Assuming average U.S. prevalence is between 4 and 10%, I wonder if all our antibody test-based studies outside of NYC would be similarly impacted by false positives, resulting in low IFR estimates. It would be interesting to compare findings by region and see if there is a direct relationship between prevalence and IFR estimates.

While IFR is interesting to me as a dork, I also see how it's possible to get bogged down in details. We'll eventually be able to narrow confidence intervals, including for different age groups and other demographic variables. For now we have a lot of evidence to support using a variety of methods to avoid uncontrolled transmission. As well as cautionary reports (some domestic) of how failing to avoid it sickens and kills people and disrupts businesses.

I've shifted focus of late to challenges with people refusing to wear masks and practice common-sense distancing, particularly indoors. Until we have better treatments and ideally a vaccine, those actions endanger our freedom. Being sick, dead, or out of work due to lockdowns in the face of widespread sickness and death is a form of imprisonment.