The state of Utah is changing how it reports to the public one of the statistics that track the spread of COVID-19.
The test positivity rate is the percentage of tests that come back positive for COVID-19 — and Utah health officials have for months been explaining that a rate of 3% to 5% indicates the virus is under control.
It has gone as high as 32.71% on Jan. 7, during the state’s post-holidays surge, but it recently has been dropping, hovering around 16%.
The new method the Utah Department of Health is using makes the percentage lower. It will help public health officials “see the benefit of our increased testing capacity,” and it will make comparisons to other states more valid, Dr. Angela Dunn, the state’s epidemiologist, told reporters recently in explaining the change.
In its daily report Monday, the department listed both the previous measurement — a rolling seven-day average of 15.4%, and the new one, which has a rolling seven-day average of 7.3%.
Here’s how those methods differ.
What does the test positivity rate show?
The percentage of tests for COVID-19 that come back positive is “a leading indicator of the spread of COVID,” Dunn said. “In addition, it also measures how much testing we’re doing.”
How was Utah calculating it?
Since the beginning of the pandemic, Utah had been using a method called “people over people,” Dunn said. That means the number of people who tested positive, divided by the number of people who have been tested.
As people were tested, the state counted their first positive test within a 90-day period — or, if they never tested positive, their first negative test. Any other test results in that window were not included. Health officials use a 90-day window, Dunn said, because that’s the period of immunity for someone after they catch the coronavirus.
So under this method, a person who gets tested multiple times in three months is counted once — the subsequent results are “de-duplicated” in the health department’s tally of how many people have been tested within that 90-day span.
What’s changing in how Utah calculates it?
The new method is called “test over test.” It takes the number of positive tests in a day, or a week, and divides it by the number of tests conducted in that same time period.
“Every single test reported to the [Utah] Department of Health is included in this calculation,” Dunn said.
The number of tests being conducted in Utah is dramatically higher now, as testing has expanded with the addition of rapid tests and regular testing is widespread among employees, students, nursing home residents and others.
Those tests are generally of healthy people; they don’t have symptoms and don’t believe they have the virus. They’re being tested to stay in class or keep working.
So counting all results, when so many presumed healthy people are being tested, will move the positivity rate down.
What’s the benefit?
The new method, Dunn said, “is going to allow us to see the benefit of our increased testing capacity.”
With all of that testing, people who catch the virus will know quickly, she said. Health officials can respond with contact tracing and move faster to stop the spread. And the positivity rate will reflect that environment.
The “people over people” method made sense at the beginning of the pandemic, she said. Back then, the federal Centers for Disease Control and Prevention required a COVID-19 patient to have two negative tests in a row before they could be considered recovered from the virus.
So a person could rack up a lot of positive tests before getting two negative tests. Using the “test over test” method at that time would have artificially inflated the positivity rate, she said.
The CDC later changed that guideline, but Utah continued to use the “people over people” method. The CDC uses both methods and a third one, Dunn noted.
Public health officials watch both calculations, Dunn said. The “people over people” method, she said, “mirrors our surging cases.”
The “test over test” formula “is really useful because it’s allowing us to see our increase in testing, and how well that’s helping us decrease our cases.”
What’s the difference in practical terms?
The new positivity rate statistic will be smaller.
Monday’s report showed that difference, with 15.4% versus 7.3% for a rolling seven-day average under the two methods.
The “people over people” method, Dunn said, “biases percent positivity a little bit higher.” The “test over test” method “biases the percent positivity a bit lower,” she said. “The truth is somewhere in the middle.”
“You’re going to notice that the trends for both of these methods are actually really similar,” Dunn said. Looking at charts for both calculations, over time, they’re practically parallel.
Is this a political move, to make the numbers look better?
Dunn downplayed that perception. She noted that at least 37 other states use “test over test” — and switching over makes it easier to do apples-to-apples comparisons with other states.
UDOH has begun using the new method on its daily report, but it will continue to publish both calculations on its dashboard, at coronavirus.utah.gov, for the sake of transparency, Dunn said.
If the daily figure looks lower, will people let down their guard?
“We have to change the scale in our head,” Dunn said.
The post-Christmas test positivity rate of over 30% “was high because that was the scale we were using,” Dunn said. “Now, 15% is going to be really high, 10% will be really high.”
Dunn said the goal she and other health officials have described for months — that the test positivity rate needs to move below 3% to 5% before experts consider the virus under control — still holds true, even with the change in measuring the rate.
That target, as set by the CDC and the World Health Organization, is “relying on that ‘test over test’ method,” she said. “And so it’s still accurate to say, you know, the lower the percent positivity, the better. … We really rely on multiple metrics, and the trends are what we really care about.”
How should people read the data?
“We know that not one metric will tell us how bad it is or when it’s going to end,” Dunn said. “We have to look at everything.”
She recommends people watch three key statistics: The test positivity rate, the number of new cases, and the hospitalization numbers, or remaining capacity.
“All three of those metrics together give us a really nice picture of where we currently are in the pandemic and where we’re going,” Dunn said.