What is my chance of dying from COVID-19? Scroll down to the third paragraph from the bottom if you want to skip the explanation.
I have a few friends on Facebook who continue to present the current mortality rate from COVID-19 relative to the total population as one’s “risk of dying” from the disease. They come up with vanishingly small numbers. Their purpose seems to assure everyone that this is all no big deal and the media are blowing this out of proportion. Indeed, it has twisted into an ugly partisan issue, which, at its core it’s not. Let me address this as briefly and simply as I can (I wrote this sentence before writing the remainder--it’s long, but I wanted to explain a lot).
The numbers they are deriving do not reflect “risk of dying.” Those numbers are not your personal risk of dying in the future (risk is always forward looking, a prediction) from COVID-19. They are simply the percent of people who have died, already, relative to the population. Those numbers mean no more nor less than that. And that’s pretty useless information since if you are reading those numbers, you are obviously not dead. Anyone reading those numbers has a 100% chance of being alive at that moment. Sounds obvious, but evidently needs to be said.
Risk is forward looking. In this case it states, “If SARS-CoV-2 continues to transmit under a particular set of conditions, X is the probability that you will die from it.” It is a prediction, but is based on previous data. When I talk to someone who has new onset Atrial Fibrillation (AF) I take various factors into account that contribute to risk of stroke or associated diseases from a blood clot resulting from AF. Those factors are age, history of stroke, diabetes, CHF, peripheral vascular disease and others. We can calculate one’s risk of having a stroke over the next 10 years using previous data and studies. If the risk is 1%-4% we will have a discussion about the use of aspirin vs. an anticoagulant vs. doing nothing in decreasing their risk of stroke. If their risk is >4% (over the next 10 years), we consider the benefits of prescribing rat poison (Coumadin) far outweigh the risk of adverse effects from taking it (or a newer alternative).
What are the factors that affect one’s risk of dying from COVID-19? Social distancing, how easily the virus is transmitted, demographics (i.e., age, comorbid diseases, possibly genetics, evidently wealth status). It’s dependent on a myriad of factors, some known and some unknown, and some of them are personal and not universal. For example, we know elderly people are significantly more likely to die, so their risk is obviously higher. To make the numbers that follow easier, we are just going to calculate a single risk of dying as if everyone were the same. It’s not accurate for any particular individual, but it will give us a sense of the total toll. In addition, we will just look at the U.S. population.
We can assume that the virus will continue to spread, since it still is (the “risk of dying” group from above has to assume everything stops right now for their calculations to be any kind of meaningful risk assessment). Since the issue with that group seems to be a dissatisfaction with various measures to mitigate the spread, let’s assume that tomorrow the entire country goes about business as usual, ignoring the pandemic. In that scenario, the virus will burn through the population (we already know it’s virulent enough to do that) until herd immunity is reached. That is until at least 67% of the population is exposed. Although the deaths will continue even after herd immunity is reached for complex epidemiological reasons, we’ll ignore those deaths to simplify things.
One argument from this group is that the actual prevalence of the virus in the population is much higher than we know, so the fatality rate is much lower. They use a number like 55x as much. There is no study to indicate this. There are a couple of studies that promote a number close to 50% (meaning for every COVID-19 positive test out there there is an additional undocumented and asymptomatic case). Those studies have serious design flaws and dubious extrapolations, in addition to using antibody tests with known notoriously high false positives. But for the sake of argument, we will assume that half our COVID cases are asymptomatic and undiagnosed. That brings the U.S. current case fatality rate (which is hovering around 6%; meaning 6% of those diagnosed with COVID-19 have died) to a “true fatality” rate of 3%. That may be true. We do know that some people are asymptomatic or presymptomatic (although the exact number is not known and difficult to tell right now without a better antibody test). In fact, we don’t have to depend on our numbers at all, because South Korea has done an excellent job of testing, tracing and quarantining. They have a fairly accurate count of the number of cases and the number of deaths. So let’s use their case fatality number of 2% since it’s lower than ours and will give a more conservative estimate.
Current US population is estimated at 328 million. 67% of 328 million is about 220 million people infected. 2% of that die (within the next year or so), which is 4.4 million. A generalized person’s risk of dying from COVID-19 if we can’t mitigate this in some way is about 1.3%. These numbers don’t include the deaths that would happen because they might have survived if they had an ICU bed, but we were too overwhelmed to get them in. Or the deaths from other diseases that would be put on the back burner because we as health care workers were overwhelmed with COVID cases. In addition, it is using a dubiously low case fatality number from South Korea (I suspect ours is so much higher partly because of our higher rates of poor health like heart disease and diabetes).
This is a hypothetical; it is based on assumptions using the data we have. But that is how one calculates risk, and we now have a lot more data than at the beginning of this. These risk predictions give us an idea of the magnitude of the situation (like a weather forecast for the next day). Obviously we are working hard as a nation to mitigate this outcome. The temporary shutdown bought us some time to get testing ramped up and PPE available to health care workers to prepare for the surge that will likely come in the next few months. The summer may buy us some more time (transmission is less likely in open air) if enough social distancing measures are universally employed. The longer it can be drawn out (assuming lasting immunity after infection, which hasn’t been demonstrated yet), the lower the number of people who need to be infected before herd immunity is reached. For all these complex reasons, please follow the advice of the health authorities and not memes put out by bots or any politician who is not listening to expert advisers. Above all, please don’t make this about politics; this crisis has the potential to unite our country, which it seemed to be doing initially. We will accomplish much more and better together.
Sorry to be “downer.” I will write some more about how you can improve your health to prepare best for SARS-CoV-2 infection this weekend.
Comments