Editor’s note: The Salt Lake Tribune is providing readers free access to critical local stories about the coronavirus during this time of heightened concern. See more coverage here. To receive top news in your inbox every weekday morning, subscribe to our Top Stories newsletter. To support journalism like this, please consider donating or become a subscriber.
The accuracy of coronavirus tests by TestUtah.com has come into question, with state data showing that the rate of positive results among people tested at its sites is less than half what it is for patients tested elsewhere in the state.
And medical experts have been raising concerns about the effort, which has funneled millions of taxpayer dollars to the Utah tech companies running it here and in two other states. While Utah’s larger health care systems have relied on national names in diagnostics for their coronavirus tests, TestUtah has obtained its tests from a smaller Salt Lake City company and is processing them at a regional hospital in Orem.
“I worry about having tests routed to a small community hospital lab inexperienced with highly complex molecular testing that uses a test from an unknown company without much in vitro diagnostic experience,” Bert Lopansri, a specialist in infectious diseases and microbiology at Intermountain Healthcare, wrote in an April 14 email that The Salt Lake Tribune obtained through an open records request after it circulated through state offices.
“A pandemic is not the time for amateurs to learn,” Lopansri concluded.
More than 108,000 Utahns have been tested statewide, mostly at Intermountain and University of Utah Health facilities and drive-up sites, as well as through other health care providers and TestUtah sites.
As of this week, TestUtah had eight testing sites around the state and had conducted tests on more than 18,000 Utahns, more than half of whom did not have symptoms of the coronavirus — despite state guidelines that only symptomatic patients should be tested.
That is likely to produce more negative results than the state’s other test sites, said executives of the companies involved with TestUtah. Mark Newman, whose health software company Nomi Health has a $5 million contract with the state to run the group’s testing, said comparing TestUtah’s results with those of other sites is “not apples to apples.”
“We’re talking about population differences,” agreed Brent Satterfield, founder and chief science officer of Co-Diagnostics Inc., which produces the test kit used by TestUtah’s sites.
But according to state data obtained by The Tribune, TestUtah has reported test results for symptomatic patients separately — and even for those patients, positive results are far below those reported at other test sites in the state.
About 2% of symptomatic patients at TestUtah’s sites have tested positive for coronavirus since April 1, according to the state’s data. That’s less than half of the 5% of patients testing positive at other Utah sites.
At a news conference Thursday, Gov. Gary Herbert said the state will work to ensure TestUtah is producing reliable results. Herbert said personnel at the Utah Department of Health and the Department of Public Safety will work with the companies behind the effort.
He noted that TestUtah sites have often been testing people with no symptoms, although the state epidemiologist has asked for only people with a symptom of COVID-19 to seek testing.
“That’s a concern to reconcile,” Herbert said, “so we make sure the results are accurate.”
State officials are watching TestUtah results
The reason for the discrepancy between TestUtah sites and others is unclear, but health officials and doctors have expressed concern over the testing practices at TestUtah locations and the sensitivity of the test being used.
“We have been watching TestUtah’s positivity rate closely and have had multiple discussions with them about different practices that may be impacting their testing results,” Nate Checketts, deputy director of the Utah Department of Health, wrote in an email obtained by The Tribune.
For example, Checketts wrote, TestUtah sites at one point were using oral swabs to obtain samples from patients — not the nasal swabs now recommended by the federal Centers for Disease Control and Prevention (CDC) to optimize virus collection.
Satterfield, of Co-Diagnostics, said he doubted the collection of the samples or laboratory protocols would explain the discrepancy in positive results. He has spoken with the clinicians in charge of swabbing patients, he said, and the collection is so thorough that swabs sometimes induce nosebleeds.
The Co-Diagnostics testing kits come with instructions that any laboratory can follow, Satterfield said. The company also consulted with the laboratory at Timpanogos Regional Hospital, which is processing the TestUtah kits, to answer any questions staffers there had.
“We consider them to be able to get the job done,” Satterfield said.
MountainStar Healthcare, which operates Timpanogos Regional and has partnered with TestUtah to staff its testing sites, defended its testing practices. HCA Healthcare, the largest hospital system in the nation, is MountainStar’s parent organization.
“We have absolute confidence in our testing process and capabilities," MountainStar spokesman Mike Graul wrote. “The state [coronavirus] task force and governor’s office [have] been in complete support of this initiative and continue to be so.”
