What if virus math wrong?

Editor's Note: We are posting select coronavirus-related stories and opinion pieces that are open to non-subscribers. We hope you will consider subscribing to our eEdition to help support local journalism in Columbia: https://www.columbianprogress.com/subscribe

The tightening grip of restrictions on American communal life, the intentional tanking of the U.S. economy, the coming $2 trillion bailout in mostly borrowed money to try to soften the nearly certain recession, all follow the premise that the coronavirus is a health threat the likes of which this nation has not seen in a century.

Most of the doomsday projections being cited to support the partial lockdown of Mississippi and the rest of the country are indeed startling: One half to two-thirds of the nation infected; as many as 1 out of 16 Americans requiring hospitalization; 200,000 to 1.7 million deaths.

But what if those numbers, which originated with the U.S. Centers for Disease Control and Prevention, are wildly wrong? What if the toll, as President Trump said before he was convinced otherwise and is now steering back toward, is about like a bad case of the seasonal flu? Would all of this trauma inflicted on the nation’s pocketbook and psyche be worth it?

Those are the types of questions being raised by John P.A. Ioannidis, a Stanford University professor of medicine, epidemiology, biomedical data science and statistics. One of Ioannidis’ specialties is doing research on research. He regularly finds that it does not meet good scientific standards of evidence.

Ioannidis does not say he can calculate with any certainty what the infection and death rates will be in the United States from COVID-19, but he says no one else can either. That’s because there is no reliable data available on how prevalent the virus actually is, and how many it kills, in a sample of the general population. For example, the World Health Organization estimates a mortality rate of 3.4%, although the documented figure so far is 4.4%. In the United States, the mortality rate has been about a fourth of that, 1.2%.

All of these numbers are inflated, though, because the vast majority of those being tested for the virus are disproportionately those with severe symptoms and bad outcomes. Consider Mississippi’s testing standards, which are similar to those throughout the country. Because test kits are having to be rationed, only those with fever and a severe cough or trouble breathing are being tested.

One situation where an entire, closed population was tested was the Diamond Princess cruise ship, the vessel that was quarantined for nearly a month off the coast of Japan. Of the 712 crew members and passengers infected, eight have died, roughly 1.1%. Because the travelers skewed older and because the data pool was small, Ioannidis said the fatality rate to project from that sample on the general U.S. population could be as high as 1% or as low as 0.05%. In other words, the range could be anywhere from 1 in 100 infected Americans dying to 1 in 2,000.

Without good data, there is no scientific way to project the true death rate. But if it winds up being similar to the flu, what the world and U.S. have done to slow the transmission of the coronavirus — that is, “flatten the curve” — has been irrational and possibly counterproductive. Instead of overwhelming the health-care system quickly and getting it over with, hospitals could be overwhelmed for not just months, but maybe years. Ioannidis compares the potential fiasco to an elephant being attacked by a house cat. “Frustrated and trying to avoid the cat, the elephant accidentally jumps off a cliff and dies,” he writes.

Only problem with his analogy is that it would be no accident. It would have been an intentional, though unwitting, series of decisions that ultimately did more harm than good.