The six percent revelation

August 31, 2020. The media are aflutter with attempts to walk back additional information about Covid deaths. A good place to see the basic problem is here.

It’s not clear to me as to disingenuous or misleading: Is it that the Detroit Free Press is using Politifact as a substitute for journalism? Or the position Politifact takes? Or that PoliFact has a ‘partnership’ with Facebook to ‘combat false news and misinformation.’

Here is the key claim cited in the article:

"For 6% of the deaths, COVID-19 was the only cause mentioned. For deaths with conditions or causes in addition to COVID-19, on average, there were 2.6 additional conditions or causes per death."

That data comes [NOTE: 'Data' is a plural noun.] from death certificates.

The key takeaway here is that in only six percent of the Covid deaths were there no co-morbidities, period. Just about every other claim is speculation or an opinion by one ‘expert.’ That is the fact. Its interpretation is open to opinion, but this fact remains. And this six percent figure is a clarification or revision of the previous, more ambiguous claims. Specificity helps. Adding specificity is a revision.

It is fair to deduce that people who die WITH Covid may not have died OF Covid. At a minimum, the presence of the co-morbidities in the vast majority of Covid deaths suggest a contributing interaction effect. And in terms of misinformation, counts of ‘positive tests’ for Covid are clearly overestimated, given the unreliability of the tests being used and the reports of false positives. This week the New York Times claimed that up to 90% of positive tests could be false. Last week, the NFL reported that all 77 positive tests collected from its players turned out to be false. So, an educated, informed opinion would note that even the claims of any Covid factor in many of these deaths is based on inexact language or interpretation of ‘fact’ or at least a clear misunderstanding of test reliability or accuracy.

Reporting causes of deaths is an inexact, complicated process, often more political, personal or economic than scientific. Asking those filling out the death certificate to use their ‘best judgment’ in completing the document opens the data to biases and unreliability. It is not even clear that there needs to be a positive test for Covid to add Covid as a cause of death. And no autopsy. Test results are not only unreliable, they take often days before the results of the test are available.

Driving to Mass Tuesday evening I was struck by two news reports, one from a doctor – ‘expert’ – commenting on the “six percent” data and the second, the daily litany of ‘confirmed’ positive Covid cases.

The doctor claimed that the six percent figure was surprisingly high, as Covid – in his estimate or experience -- was behind numerous co-morbidities, such as a patient entering the hospital with Covid who subsequently died of pneumonia ‘caused’ by Covid. To be clear, he provided no specific statistics, just some examples, perhaps hypothetical. He also had not apparently actually read the 150,000 or so death certificates in making his speculations, er, offering his ‘expert’ opinion. The radio reporter claimed that this was an effort by the doctor to ‘clarify’ the six percent claim.

Besides his lack of data, the doctor ignored the list of primary co-morbidities associated with ‘Covid deaths,’ such as hypertension, lung disease, obesity and diabetes. And death certificates include 'Covid' deaths accompanied by dementia, suicide. Covid caused none of these conditions. At the least he could recognize the prevalence of pre-existing co-morbidities. Again, no data, just anecdotes.

And this leads to the second, fear-mongering report about ‘record’ numbers of ‘confirmed’ positive tests in Wisconsin. But considering the cases of false positive tests for Covid, consider this: Has there been a single report of a patient who tested negative for Covid who then died of Covid two weeks later, i.e., the patient had a false negative Covid test? Probably not. On the other hand, false positive test reports are common and, as noted above, could represent as many as 90% of the positive tests, i.e., there might be ninety percent fewer cases of Covid than those reported by the ‘confirmed’ tests.

The reporter was lazy and failed to note how many tests had been run, period. Nor did her note the percent of tests that were ‘confirmed positive.’ Absent this context, it is no wonder that the best argument for reducing Covid cases may be simply to stop administering the tests other than to very sick people, those with strong indications of the illness.

So what does this all say? Here is one good summary.

When the dust settles on the year 2020, we will finally accept that the United States shut down the greatest economy in the world, that the governors in the worst hit states issued executive orders turning their nursing homes into death traps, and that we significantly increased deaths from otherwise treatable illnesses — all because of a bad flu.