OOPS: these are per 10,000, not 1000.
You’re seeing lots of graphs and tables on COVID 19. In a disgusting display of fear mongering, networks are now keeping running counts of the cases and deaths on screen as if tracking stocks or something. But all of that lacks context. How does COVID19 compare with a bad influenza outbreak? We’re starting to get enough data to seriously answer that question. I’m using the 2017 H3N2 outbreak for reference, which was a bit worse than an average season, and the data from three areas that are relatively farther along in the process: Hubei, China, Daegu, ROK, and Lobmardy, Italy. Influenza is in blue; Hubei is orange, Daegu yellow, and Italy in green. We start the clock on our graphs at the first known case and plot cases per 10,000 population by week:
Oh. That’s not very interesting. The COVID19 cases barely show up! What is going on? Well, to start with, we are probably only detecting/reporting a fraction of cases. Lets scale H3N2 flu to assume we are only detecting 5% of COVID-19 cases (which seems to be the range in the literature at the moment):
That’s a lot more interesting – and really illustrates how COVID19 is both different from and more stressful for the health care system, and not as bad for the general population, as influenza. Notice how rapidly the cases explode for COVID19. This is why the outbreak is so stressful for hospitals: the cases flood in over 3-4 weeks, as opposed to 20 weeks for a flu outbreak.
There is a lot to learn from these numbers and graphs. Notice the sharp break in the Hubei China curve. There are likely two reasons for that. First, they instituted rather draconian travel restrictions. Second, they are likely not being entirely honest about their reporting, either internally or externally. From the Korea curve. which is probably pretty reliable, it looks like COVID19 cases will level off between 1.5 and 2 cases per 10000. H3N2 leveled off at over 60 per thousand, but if we scale it to the same detection rate we suspect we are seeing for COVID19, then 3 to 3.5 is the range, and therefore COVID will have maybe 2/3 the impact of a bad flu season in terms of total number of cases, and mortality. Italy is in the steep part of the curve. We should see their case rate slow over the next week and level off, probably in the 2.0 to 3.0 /10,000 range (a bit higher than Korea due to the older population and later implementation of control measures).
What about the US? We are just entering the steep part of the curve. It will be very scary as cases explode – but keep it all in context. If these trends hold, the US can expect about 30 to 35 million people to be “symptomatic” (most mild), 400,000 to 500,000 need hospitalization, and 30-40 thousand deaths. Compare to the 2017 influenza season: 45 million symptomatic, 810,000 hospitalized, 61,000 died. HOWEVER, rather than coming over 20 weeks or so, those cases will come over maybe 4 weeks -five times faster. The US health care system can’t really keep up with a normal flu season; there is no way it can handle this flood. That is why COVID19 is so dangerous, and why everyone needs to take this seriously, following the CDC guidelines, exercising social distancing and hygiene protocols, and otherwise doing everything you can to try to slow down the spread. It’s more than likely not about you. It’s about that 1% of so of the population who will get very sick, and may not get enough care because the system will be overloaded. The state of the US system is a disgrace, and its inability to handle this outbreak is the result of health care policy decisions going back decades. That will likely be the subject of an upcoming vehement rant …
Influenza Hospitalization Surveillance Network (FluSurv-NET), US Centers for Disease Control.
2019 Novel Coronavirus COVID-19 (2019-nCoV) Data Repository by Johns Hopkins CSSE