There has been much in the news about why the COVID-19 virus (officially known as SARS-COV2) death rates (also known as mortality rates) differ so much by country, and that the true infection rates are probably vastly undercounted.
The only way to really understand the mortality rate, and true COVID-19 infection rates overall, is to test people systematically, whether they are sick or not. Ideally, we would test everyone, if that were feasible. In a country as large as the US, instead an epidemiologist would try to target a representative sample – i.e., a subgroup who ‘looks like’ the greater population as a whole with respect to age, gender, ethnicity, and many other risk factors.
Right now, in the US, only the sickest COVID-19 patients are able to get tests because of the shortage of tests and delays in obtaining results. Here is a very simple example to help illustrate why testing as many people as possible, whether sick or not, is important:
Testing Only the Sickest:
In one neighborhood, 105 people have high fever, cough, and shortness of breath. They all get tested, and of the 105, 100 of them are positive for the COVID-19 virus. Unfortunately, 3 of them die.
Mortality rate: 3/100 = 3%
Infection rate: 100/105 = 95%
This infection rate seems astonishingly high, but only people who are highly likely to be sick with COVID-19 actually got the test.
However, if you were to test everyone (or nearly everyone) in a small city, regardless of symptoms, you might see something like this:
# People tested = 12,500
# People found COVID-19 Positive = 10,000
# Deaths from COVID-19 = 50
Mortality rate: 50/10,000 = 0.5%
Infection rate: 10,000/12,500 = 80%
This “Testing Everyone” example shows why it’s so important that more people get tested. While obtaining a more accurate death rate is very important, one needs to know what proportion of the population has the infection, even if they don’t know it. If someone knows they are infected, they can quarantine to prevent further spread.
Countries such as South Korea and China weren’t able to test everyone, of course, but they were able to test much larger numbers of people (and get results more quickly), which helped them identify cases quickly to encourage self-quarantine to reduce further spread. Of course the reasons for their reduction in cases are more complex than just testing, but it is a major component of their successful control.
Many epidemiologists who are intensely researching COVID-19 in the US right now estimate that we are probably undercounting actual infections by ~10-11x (Li, R., et al, 2020). In other words, for every 1 person that is COVID-19 positive, likely there are another 9-10 people with the infection who may have only mild or no symptoms, spreading it to others.
Please be safe, stay home as much as possible, and remember to wash your hands often and well.
While this COVID pandemic situation is extraordinary, I feel fortunate to be able to use my epidemiology training to help educate why testing is so important. UBC® advises clients on a range of epidemiology topics, particularly using observational designs and various real world data (RWD) sources to better understand the safety, effectiveness, and usage patterns of medications, and to better understand diseases and patient populations more generally.
To learn more about our epidemiology capabilities, contact us.
© 2020 United BioSource LLC, all rights reserved.
UBC® is a trademark of United BioSource LLC.