Florida’s hidden data skews COVID-19 test results

Florida might be minimizing the depth of its COVID-19 problem by underreporting its rate of positive tests, experts say.

The method used to calculate the “positivity rate” — a critical measure of the pandemic’s progression — puts more emphasis on negative tests, skewing the results in that direction.

A person who tests positive is counted only once, but negative tests can be counted repeatedly if the same person got more than one test.

In addition, just as the pandemic raced out of control this month, the state changed the formula. It now mixes two different types of tests, including one that produces more false negative results.

The upshot of both factors is that the rate of positive tests, as quoted by the state, could make the situation look significantly better than it is, experts say.

How much the rate would change, if calculated differently, is impossible to know because the state refuses to release key pieces of raw data.

“There is a peculiar odor around the data in Florida and there has been for some time,” said Dr. William Hanage, associate professor of epidemiology at Harvard’s T.H. Chan School of Public Health.

Olivier Lacan, a volunteer for the COVID Tracking Project, which has followed data across the country, said Florida’s lack of transparency is fueling distrust.

“The right wing is saying the numbers can’t be trusted; the left is saying the numbers might be higher. Just show the math!” he said.

Governor Ron DeSantis frequently cites the positivity rate to justify reopening schools and businesses. He tweeted the statistic 12 times between late April and the end of May. He last tweeted the figure on May 26, but he has highlighted it during several news conferences.

Recently, his spokeswoman, Helen Aguirre-Ferré, has tweeted the statistic twice, pointing to a four-day-long decline in the measure as a silver lining in Florida’s looming pandemic thundercloud. That trend reversed on Monday.

How it’s calculated

Here’s how state officials say the positivity rate is calculated:

“We only count the positive test once,” Alina Alonso, the head of the Palm Beach County Department of Health, told Palm Beach County commissioners on July 7.

“We do count the negative tests more than once because there are reasons for people testing negative and getting multiple test results. But the positives by name are only captured once,” she said.

That means that the same person with multiple negative tests can be counted several times.

That’s fine for the day-to-day positivity rate, said Jason Salemi, associate professor of epidemiology at the University of South Florida’s College of Public Health, but a problem arises when you try to calculate an average over a series of days from the daily statistic.

“If the same people are testing negative and getting reported over and over … well, the results could be quite different from a true person-level analysis,” he said.

The number of people being retested on a daily basis is potentially very large, Lacan said.

On Tuesday, he estimates that at least 21,000 of the day’s approximately 53,700 negative tests were likely retests, since 45,753 people were tested for the first time that day, but about 67,000 tests were reported.

Some 12,290 people tested positive for the virus, according to results released that day.

Aguirre-Ferré acknowledged May 22 that the government was counting only people who tested positive once, but made no mention of negative tests in her statement.

Late Wednesday evening, Alberto Moscoso, the spokesman for the Florida Department of Health, said the state counts multiple negative tests for the same person because “a negative test only reflects the individual’s status at the moment the specimen was collected. Since a person who tested negative previously remains susceptible to the virus, the subsequent test is still included in the denominator of the positivity rate calculation.”

Combining tests

Experts say the state’s decision July 1 to include Antigen tests in their count made the situation even more murky.

Before then, the state had published results for PCR tests only. PCR tests are the “gold standard” of virus tests, since they detect pieces of the virus genetic material with a high degree of accuracy. But they can be slow to process.

Antigen tests, which instead look for specific proteins on the exterior of the virus, are much faster to process. However, they are significantly less accurate. One popular test, produced by Quidel, a pharmaceutical company, takes as little as 15 minutes but can miss up to 20% of positive infections.

Combining the two types of tests into one statistic could drive down the positivity rate, since a significant number of Antigen tests may be false negatives.

“Do not combine those cases, just don’t,” said Hanage, the associate professor at Harvard. “That’s lunatic. If that’s the case in Florida, people should be up in arms.”

It is unclear how many antigen tests the state has performed since July 1. The state does not disclose this figure in its daily reports and has not responded to questions about the number of tests performed.

Dr. Amira Roess, professor of Global Health and Epidemiology at George Mason University, said that no matter how the data gets parsed, the conclusion right now is pretty clear.

“The bottom line is that their percent positivity is high, and it keeps going high,” she said. “Florida’s hospitalization is getting very close to hitting capacity in some cities. It’s very clear that they’ve got a COVID hot spot.”

Staff writer Karina Elwood contributed to this report. Staff reporter Adelaide Chen contributed to this report.