Addressing the Problem of False-Positive PCR Results in COVID-19 Tests

There are some remaining contamination issues and human errors that produce false positives.

In previous articles, I discussed the frequency of false positives in COVID-19 PCR tests (False Positives in PCR Tests for COVID-19) and the resulting impacts (Impacts of False Positive Results in COVID-19 Tests).

Ever since PCR tests were developed to aid medical diagnoses in the late 1980s, it has been recognized that they produce false positives. False-positive rates were initially quite high, sometimes in the double digits, and efforts were made to reduce them. These efforts focused on designing tests to ensure that they don’t mistake some other organism’s genetic material for the DNA or RNA of the pathogen targeted by the test, a problem called cross-reactivity.

Thanks to improvements in test design, this is now quite rare. Other efforts developed procedures to prevent a PCR test from being contaminated by genetic material from a previous run, called carryover contamination. But despite these efforts, there are some remaining contamination issues and human errors that produce false positives.

Two approaches have been used in the past to address the problem. First, health authorities restricted testing to individuals with signs, symptoms, or exposure, who were more likely to have an infection. As I explained in the first article, if the infection rate among the individuals tested is higher, the likelihood that a positive result will be false is lower. Second, diagnoses were never made solely on the basis of a single positive PCR result, but required corroboration by signs or symptoms, or confirmation by additional PCR tests or other types of tests.

In every prior epidemic that we examined – SARS, MERS, H1N1, Ebola, and Zika – guidance provided by the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC) required some combination of these measures. With SARS, for example, when the infection rate was thought to be extremely low or possibly zero, the CDC required six positive PCR results in order to diagnosis an infection: the initial test, a second test of the initial sample by the same lab, a third test by a different lab, and then the same series of tests on a second sample taken from the patient. In other epidemics, the expected infection rate was higher and the guidance was less extreme, but still, some combination of restricted testing and corroborating evidence was required – that is, until 2020 and the current pandemic, when for some reason all such measures were dropped. That’s what led to our current problems with false positives.

What can we do about it? Well, we don’t need to reinvent the wheel. First, even if we do nothing else, simply being aware that positive PCR results can be false (and when the infection rate is low, are likely to be false) would help a lot. For example, hospital and nursing home administrators would think twice, as they should, about moving a patient or resident into a COVID-19 unit based a single positive PCR result. But in addition, at least in situations with low infection rates, or where asymptomatic individuals are tested, we should check positive results with additional tests. There are many situations in which this applies, even during the current surge: mass testing of residents and staff at long-term care facilities, testing of athletes in sports programs, individuals seeking testing prior to travel, screening of hospital and pre-surgical patients, testing of healthcare and other frontline workers, and testing of students or staff at colleges, schools, and workplaces.

Finally, if we can figure out more precisely what is causing false positives, we may be able to develop procedures or tools to reduce them at the source, as we were able to do for cross-reactivity and carryover contamination. The enormous number of tests now being run for COVID-19 provides an unprecedented opportunity to improve our understanding of what causes false positives. This would not only help to improve the accuracy of COVID-19 PCR tests, but could also improve the accuracy of PCR tests for other pathogens, and possibly of other diagnostic tests as well. But because we’re not addressing the problem, we’re squandering the opportunity.

For further information on false positives in COVID-19 PCR tests, see https://www.medrxiv.org/content/10.1101/2020.04.26.20080911v4 and https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3656876.

Programming Note: Listen to Andrew Cohen report this story live today during Talk Ten Tuesdays, 10-10:30 a.m. EST.