No Good Evidence on Accuracy of Coronavirus Antibody Tests: Study
WEDNESDAY, July 1, 2020 (HealthDay News) -- Do you wonder if you've been exposed to the new coronavirus in the recent past?
Good luck finding out for sure: A new review finds there's little good evidence of the accuracy of blood antibody tests for COVID-19, especially those performed outside a lab.
The new findings "indicate important weaknesses in the evidence on COVID-19 serological [blood] tests, particularly those being marketed as point-of-care tests," said researcher Mayara Lisboa Bastos, of McGill University Health Center in Montreal, Canada, and colleagues.
And, the researchers believe that point-of-care tests -- performed directly with a patient outside of a laboratory -- are especially inaccurate at telling people whether or not they've been infected with coronavirus. Bastos and colleagues said that such tests should no longer be used.
In order to determine the diagnostic accuracy of COVID-19 antibody screening, the investigators reviewed 40 studies that assessed the sensitivity and/or specificity of a variety of such tests.
Test sensitivity is the ability of a test to correctly identify those with the disease (true positive rate), whereas test specificity is the ability of the test to correctly identify those without the disease (true negative rate).
Of the 40 studies, 70% were from China and the others were from the United States, the United Kingdom, Denmark, Spain, Sweden, Japan and Germany. Half of the studies were not peer-reviewed and most had problems in study design that might have influenced results, the researchers said.
Only four studies included outpatients and only two evaluated tests given at the point of care.
When results for sensitivity (how often a positive test result got it right) in each study were pooled together, they ranged from 66% to 97.8%, depending on the type of test method used. That means that anywhere from 2.2% to 34% of patients with a current or prior infection would be missed.
Results for specificity (how often a negative test result was wrong) ranged from 96.6% to 99.7%, depending on the test method used. That means that anywhere from 3.4% to 0.3% of patients would be wrongly identified as having been infected with coronavirus, when in fact they were negative.
Sensitivity results were consistently lower for the lateral flow immunoassay (LFIA) test compared with other test methods. The LFIA test is the potential point-of-care method that is being considered for so-called "immunity passports" that would allow people to return to work, for example
If an LFIA test is used in a population with a COVID-19 rate of 10%, for every 1,000 people tested, 31 who never had been infected with the coronavirus will be incorrectly told that they are immune, and 34 people who had been infected will be incorrectly told that they were never infected, the researchers said.
Sensitivity results were lower with commercial test kits (65%) than with non-commercial kits (88.2%), the study found. Sensitivity also tended to be lower for tests conducted in the first and second week after the start of symptoms, compared with after the second week.
The review was published online July 1 in the BMJ.
One big problem is the paucity of high-caliber studies into the accuracy of COVID-19 blood antibody tests, the study authors stressed in a journal news release.
"This review underscores the need for high-quality clinical studies to evaluate these tools," the team concluded. "With international collaboration, such studies could be rapidly conducted."
Dr. Amesh Adalja is an infectious disease physician and senior scholar at the Johns Hopkins Center for Health Security, in Baltimore. Reading over the new findings, he agreed that "while there is much interest in serological tests, it is important that we realize that there are many test kits out there that do not have adequate operating characteristics for reliability."
The new review "underscores the need to standardize how these tests are interpreted, what are the acceptable sensitivity and specificity parameters, and how test results can be operationalized," Adalja said.
The U.S. Centers for Disease Control and Prevention has more on COVID-19 antibody testing.
SOURCES: Amesh Adalja,MD, senior scholar, Johns Hopkins Center for Health Security, Baltimore, Md.; BMJ, news release, July 1, 2020