Satterfield said that lab managers on Wednesday pointed out to him that people who are certain they are sick are more likely to seek services through a hospital. People who are only curious as to whether they have contracted COVID-19 are more likely to fill out TestUtah’s online form and get tested at its sites, he said.
“I probably wouldn’t want to show up at a hospital with people I know are sick if I was at all unsure that I could be,” Satterfield said.
In a prepared statement, Nomi Health said lower positive rates are a “natural” result of self-reporting by people whose symptoms are so mild that a doctor at another site may tell them they don’t qualify for testing.
“TestUtah is about community health, enabling as many Utahns as possible to get tested and have the peace of mind that comes along with that,” the statement read.
“This kind of broad community testing will naturally result in a lower positive rate, as even those who show minimal to no symptoms can screen themselves as ‘symptomatic,’ compared to clinically ordered and have the opportunity to be tested," it said. "All testing is conducted within CDC guidelines.”
A task force doctor urged scrutiny of TestUtah’s testing
TestUtah staffers reached out to health officials to help “validate their technique and processes,” but there was no reply, according to an April 17 email written by Kristen Cox, executive director of the Governor’s Office of Management and Budget.
Cox sent the email to Herbert, interim Utah Department of Health executive director Jeff Burton and several state employees. She accused health department employees of a pattern of “nonresponse.”
A health department spokesman did not respond Wednesday to The Tribune’s request for comment. But in an email replying to Cox’s complaints, Burton wrote: “I am on a calm and measured rampage to get everyone together on this.”
Cox also asserted that Lopansri sent his criticism to HCA Health, MountainStar’s parent company. Cox wrote that it was “inappropriate for [Intermountain] to ask HCA to stand down” and added: “I’m not sure if this is all about how to divide the pie.”
“HCA headquarters is apparently furious,” Cox wrote, saying she learned about Lopansri’s message when she “received a call of frustration from Silicon Slopes” about his email.
Promoted by nonprofit industry group Silicon Slopes, TestUtah.com launched in early April to plaudits from Herbert and other state officials as a philanthropic initiative. As news of Nomi Health’s profitable contracts with Utah, Iowa and Nebraska emerged last week, Silicon Slopes Executive Director Clint Betts responded that the project had become too large to operate on a volunteer basis.
Graul, the MountainStar spokesman, wrote Wednesday that Timpanogos Regional “did receive a letter from a competitor."
But Intermountain Healthcare on Wednesday supplied Lopansri’s email in its entirety — and the message appears to be addressed to Checketts, not HCA or any of its entities.
Lopansri sits on the state’s task force on coronavirus testing and wrote the email to Checketts, who is chairman of the committee, said Intermountain spokesman Jess Gomez.
Each organization on the task force “has made changes based on what we have learned together,” Gomez wrote in a prepared statement.
“The letter ... was not sent directly to any one health organization,” Gomez wrote. “It expressed concerns and questions that Dr. Lopansri was seeking clarification for, which is consistent with the open dialogue of the task force as they work together collaboratively to address these important health issues.”
But in the full email, Lopansri doesn’t mince words for TestUtah.
“What alarms me the most is that they are expanding collection and testing with these unknowns about how their test performs. If correct, I urge you to halt their testing until we understand why their results differ so much from what other labs are reporting," Lopansri wrote.
“This is a potential public health disaster that will be compounded by the fact that they are constantly promoting themselves publicly while we are not."
Critics question, and Co-Diagnostics defends, TestUtah tests
Meanwhile, some experts have raised questions as to the accuracy of the Co-Diagnostics test used at the TestUtah sites.
Shares in Co-Diagnostics tripled in late February, when it received certification to sell its coronavirus test in Europe. On April 3, the day the state signed its contract with Nomi, Co-Diagnostics received emergency use authorization from the Food and Drug Administration (FDA) to sell its test in the United States.
Now the company is supplying thousands of tests in Iowa and Nebraska, where state officials last week issued no-bid contracts worth more than $50 million to Nomi and the TestUtah team.
But the Co-Diagnostics test has a “higher limit of detection compared to tests offered by more established vendors,” Lopansri wrote, referring to the test’s sensitivity. A higher limit of detection means more virus is required in a sample to trigger a positive result.
An analysis by the life sciences publication BioCentury showed that at least 16 of 22 comparable tests authorized by the FDA report a lower limit of detection, or greater sensitivity, than Co-Diagnostics’ tests.
