A problem with comparing coronavirus deaths to flu deaths

It’s fairly common to compare deaths attributed to the Wuhan coronavirus to deaths attributed to the flu in a bad season. The number used for flu deaths in a bad season is typically around 61,000, based on the CDC’s estimate for the 2017-18 season. The number of deaths attributed to the corornavirus is around 65,000 and rising fast.

Attempting the comparison is reasonable — certainly more so than comparing coronavirus deaths to those resulting from auto accidents, the Vietnam War, and even health-related events like heart attacks. Coronavirus deaths, like those from the flu, are the product of an infectious disease. In theory, the comparison approaches apples to apples.

But there is a problem with using flu-related deaths as a comparator. The coronavirus number is a count, albeit an imperfect one. The flu number isn’t a count, it’s the product of equations.

That is the gist of this article by Robert VerBruggen, which cites this piece in Scientific American by Jeremy Samuel Faust. VanBruggen says “the CDC’s reported number of actual confirmed flu deaths — that is, counting flu deaths the way we are currently counting deaths from the coronavirus — has ranged from 3,448 to 15,620” a season.

How do we get from a high of less than 16,000 to a high of 61,000? Via adjustments. VerBruggen quotes this explanation by the CDC:

First, we adjusted the reported annual hospitalization rates from FluSurv-NET [a system that collects flu data from a small subset of U.S. health-care providers] . . . using multipliers that included the probability of being tested for influenza and the sensitivity of influenza testing. . . . Rates of influenza mortality were calculated by multiplying the adjusted rates of hospitalization by the ratio of deaths to hospitalizations.

How is the “ratio of deaths to hospitalizations” determined?

Not all persons who die with influenza are admitted to a hospital prior to their death, and others may die after hospital discharge, thus hospital surveillance does not fully capture deaths due to influenza in the catchment area. To estimate a more complete ratio of deaths to hospitalizations, we also included data on the probability that a person with a respiratory infection would die outside of a hospital admission. For this we used publically available mortality data from the National Center for Health Statistics for the U.S. population in 2010 to identify the deaths attributable to pneumonia and influenza (ICD-10 codes: J10-J18) and the proportion that occurred while hospitalized vs. outside of a hospital admission (e.g., at home, on arrival, in the emergency department, in hospice or long-term care facility).

Faust criticizes this methodology. He claims it seriously inflates the number of flu deaths. I can’t evaluate this contention but it does seem odd that the actual number of deaths from the flu in a given season would be four times higher than the observed number. And, says Faust, the CDC acknowledges that its count includes deaths from pneumonia.

The coronavirus death count is also imperfect. The imperfection results from ambiguity in some cases about whether a particular death was due to the virus or some other cause.

This is a different kind of imperfection than one caused by the use of multiple equations based on multiple assumptions. In that sense, comparisons between flu deaths and coronavirus deaths aren’t really apples to apples.

Responses