The United States is being told to brace itself for anywhere from 100,000 to 2.2 million deaths from the COVID-19 epidemic. These numbers are completely unreliable because we don’t have adequate baseline information from which we might have made informed estimates, nor is it all clear over what time period we are being told to anticipate these deaths. A 100,000 deaths in one month would be a catastrophe. Two million deaths over a two year period would be tragic (especially for the victims and their families) but wouldn’t be catastrophic.
There are two reasons for this lack of real information: a basic lack of data and poorly expressed public statements that aren’t rigorous enough to be reliable.
The COVID-19 death tolls now being reported around the world are specifically in-hospital deaths attributed to the virus. This is because out-of-hospital deaths are not being tabulated by anyone with a vested interest in reporting the statistics. (Coroners offices generally do this, but the collection of cause of death statistics has always been rather spotty and no on is listening to them anyway.)
We actually have no idea how many people have been exposed to the COVID-19 because we only know how many people got sick enough to go to a testing site to be tested for the virus.
In order to determine the actual mortality rate for this disease, we would have to test the entire population ON THE SAME DAY (which is an obvious impossibility) but even a ten percent coverage of the population would give us enough of a statistical baseline to make some informed assumptions about the actual extent of the epidemic. Unfortunately, we don’t have the capacity to do even that much. This leaves us with only the people who have actually made it to a hospital as the baseline, but that also limits the baseline to the number of people the hospitals have the capacity to handle, which is an unknown number since it keeps changing from day to day.
We really have no idea how many people have died from the virus who were not hospitalized before their deaths. We have no idea how many people have been exposed to the virus because we haven’t done any testing of people who have not had obvious symptoms and therefore we do not know what the real rate of infection is, nor do we really know what the real death rate is because we can only determine that once we know how many people have actually been infected.
Statisticians like to make do with the data they have, so they concoct tables and charts based on what we do know, but they rarely ever mention the potential sources of error built into their statistics and, when they do issue such caveats, the chances are that the reporters covering the stories will not mention them. This is because mentioning those sources of error undermines the validity of their reporting, which is based on the statistics they have been given, not the statistics they have not been given.
According to a 2020 UN study released before the epidemic began, the projected death rate for the United States was 8.880%. That works out to an average of 2,427,200 per month. Therefore, we can estimate that over the first three months of 2020, 7,281,600 Americans have probably died of OTHER causes.
This placed the United States at number 28 on the UN’s partial death rate list, below Germany, Japan, Portugal, Italy, Finland, Denmark, Belgium, the United Kingdom, France, Spain, Sweden, and the Netherlands. The lowest death rate on the list belonged to Qatar, where the death rate is just 1.298%. (Maybe we should start comparing our medical care to Qatar’s rather than Sweden’s.)
The total population of the United States is now approaching 328 million. The total number of deaths to date from the virus here in the United States has now reached 4,000, which means that the death rate (deaths from virus/total population) is around 0.0000122. That’s 1.22 Hundred-THOUSANDS OF ONE PERCENT. The four thousand deaths to date from the virus works out to 0.0005493 of the total deaths from all other causes or around five ten-thousandths of one percent of total deaths.
In other words, so far, to date, the coronavirus epidemic hasn’t changed the death rate in the United States by more than five-ten-thousandths of one percent, a statistically irrelevant number.
The number that the media has focused on is the ratio of reported cases to the number of deaths. As of April 1, the CDC is reporting 186,101 cases and 3,603 deaths, for a death rate of 0.0193605 or 1.93%, which means that if you are hospitalized because of a COVID-19 infection, your chances of survival are presently 98.1%. That number tends to go down as your age increases but that is only because of the increasing prevalence of underlying conditions among older patients. Recent reporting is now indicating that an increasing percentage of younger people are dying from the Coronavirus. Authorities profess to be perplexed by this development which is obviously due to the fact that younger people aren’t taking the same precautions as older folks are because younger people are, by nature, risk takers.
The number of cases is significant because that tells us how well our containment efforts are working. The number of deaths is not significant, except to those who have been touched by those deaths, because we cannot cross-correlate the number of deaths according the nature, scope or extent of the interventions for each of those deaths.
The COVID-19 disaster is actually an example of what happens when incompetent or self-interested people play with statistics. It is also a demonstration of what happens when factions manipulate data to achieve political ends.
Liberals will use the numbers to beat Donald Trump up as often as they can, and the more power them. Conservatives will continue to privately question whether killing off two million Social Security recipients wouldn’t be a good thing for them. (It won’t be. Killing off your best customers is never a good idea.)
In the end, however, we will continue shooting in the dark, grasping at straws like wearing masks, spraying alcohol on everything while waiting for our symptoms to emerge.
93 total views, 1 views today