The Changing Models Show: One Size Fits All Isn't A Good Strategy For Coronavirus

People observe from their window a crucifix displayed outside the Santa Maria della Sanita church during Good Friday, as Italy remains on lockdown during the Easter period to try and contain the spread of coronavirus disease (COVID-19), in Naples, Italy A
April 12, 2020 Topic: Security Region: Americas Blog Brand: The Buzz Tags: CoronavirusCOVID-19PandemicEpidemiologyModeling

The Changing Models Show: One Size Fits All Isn't A Good Strategy For Coronavirus

It's becoming more clear as the data gets better.

The Institute for Health Metrics and Evaluation at the University of Washington provides a forecasting model respected and used by the White House Coronavirus Task Force to plan the response to the COVID-19 pandemic. Last Sunday it revised its fatality projections downward to 81,766 deaths (range 49,431 to 136,401) by early August, with the peak day occurring on April 16. It now projects 60,401 deaths (range 31,221 to 126,703) by August 4, with a peak day of April 11.

This is very encouraging news. As Dr. Anthony Fauci of the task force previously stated, “Models are as good as the assumptions you put into them.” And as more data come in and we develop more knowledge about the virus’ impact, the assumptions put into the models change and the projections change with them.

These improved projections are little comfort to the people living in the New York/​New Jersey metropolitan areas, as well other highly impacted areas such as New Orleans or Detroit, whose health systems are overwhelmed and whose fatality rates are striking, But it is important to bear in mind that different portions of the country are experiencing different case rates and fatality rates.

As the Wall Street Journal pointed out in its April 8 editorial, demographic and geographical differences probably explain a lot about these variations. That’s why I have argued that policies implemented in response to the public health emergency should not be one‐​size‐​fits‐​all, but should be customized to fit the facts on the ground in various regions of the country based upon local knowledge. California, which began its statewide lockdown on March 22, just three days before New York, had 485 fatalities reported on April 8, compared to New York which had 6,268 fatalities. It is reasonable to ask if the same policies that are appropriate in New York are completely necessary in California. It is also reasonable to ask if perhaps Californians were exposed to COVID-19 months ago, and herd immunity may have already developed there, slowing the virus’ spread. (More on that later.)

Speaking of local knowledge, on March 16 the Food and Drug Administration responded to criticisms that its cumbersome regulatory process is slowing the development and release of COVID-19 tests by delegating authority to the states to approve tests within their borders.

Public health officials are hoping that tests for antibodies, which indicate if a person has already had the infection and therefore developed immunity, will soon become widely available. This is key to a plan in Germany: people there who display immunity will be given “immunity certificates” and allowed to go back to work. They can’t become infected and can’t infect any contacts.

The FDA finally gave emergency use authorization for an antibody test made by the biotech company Cellex on April 4. However, it specified that the manufacturer must state on the test label that it is not formally FDA‐​approved, and testing is limited to FDA‐​certified labs. The tests are not available to private labs or for home testing. They should be available soon.

While we wait—and exercising the authority delegated to it by the FDA—California, on April 1, allowed an antibody test made by Premier Biotech to be used by ARCpoint Labs, a commercial laboratory in Monterey, CA. The test is not FDA‐​approved.

And on April 3 researchers at Stanford University School of Medicine began a study using that same Premier Biotech test, hoping to learn if California’s low fatality rate might be the result of herd immunity. California is the number one U.S. tourist destination for people from China. Chinese health authorities recently revised the date that they believe the COVID-19 virus came on the scene in Wuhan, from early January to early November. The Stanford researchers believe it might have developed even earlier, and was being brought to California by asymptomatic or mildly symptomatic Chinese tourists in the fall—when California doctors were reporting an earlier than usual flu season.

On April 3rd and 4th the Stanford researchers performed the test on 3,200 people at sites in San Jose, Los Gatos, and Mountain View. The tests just take a few minutes to run. They hope to complete the study and release their findings in the next few weeks. Meanwhile, ARCpoint Labs has already conducted 500 of the tests, with some positive results, and is sending in all of the results to the Monterey County Health Department.

If it turns out that Californians are approaching herd immunity, life there can return to normal sooner than the national models would otherwise predict.

California provides a good example here of how decentralization and deregulation, using local knowledge, can produce rapid and community‐​specific responses when unexpected emergencies arise.

This article by Jeffrey A. Singer first appeared in CATO on April 4, 2020.

Image: Reuters.