too much time with technology and difficulty finding healthy food options.  Social distancing compliance varies widely by age group.

Another paper on my favorite topic, where is this virus hanging out, other than in our bodies.  This paper finds that it can linger for a while on surfaces and in another, pretty gross place.  (Medrxiv Paper)   At room temperature, the virus can stay viable on a smooth surface for several days, and on typical clothes fabrics for 3 or 4 days.  It also appears to be viable in feces for an hour or two and in urine for 3 to 4 hours.  So please, don’t be touching that stuff.

The next paper covers transmissibility of the virus during a person’s incubation period.  (Medrxiv Paper)  The authors built a model state by state for the US and attempted to infer the epidemic flow.  They used a fairly traditional SEIR model, but added some buckets for people who are infected but not symptomatic and for infected people who are and aren’t detected as transmitters.  Their model estimated that about 16% of transmission was from the latent bucket and 29% from the undetected infected transmitters.  They also found a fairly high transmission rate early in the epidemic.  Some of their projections, however,  probably were susceptible to error due to not tracking or accounting for variable and changing testing strategies.

Up next, a paper that looks at the effect of various factors on infection rates, including population density.  (Medrxiv Paper)   This group accounted for population density, socio-demographic factors, social distancing and testing data.  They claim their model was highly predictive, but not clear if they were just fitting the model to the data.  They did find that high population density, poverty index and minorities were correlated with more infections.  Social distancing is associated with lower rates of infection.  More testing, what a shock, is associated with more infections.

And finally a couple of notes on comments by Scott Gottlieb, the former FDA Commissioner, who has been pretty vocal about the epidemic.  This morning he said there are about 10-20 missed infections for every positive test result.  Note this makes all case ratios, like hospitalizations per number of cases and deaths per number of cases, one-tenth to one-twentieth lower.  The reported number of cases in the US by infection testing is about 1,394,000, so really 14 million to 28 million cases.  And 82,471 deaths.  So either a .6% or .3% case or infection fatality rate.  And again, heavily, heavily skewed toward the elderly.  In the general population, it is less than half that level.

The other thing he mentioned on Twitter, I believe, was the study I posted on a week or so ago indicating that high variability in infection susceptibility significantly affects population immunity levels.  Early on I mentioned that this epidemic looked front-loaded, which would screw-up traditional statistical and modeling assumptions, so it is good to see the research bearing that out.