Its limit of detection also is higher than that of the CDC’s test, wrote Dan Diekema, director of the infectious diseases division at the University of Iowa’s department of internal medicine.
“Based upon concerns I’ve heard from others re: test performance, it would be irresponsible to move forward without evaluating their test using sample exchange [with] Iowa’s State Lab. Or do all tests at [the state health lab],” Diekema said on Twitter. Diekema declined to comment further.
But Satterfield maintained that in a real-world environment, people who have contracted COVID-19 have thousands or even hundreds of thousands of particles of the virus in their samples, and that’s plenty to trigger a positive in his company’s tests.
He added that Co-Diagnostics’ COVID-19 tests scored between 99.52% and 100% in evaluations conducted by the FDA and in Europe. He said those evaluations often put his kits side by side against other available kits to gain consensus results.
“We are in the same ballpark as the other tests,” Satterfield said.
And there have been no complaints from nearly 50 other countries, where Co-Diagnostics has sold the majority of its tests, Satterfield said.
TestUtah’s testing of asymptomatic patients raises concerns
TestUtah has since April 1 been reporting COVID-19 test results for patients with and without symptoms, according to the state’s data.
The decision to test asymptomatic people has been a point of contention. State epidemiologist Dr. Angela Dunn said that health officials have asked TestUtah staff to comply with state testing guidelines that require patients to have at least one symptom to receive a test.
“We’ve given them our advice and the state guidelines that have been decided upon,” Dunn said, “and we’ve asked them to implement them."
Instead, TestUtah has continued asymptomatic testing for about two weeks after Dunn first acknowledged that health officials had asked them to stop.
“We were early on asymptomatic testing,” Newman, of Nomi Health, wrote in response to a Tribune inquiry Wednesday. “Now we are all working together on a unified assessment and testing criteria with the DoH [Department of Health] as we’ve been really successful working together on testing in Utah.”
Newman referred all further questions to Cox, of the management and budget office; she did not reply to The Tribune on Wednesday.
On Wednesday afternoon, a Tribune reporter filled out the questionnaire at TestUtah.com and received a notice that read, “You qualified to be tested,” despite reporting no symptoms, risk factors, or known contact with infected people. State guidelines for testing still called for at least one symptom.
“There is a place for asymptomatic testing in this pandemic, but it’s necessary to have a very thoughtful approach in terms of recruiting who should be tested if they don’t have symptoms ... so we can accurately interpret those results and use good data to inform policy moving forward,” Dunn said in a recent news briefing.
Because the accuracy of the test in asymptomatic patients is not known, health officials are wary of the risk of false negatives — which could mislead public officials and individual patients — as well as false positives, which immediately trigger a thorough contact-tracing investigation by state and county employees.
And scientific protocols exist to design a truly randomized study of infection rates, to protect against the biases that may come from relying on a pool of volunteers on the internet. That’s something the University of Utah planned to launch this week with ARUP Laboratories, using a new antibody test, said Julio Delgado, ARUP’s chief medical officer.
Economic officials for Herbert had said in mid-April that unrestricted testing should continue.
“Opening up testing to anyone who wants it on demand is the right direction,” Cox wrote in the April 17 email, stating that health department director Joseph Miner had earlier sanctioned that plan; he later stepped away from daily oversight of the department due to health concerns. “It provides reassurance to people who need it while also providing important insights into prevalence and other outstanding questions.”
But that testing comes at a price to taxpayers. Although federal rules require all insurers to cover coronavirus testing, TestUtah is not submitting claims to patients’ insurers. Instead, the state is paying Nomi directly; testing accounts for $3 million of the $5 million contract for TestUtah.com.
Betts, of Silicon Slopes, speculated that insurance billing may deter some patients from seeking testing.
“As I understand it, the goal of community testing is to make it as available and accessible as possible," Betts said.
“Adding insurance billing to the process adds another level of complexity to accomplishing that goal," he said. "That may be why insurance billing [and] reimbursements hasn’t been part of TestUtah up to this point but, again, that’s not a decision Silicon Slopes has been involved in nor is it our organization’s decision to make.”
Editor’s note • Clint Betts serves on the The Salt Lake Tribune’s nonprofit board of directors.
— Tribune reporters Bethany Rodgers and Robert Gehrke contributed to this report.