As Pennsylvania clarifies coronavirus reporting, health experts call process into questionThe Tribune-Review, Greensburg — Megan Guza And Dillon Carr The Tribune-Review, Greensburg
May 23-- May 23--The Pennsylvania Department of Health on Friday took the first step to clarifying its reporting process for coronavirus test results, which infectious disease experts say has muddied the picture of the outbreak in the state.
Pennsylvania was among several states that combined two types of test results, according to a report in The Atlantic magazine. The Centers for Disease Control and Prevention, the federal agency leading the fight against the virus, is doing the same thing, The Atlantic reported.
Diagnostic tests show whether a person currently has the virus. Antibody tests gauge whether someone has the antibodies associated with the virus, meaning they have had it or have been exposed to it.
Pennsylvania Department of Health officials on Friday began reporting a separate antibody result number as part of the agency's daily update, but the department continued to include those test results in its overall case count.
The state tracks how many of the positives were the result of antibody tests but does not break down the negatives, calling into question the number of people who have been tested for the virus as opposed to being tested for antibodies.
Secretary of Health Rachel Levine has repeatedly said the department follows CDC guidelines in its reporting.
A Pittsburgh-based infectious disease and critical care physician, Dr. Amesh Adalja, said intermingling viral and antibody test results clouds the picture of what's happening with the outbreak: The two tests, while both valuable, measure different metrics.
Diagnostic tests paint a picture of the here and now. Antibody tests, he said, look backward.
"If somebody took an antibody test today, and it showed they had it in March and they're recovered, how does that tell us what we should do today?" Adalja said. "You have to make sure those numbers (that the state reports daily) are reflective of acute infections -- where you are today and where you're going to be tomorrow."
Nate Wardle, a health department spokesman, said 97.5% of the state's cases are confirmed through a diagnostic test, and another 1.8% are probable due to symptoms and exposure. Positive antibody tests make up about 0.7% of the state's positive cases, he said.
As of Friday, the Department of Health was reporting more than 66,000 covid-19 cases since the first were identified March 6. Positive antibody test results accounted for less than 500 of those cases.
Wardle said negative antibody results are counted toward the overall number of negative tests across the state, which stood at close to 313,000 on Friday. He said the department does not have a breakdown of how many negative results are from antibody tests.
Dr. Thomas Walsh, an infectious disease specialist with Allegheny Health Network, said combining the test results doesn't make sense, "especially if you lose your ability to separate them out."
The Department of Health has long noted that the overall case count it reports every day includes "confirmed" and "probable cases."
An individual is confirmed to have covid-19 only after a diagnostic test -- generally a nasal swab -- indicates as much. An individual is deemed to be a probable case if he or she has symptoms of the virus and has been in close contact with a confirmed case.
On Thursday, however, Levine said individuals who have a positive antibody test also are being included in the probable case count.
"There's a lot of potential downside to doing these sort of things," Walsh said. "You look at those tests separately because they give you much different pieces of information."
Lumped together, their "usefulness is really diminished," he said.
Having an inflated number of test results by combining the two measures could lead to the perception that more people were tested for the virus, leading to a positivity rate that looks lower than it actually is, Adalja said.
It's hard to tell what the data is saying when results are combined, he said.
"It's much harder to be confident in the data you're using if the data doesn't have high fidelity," Adalja said. "(The tests) shouldn't be lumped together."
All of this can erode public trust, Adalja said.
"When the general public sees these errors in data, they become less trustful of public health authorities," he said. "We have to be careful with data reporting and mindful of conspiracy theories."
(c)2020 The Tribune-Review (Greensburg, Pa.)
Visit The Tribune-Review (Greensburg, Pa.) at www.triblive.com
Distributed by Tribune Content Agency, LLC.