Making sense of COVID-19 fatality rates

The case fatality rate for the seasonal flu is about 0.1 to 0.2%. It means that for every 10,000 people with a documented case of the flu, 10 to 20 will die.

This method of calculation is often what people use if they casually cite a death rate. But it's not the most accurate way to estimate your personal risk of dying if you become host to the agents of a disease. That would be the infection fatality rate. For the flu, by one account, it's 0.04%. That means for every 10,000 people who are infected, whether they report it or not, four ultimately will perish.

You'll see that the extra numeral after the decimal point makes a difference. But you'll also notice that we've gone from a rate that can be calculated with hard data to one that must be inferred.

Store those figures in your temporary memory. They'll be relevant for comparison as Minnesota considers not just whether to lift orders to stay at home during the COVID-19 outbreak but also how best to arrange things if it does. The current order is set to expire next Monday, so another update is coming soon.

Fatality rates are not fixed. They can vary by patients' ages and overall health, among other things, and they can be bent by what medical professionals learn about treating a disease. The circumstances around this new strain of coronavirus have been nothing if not nebulous, but two months into the arc of the pandemic, we can begin to discern a shape.

As of Monday in Minnesota, 11,799 cases of the disease had been confirmed by laboratory testing. There were 591 deaths. Case fatality rate? Simple math. Divide the smaller number by the larger, and you get 5%. That's in range of the 6% figure globally for people with a diagnosis.

Infection fatality rate? Here you must make an assumption: How many unreported cases might there be? Last month before testing ramped up in Minnesota, state officials thought it could be 100 times more than reported ones. Provisional antibody tests in other parts of the country have put it closer to 10 times more. This results in an ambiguous estimate for the rate — anywhere from 0.05 to 0.5%.

Those arguing against stay-at-home orders are banking on lower rates for COVID-19 in the end analysis. Figures of less than 1% for the case fatality rate and as low as 0.1% for the infection fatality rate have been suggested.

So far in Minnesota, about 4 of every 5 fatalities have occurred in nursing homes or assisted-living facilities, and 99.24% of those who have died had an underlying health condition (as discussed in an editorial Sunday). It is indeed plausible that the rates will settle as the virus works into less vulnerable segments of the population. By how much?

There's an example in which nearly an entire population was tested: that of the Diamond Princess cruise ship, which had 3,711 passengers and crew members aboard and was the subject of a follow-up study. (Though not an exact comparison to poll methodology, that number of people would be a statistically significant sample size if applied to Minnesota's population — for a recent Star Tribune/MPR News Minnesota Poll, 800 people sufficed.) As a caveat, a cruise ship can be a rather ideal petri dish, but among those 3,711 people, the case fatality rate was 2.6%, and the infection fatality rate — which includes those who had no symptoms at the time of the test — was 1.3%.

Outcomes can depend on how much virus a person is initially exposed to. The implication is that even as restrictions are lifted or adjusted, distance will matter. May keeping our distance be as easily done as it is said.

Minnesota's stay-at-home order began March 27. When weeks stretch to months, anyone's resolve can be tested. But for now — despite the skepticism, despite the protests — it appears that most Americans prefer to go slow in the face of uncertainty. A collection of recent polls found support for an immediate reopening of the economy ranging from 10 to 35%.