A Johns Hopkins research team has determined that tests are more likely to result in false negatives for the COVID-19 virus if taken too early in the course of infection.
A Johns Hopkins team of investigators has found that testing infected individuals too early for the novel coronavirus disease 2019 (COVID-19) may result in a false negative, even though they may test positive for the virus later.1
The study, published in May in the Annals of Internal Medicine journal, examined 7 previously published studies that provided data on the performance of a common method called reverse transcriptase polymerase chain reaction (RT-PCR) in diagnosing COVID-19.1
RT-PCR function tests function by copying and detecting a virus’s genetic material when a group of infected cells are swabbed. Though capable of rapidly generating results, these tests have a history of producing negative results, such as when cells are missed, or if virus levels are too low during the earliest stages of an infection.2
The 7 selected studies covered a total of 1330 upper respiratory swab samples collected from both inpatients and outpatients using the RT-PCR method. From them, the researchers calculated the probability that someone who later tested positive for COVID-19 tested negative initially. False negatives were estimated at a daily rate since exposure and the first appearance of symptoms.1
Over the course of 4 days, the team estimated that the probability of getting a false-negative decreased from 100% on day 1 to 67% on day 4. The day that symptoms started appearing, day 5, the median rate for false-negative testing was 38% and decreased to just 20% on day 8. The rate increased to 21% on day 9 and to 66% on day 21.1
Lauren Kucirka, MD, PhD, MSc, obstetrics and gynecology resident at Johns Hopkins Medicine, explained that this demonstrates that the later a person suspected to have contracted COVID-19 is tested, the less likely they are to result in a false-positive. “A negative test, whether or not a person has symptoms, doesn’t guarantee that they aren’t infected by the virus,” she said, arguing that individuals who have high-risk exposure should still be treated as if they are infected, especially if they exhibit symptoms consistent with the virus. The RT-PCR tests and their shortcomings should also be conveyed to these individuals.2
The study investigators also noted that its estimates in probability were imprecise due to the diverse pool of patients on which the results of the 7 chosen studies were based.1
The investigators concluded that care must be taken when interpreting any test for COVID-19, especially when using results as reason for removing precautions that are intended to prevent further transmission. New and improved tests are needed to diagnose patients accurately and quickly, which will allow for better control of the virus’s spread, the investigators said.2
References