Since I first posted country case growths on a common timeline, our line has turned sharply up compared to many other countries. Nearly all of this growth is due to cases that existed BEFORE we went to extreme social distancing, we're just learning about them now. First I'll explain why we're seeing a rise in cases, then I'll write about why the increase could even be a good thing.
The reason why we're seeing a spike in cases is that we are doing so many more tests daily. There were already many more cases that we counted, so we are bound to find more cases, even if they aren't severe or show no symptoms at all. Even if nobody new caught the virus yesterday, we would still be seeing this rise. While complicated, the graph below from an article from JAMA may help (a bigger version is here).
Look at the gold bars growing on the right hand side. Those are the number of cases the Chinese Government identified. The Blue bars before it show when people actually contracted COVID and was reconstructed by interviews with patients after the fact. If we had a graph like this for the US, we would currently be in the time above the light yellow shading where the blue bars are much larger than the yellow bars -- we are observing the growth that happened weeks ago before we started taking serious action.
To get an idea of our actual count, we can try to build it from what we do know - the current death count. Let's assume the mortality rate for COVID is 2% (the research is still unclear on the actual number, but this is in the average). As of March 20, there have been 256 deaths in the US due to COVID. If that was 2% of people carrying the virus, it would mean that there had to be 12,800 cases to result in that number of deaths. We also know it takes 17 days on average for someone to die of COVID, so that means that 17 days ago, March 3, we probably had 12,800 cases of COVID in the US even though the number we were actually reporting was 124 cases to-date. We had 100 times more cases than reported! Of course we can only calculate that looking back 17 days so its hard to know how many we have now, but even if it doubled every 6 days, as some research suggests, we would have over 100,000 cases now (I know this may be shocking, but I'll explain later in this post why its not as bad as it seems).
If you accept we have something like 100,000 cases in the US, then the testing results we're getting now isn't really about new cases, its about us finding the cases that have been here for a while. The number is rising sharply because we've gotten much better at testing this week.
Even though our case rate is curving up like Spain and Italy have, it doesn't necessarily mean we're at all in the same situation. Keep these things in mind (from this source):
- 81% of those who test positive for COVID-19 have mild cases who at worst get mild pneumonia .
- 14% have severe cases
- 5% are critical, of which 2.3% result in death.
So more than 4 out of 5 cases we identify are relatively mild. Given the more acute cases are more likely to need medical care and get tested, it may be that our accelerated testing is finding mostly mild cases. One way to see if this is true is to look at the common timeline of case growth and the common timeline of fatality growth side by side. Here's the chart I showed you at the beginning of this post:
And Here's the corresponding chart of fatalities (both from my Covid sheet)
Looking at the fatter line that shows the US growth, the number of fatalities are not spiking upwards like the number of cases did. While its possible this could change, at the moment, the most sobering aspect of this pandemic is not growing in the US nearly as dramatically as the case growth at the moment. You can also look at the relative % of deaths to cases in my current tab.
I've explained why the rise might not be as alarming as it first appears, but why is the rise possibly a good thing? Because we're making good progress on identifying the unidentified cases that were here already. Assuming anyone who tests positive self-quarantines and watches for symptoms, all these discovered cases reduces the chance for the virus to spread to other people or that someone develops a serious case without getting treatment. One other reason is that before we can see the effect of how the social distancing and other measures we've taken as a country, we need to identify the cases that existed before we took those steps. I'm certain the hard and expensive steps we're taking as a country will bear fruit, but we're going to see numbers rise before we see the effect. I'll look into the effect of social distancing in an upcoming post.
Given the calculation from the death rate, I think you are right that there are likely at least 100K cases already (almost 2/3rds not confirmed due to being asymptomatic or lack of testing). I fear the number is higher, but it is so hard to know given how many people are asymptomatic. What I hope though is we experience a lower death rate like in Germany, because if we experience anything like Italy’s per capita rate, the number of deaths will be extremely sobering. I feel morbidly curious (wrong term I know) about why in some countries the rate of death seems to be much lower, like Germany's. It may mostly be linked to the unknown number of cases and lack of testing, but I wonder if it is also demographics, weather, when the country got serious about physical distancing, etc.
ReplyDeleteDan