Tuesday, June 16, 2020

Deaths Data

For the 2009 fall semester the University asked me to add to my teaching load one of what it was calling "First Year Seminars." These were small-section classes for entering first years taught by selected professors from a variety of colleges and departments, with subject matter of the professor's choice.

My choice? General Semantics. It probably had not been taught at Iowa since my father's lectures in the 1940s and '50s, when it was said to have been one of the most popular among students, and chosen for broadcast throughout the coverage area of the University's AM radio station, WSUI.

My book, What Do You Mean? and How Do You Know? An Antidote for the Language That Does Our Thinking For Us [Prairie Lights; Amazon], created for that seminar, consisted of four chapters from my father's writing and eight of my own. (Two K-12 teachers, present for my presentation with that title at a Herbert Hoover Presidential Library event in 1997, used these ideas in their teaching "American History from the Perspective of Minorities and Women," which course was described in a third person's doctoral dissertation and subsequent book.)

This is not the time to attempt a full explanation of general semantics. Perhaps that should be saved for an essay on its relevance as a necessary first step toward the elimination of systemic racism. But this brief quote at the beginning of the first chapter may be useful:
"General semantics" is not about "semantics" -- as in the expression, "Oh, you're just arguing about semantics." It's not about "defining" words, their historical etymology, or cognates. General semantics deals with language as human behavior; for example, the disparity between what we say (that largely reflects what's going on inside our brain) and the "reality" we think we're describing -- and the consequences of that confusion.
With this long introduction, and short explanation, let's take a look at "what we mean" and "how we know" about the number of Iowans' deaths from COVID-19.

What Do You Mean?

For the most part we know what we mean by "dead," the word and the reality are fairly clear (though there is some ambiguity in cases of "brain dead" and other variations).

The official State of Iowa "Covid-19 in Iowa" site labels its data under "Current Cases": "Individuals Tested," "Individuals Positive," "Total Recovered," and "Total Deaths."

We can assume that "deaths" means an individual's death was in some way associated with COVID-19. If a person has tested positive for COVID-19, had symptoms associated with the virus, no other medical conditions weakening their immunity, and died, it would be difficult not to add their number to "total deaths." But what if they had not been tested before they died, but had one or more of the symptoms of COVID-19 at the time of death? Are they part of "total deaths"? What if (whether tested or not) they had COVID-19 symptoms that weakened their immunity, but also preexisting conditions (say, pneumonia)? Is whoever determines "cause of death" free to report they died of pneumonia and not record it as part of "total deaths" (from COVID-19)? In other words, does it make a difference in the counting whether someone has died from COVID-19 or just with COVID-19 -- along with, perhaps, many other conditions? [Chart credit: Nicholas Johnson, using State of Iowa data.]

Where and by whom, within the healthcare system, is the determination made as to whether a deceased should be counted among "total deaths" or not? Is there, somewhere, written instructions regarding these choices? Are they publicly available online? If so, is there any oversight of the compliance and noncompliance with these procedures and definitions?

How Do You Know?

There are powerful incentives for elected officials (state, county and local), meat packing plants, nursing homes, prisons, and other institutions to minimize their responsibility for the spread of this global pandemic, Americans infected, with the stresses it has created on our hospitals, healthcare workers, families and businesses of all sizes -- not to mention the more than 116,000-and-rapidly-climbing deaths of Americans.

The Iowa governor announced that nursing homes need not report cases or deaths unless they had an "outbreak" -- which she defined a three or more cases. The Nebraska governor explained that packing plants did not have to report at all. "Vice President Mike Pence encouraged governors on Monday to adopt the administration’s claim that increased testing helps account for the new coronavirus outbreak reports, even though evidence has shown that the explanation is misleading." "Pence Tells Governors to Repeat Misleading Claim on Outbreaks,"New York Times, June 16, 2020.

What is the process by which (and by whom) deaths -- in general, not just COVID-19-related -- are reported? Who reports, and to whom, when someone dies in a hospital, at home, or elsewhere? What oversight is there for these alternative reporting channels -- hospitals, long term living and nursing home facilities, funeral homes, those performing autopsies, police at the scene of an accident causing death, someone who dies at home?

Where can one go online to find, say, total deaths in Iowa for the past five years, by months? (The latest I found was 2017, broken out by conventional causes of death.) An alternative way to measure COVID-19-related deaths would be to look at Iowa's mortality statistics for February through June for 2016-2020. One would need to correct for variations in 2020 data from prior years' averages; for example, a decrease in deaths from automobile accidents during 2020 due to less driving during the pandemic. Having done that, it would be possible to see any increase in total deaths during those months in 2020 that could rationally be related to COVID-19. As such, it could be one way of confirming (or questioning) the State's official "total deaths" statistic.

At a time when trust is in increasingly short supply, one recalls President Ronald Reagan's advice, borrowing the Russians' expression, "trust, but verify." One of the first steps in verification is to ask the two questions, "What do you mean?" and "How do you know?"

# # #

1 comment:

Nana said...

Excellent post! I hope someone is doing this kind of analysis of reported deaths in Iowa.