Warning, LOL, super long post here… I am replying to several posts in the Cabin Camp Forum about Covid 19, but I didn’t want to hijack that Forum, so that’s why I’m starting a new thread here.
@Peg_E - thanks for giving pertinent, factual information and excellent advice for prevention in the Cabin Camp Forum! I was easily able to verify your flu statistics on the CDC website (the data has just been updated now to show CDC estimates at 34 million flu cases, 350,000 hospitalizations, and 20,000 deaths from the flu).
I’m an actuary and I thought I would start looking into a little bit of the data available on Covid 19. I could not find any data on CDC or WHO websites regarding testing, cases, mortality, etc. for the coronavirus. As of March 6, the CDC does report 164 cases in 19 states with 11 deaths. The CDC says that the individual states’ data should be considered the most up to date. The vast majority of reported cases are in WA (70 cases), CA (45) and NY (14). Forty states report zero cases, and seven state report 1-5 cases each. I read one report that in the U.K. there have been 18,083 tested with 115 positives and 1 death. Still too small a dataset to draw conclusions from but the net they’ve cast appears to be a little larger than the U.S. at this point based on the relatively low number of positives, plus the fact that the U.K has a more centralized health care system than the U.S. leads to a potentially broader reporting scope in the U.K vs. the U.S. Obviously @Archie would know more about the situation in U.K. than I.
I am skeptical of coming to any conclusions regarding the mortality rate of Covid 19 at this point for several reasons: 1) The CDC is no longer reporting negative tests - this is a number that would contribute to the denominator of a mortality rate, making it lower (although they supposedly have announced that they will resume, but I haven’t confirmed), 2) We aren’t getting a clear number of the number of tests performed for Covid 19 from all of the individual states and some like FL will not report that number at all (Miami Herald, 2/18/2020), 3) What little data on actual testing that has been provided by states shows a paltry number of tests, most of which have been targeted at high probability patients (travel, known contacts with Covid 19, obviously sick, etc.) According to articles I have found (I have not personally verified at the state level), CA has only tested 516 people as of 3/5/2020. That’s an incredibly low number for such a populous state. And again, those tests have been highly targeted. They now have a total capacity for doing 6,000 tests per day and are expected to ramp up their testing. WA can test 1,000 people a day. OR can only test about 40/day, and TX about 30/day (according to the Texas Tribune). 5) Many cases are mild enough that the symptoms are minor or don’t even present at all, and those people will not be tested – more data that should go into a valid mortality rate but will not. 6) The incubation period from Covid 19 seems to be fairly long (2-14 days according to the CDC, 7) The people getting tested now are obviously in a select group of people who are sick enough to go get tested and/or who have had recent activity that puts them in a high-risk category. So there is an extreme selection bias in that data. Selection bias can completely destroy any ability to draw valid conclusions from a dataset. Of course the mortality will initially be overstated in any situation like this until more complete data is collected on a legitimate sample population. The “law of large numbers” requires a carefully inspected set of representative data, not a biased set of data, before applying conclusions to a general population.
Points 1-7 indicate that the number of actual cases of Covid 19 is likely significantly under-reported at this point. A mortality rate due to a particular cause (very simplified) is #ofdead/#ofsick from that cause. The number of dead is probably a reasonably sound number. The # of sick is likely significantly under-reported. I personally do not give any credibility to any estimate of mortality rate at this time.
- My points 1-7 are focused on the U.S. These same factors should be considered regarding other countries you get information from, plus you also have to consider that living conditions, age of population, health of population, testing capacity, medical treatment capacity, transparency of reporting, and many other factors will vary widely among countries. This means I will take a long pause before drawing any definitive conclusions on the pervasiveness or the mortality of Covid 19 at this point. I would hazard a guess that pervasiveness is higher than being reported, and that the mortality rate is lower than being reported.
In other words, as stated in today’s WSJ, “But once the number of asymptomatic or minimally symptomatic cases is known the real fatality rate for new coronavirus may be less than 1%, says Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, in a New England Journal of Medicine article published recently… the exact rate won’t be known until experts know the true denominator, which is the total number of people infected, including those who are asymptomatic or never got tested.”
Numbers that get thrown around in the press and presented like a current mortality rate are unsupported, in my opinion. It’s hard enough to get the raw data (# of tests done, on whom, etc.), but going a step further and drawing conclusions from incomplete data and publicizing those conclusions as fact is irresponsible, in my opinion.
I take a very analytical view of data. I always question the validity of the data and the use of data. The saying popularized by (but interestingly not attributed to) Mark Twain goes, “There are three kinds of lies: lies, damned lies, and statistics.” LOL! I’m not saying that anyone is intentionally distributing bad data. I am saying that garbage in = garbage out, and so we have to think critically about assumptions before applying a rational thought process to come to conclusions. I think that the biggest mistake people tend to make is not necessarily in using factually incorrect data (although that’s very common), the more insidious problem lies in not having or not using all of the relevant data needed. It’s not knowing what you don’t know, but plowing forward anyway.
Bottom line for me: I will heed @Peg_E 's excellent advice in the Cabin Camp forum, continue to monitor the facts (not breathlessly reported fantastical headlines), and work on my song list for Cabin Camp so I can come to camp as prepared as possible!