Monday, January 4, 2021

A problem in practical epistemology

Co-Morbidity and the Pandemic

 

            Jack: They say Mrs. Black died from Covid-19, but she had diabetes and was overweight.  But she tested positive for Covid-19, so they put that on her death certificate.  It makes me suspicious of the numbers they report on the news.”

            Jill: “No kidding.  You don't know who to trust anymore.  I read an essay by a doctor who said the pandemic is already over, because people are developing natural immunity.  And the vaccines they are promoting are completely new; no truly long term safety studies.  So I’m not going to be in any rush to take a vaccine.”

 

            Fictional "Jack and Jill" illustrate a problem in philosophy.  They find themselves in doubt, troubled by skepticism.  In philosophical jargon, they have a problem in practical epistemology.  That is, they don't know what to believe about an important question, and that question has important implications for what they will do.  Never fear!  What follows will not be an essay in epistemology.  I will examine just one tiny aspect of the Covid-19 pandemic.

            Pretty obviously, there are many causes of death.  People die from drug overdoses, car or airplane crashes, drownings, falls, murders and other sudden causes.  Usually, diseases kill more slowly: heart disease, influenza, cancers of various kinds, strokes, and many more.  “Cause of death” gets a bit tricky when a person suffers from multiple diseases we recognize as typical causes of death.  A recently deceased person had heart disease and cancer and high blood pressure; while driving late at night he crashed into a logging truck.  What was the cause of death?  The crash?  His loss of consciousness which caused the crash?  The high blood pressure and heart disease that may have caused a stroke?

            “Co-morbidity” describes cases in which persons suffer from multiple deadly causes.  Doctors who prepare death certificates for recently deceased persons often have a clear choice.  (Dr. G wrote “uterine cancer” on Karen’s death certificate.)  But it’s easy to imagine cases in which various causes “cooperate” in killing some person.  The patient’s lung cancer probably would have killed him in a few years, but because his breathing was so impaired what should have been a mild case of flu ended his life.  Which killed him, the underlying cancer or the flu?

In cases of co-morbidity the right answer could be either.

            Most cases of Covid-19 death are cases of co-morbidity.  From the beginning of the pandemic, medical experts have warned that the disease is most dangerous for people with pre-existing conditions.  This is why most Covid-19 deaths have been among the elderly; they have lived long enough to accumulate pre-existing conditions.  In the United States there have been more than 350,000 Covid-19 deaths; it is no surprise that lots of people “know someone who knew someone”—cases in which the official cause of death was Covid-19 for a person with some other deadly disease.  My fictional Jack knew such a case.

            Does this mean we should be skeptical of reported Covid-19 deaths?  No.  Doctors and statisticians at the Centers for Disease Control are well versed in co-morbidity.

            But suppose you are like Jill.  Skepticism clouds your attitude toward authorities like the CDC.  How can they know which deaths are properly attributed to which disease?  Aren’t their totals subject to the prior decisions of attending doctors?

            Well, no.  There is a different way to analyze the problem.

            In recent years the United States deaths have numbered around 2,700,000 or 2,800,000.  The number of deaths increases almost every year as the total population increases.  In raw numbers, deaths increase by about 20,000 or 30,000 each year.  But in 2020, the total number of deaths in the US went up by at least 300,000.  The death rate in the United States in 2020 is far higher than we ordinarily would expect.  Increased gun deaths or opioid overdoses will explain some small part of the increase, but Covid-19 has to be the cause of at least 250,000 excess deaths.  (At least.  The true number will be higher.)  Now, 2020 is not the most extreme year on record.  In 1918 the combination of “Spanish” flu and deaths of soldiers in World War 1 produced the all-time spike in death rate.

            Full disclosure: the final numbers of deaths for 2020 have not been finally reported.  There have even been internet claims that deaths in 2020 are pretty much the same as 2019.  And (for those skeptical of the CDC) the official numbers are collected by the CDC.  When the final numbers are published in a few weeks, 2020 deaths will probably exceed 2019 by more than 400,000.  But if you think the CDC is inflating numbers, what can you do?

            You can wait.  The pandemic will continue to kill tens of thousands in 2021.  But the vaccines (which are, according to the FDA, safe and effective) will slow the pandemic as the months go by.  By 2022, the pandemic should be over.  And that means the American death rate should fall in 2022.  (It’s possible, of course, that a foreign dictator could launch nuclear missiles at us.  In that case, and other such scenarios, all bets are off.)  It’s fairly easy, for those of us who trust CDC experts, to predict that the raw death numbers in 2022 will actually fall.  And that’s even with the usual overall population increase.  Covid-19 has co-morbidly carried off so many people that we will actually see a decrease in death numbers.

            Of course, if you are totally skeptical, you should say deaths in 2022 is inscrutable—totally unknowable.  So what do you think?