Press "Enter" to skip to content

Are counties cherry-picking coronavirus test numbers to look better?

How many actual human beings have been tested for coronavirus in California?

Surprisingly perhaps, that’s hard to say.

Some officials seem to tell you outright: In Los Angeles County, 611,950 unique persons were tested as of May 31, according to officials and the county’s data dashboard. In San Bernardino, 62,989 patients were tested as of June 1.

In Orange and Riverside counties, though, it’s less clear. Both counties report the number of tests performed, not the number of people tested.

What’s the difference? Consider a front-line health care worker who’s been tested 25 times. If you’re counting the number of people tested, that’s one. If you’re counting the number of tests, that’s 25.

A man wears a ‘COVID-19 is a hoax’ mask as he joins thousands of people at the corner of Main Street and Pacific Coast Highway to protest coronavirus (COVID-19) closures in Huntington Beach, CA, on Friday, May 1, 2020. (Photo by Jeff Gritchen, Orange County Register/SCNG)

Counting tests, rather than people tested, can make it appear that testing is quite robust — and that, in turn, can be used to justify a swifter reopening of the economy.

“Maybe I’m ignorant to the details, but the only beneficiary of reporting total tests would be those that are failing at actually expanding the testing network, failing to implement contact tracing,” said a number-cruncher who worked for state, but doesn’t want to be named for fear of reprisal.

“Creating a false image of a virus under control (high tests, low cases) would benefit those desiring economic recovery before we have even begun to respond adequately.”

Experts said that it’s not an either-or question. Both numbers are important, and both should be tracked.

Still, charges of cherry-picking data are not exactly new. In Florida, a woman who maintained that state’s COVID-19 database claimed she was fired after objecting to reporting changes that she said were designed to justify an early re-opening.

The California Department of Public Health, which collects COVID-19 numbers for the entire state, counts the number of tests given, not the number of people tested. That allows officials to keep an eye on the state’s total testing capacity, a spokesperson for CDPH said.

As far as the state is concerned, individual counties can choose to report whichever number they wish — total individuals tested, or total tests administered.

How many tested?

In Orange County, 136,098 tests have been administered as of June 1, to an unknown number of people, according to officials and the county’s data dashboard.

In Riverside, 119,824 tests have been administered as of June 1, also to an unknown number of people, according to officials and its data dashboard. But Riverside reports total individuals tested to its Board of Supervisors regularly, a spokesman said.

In Orange County, the number of individuals actually tested wasn’t always a mystery.

“Very early on, everyone was running multiple tests (swabs) on the same person as part of the testing, so we were cleaning the data to only report the number of unique persons tested,” said an email from Marc Meulman, operations chief for O.C. Public Health Services. “Once case counts picked up, that changed to one swab per person.”

Change kicked in as of May 6, according to the county’s data dashboard.

“More recently, there has been a significant push for total tests, including repeat testing of first responders, especially health care workers. Each test now is a unique effort to determine infection, so it makes more sense to report total tests,” Meulman wrote.

He gave the example of a worker at a skilled nursing facility, where the environment is high-risk. That worker needs to be tested regularly, and all those tests are important to count, the email said.

“As there is more and more serial testing of asymptomatic essential workers, this becomes even more important so we are capturing all that testing.”

Because of this, though, “we won’t know how many individuals received tests.”

Vital data, MIA

Reporting the number of tests — without also reporting the number of people tested — is misleading and encourages a false sense of security, several experts said.

“Excuse my language, but that’s a half-a–ed way of doing it,” said Jerika Lam, an associate professor of pharmacy practice at Chapman University. “I see where they’re coming from — they’re looking at the low-hanging fruit. What’s easiest to capture?”

A health care worker prepares to collect a sample to test for COVID-19. (AP Photo/Wilfredo Lee, File)

Certainly, staffing is tight and funding is a huge issue right now for governments at all levels, “but if I distribute 1,000 tests a day, it doesn’t mean that 1,000 people got tested,” she said. “That’s what people assume when they see those numbers, but that’s false. It’s incorrect.”

Georgiana Bostean, an associate professor of environmental science, health and policy at Chapman, pointed to the expertise of Johns Hopkins University — which has emerged as the nation’s definitive data aggregator in this pandemic — and says it’s imperative to know both numbers.

“It is important to track the testing that states are doing to diagnose people with COVID-19 infection in order to gauge the spread of COVID-19 in the U.S. and to know whether enough testing is occurring,” Johns Hopkins says.

“When states report the number of COVID-19 tests performed, this should include the number of viral tests performed and the number of patients for which these tests were performed. Currently, states may not be distinguishing overall tests administered from the number of individuals who have been tested. This is an important limitation to the data that is available to track testing in the U.S., and states should work to address it.”

Like many raw counts, the total number of tests doesn’t really tell you much by itself, especially given the population size across counties and the fact that some people are tested multiple times, said Richard M. Carpiano, professor of public policy and sociology at UC Riverside.

Knowing the number of tests per capita would at least enable comparison across counties; and knowing the total number of tests performed per confirmed case, or the percent of total tests that are positive, would be useful as well, he said.

Michael Urban, a senior lecturer at the University of New Haven’s School of Health Sciences, said that tracking the number of people who have been re-infected by the virus would provide vital information as well. That work is picking up, but still has a way to go.

Experts agreed that both the number of individuals tested, and the number of total tests, should be tracked.

“With any statistical analysis, the more data points we have, the more confident and sure we can be that the result is accurate and precise,” said Lam of Chapman University. “If we’re just looking at one angle of the prism, we’re not seeing the whole pyramid.”

Source: Orange County Register

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *