Few Covid-19 rapid diagnostic tests meet WHO standards for accuracy

A health worker takes a blood sample of a passenger at a Covid-19 rapid test facility at Pasar Senen train station in Jakarta, Indonesia. Credit: EPA / Bagus Indahono

Rapid diagnostic tests (RDTs) could play a critical role in decentralising testing and reaching remote communities as WHO launches an Africa roll-out, but a recent study has found that only two of the six RDTs making it to the final round of testing met WHO accuracy standards.

RDTs test for proteins produced by the SARS-CoV-2 virus, collected using nasopharyngeal or oropharyngeal swabs. Typically, they are more reliable in symptomatic patients, with a high viral load, especially in their upper respiratory tract, and their ease-of-use makes them preferable to other means of testing in critical situations.

What are the findings? The study, led by the Foundation for Innovative New Diagnostics (FIND), evaluated the performance and ease-of-use of six commercially available RDTs in people presumed to have Covid-19, in testing centres across Brazil, Germany and the UK. The 2,400 adult participants enrolled in the study were identified as at risk for infection according to their local department of public health. Patients were excluded if they had previously been diagnosed with SARS-CoV-2.

The results, which were released at The Union World Conference on Lung Health and also pre-published on medRxiv.org, were then compared to the accuracy requirements approved by WHO. According to this profile, RDTs must have a minimum sensitivity of 70 per cent and at least 97 per cent specificity, but only two tests met the proposed minimal requirement for both categories.

The Standard Q, an RDT made by SD Biosensor Inc, performed best overall, with 76 per cent sensitivity and 99 per cent specificity in Germany, and with 88 per cent and 97 per cent in Brazil. It also scored 86 out of 100 for ease-of-use.

“These tests are likely to be useful in settings where the Covid-19 positivity rates are higher to minimise the number of false positive tests that are detected,” said Jilian Sacks, an author on the study. Because data suggests RDTs have improved accuracy in people with higher viral loads, this would be a good way, she added, “to be sure you're diagnosing someone with Covid-19, and then take appropriate health measures.”

Tests for antibodies post-infection seem more sensitive. On a more positive note, Sacks said, the blood tests being rolled out now to test antibody levels in the general population seem more reliable. The FIND study evaluating those tests, found that of the 15 enzyme-linked immunosorbent assays (ELISA), 11 met performance targets. ELISAs detect antibodies 14 days or more after symptoms appear, with a performance time of up to two hours.

This makes such tests an excellent way to track infection and immunity in large groups of people, but the sophisticated equipment required makes them inefficient for use in the rapid diagnosis of people who are actually ill and infectious.

Why is this important? Effective control of the virus is largely dependent on the quick identification of infected individuals, followed by a period of isolation and clinical support, if needed. The current gold standard of testing - polymerase chain reaction (PCR) tests - can only be conducted in a laboratory as they detect nucleic acid from SARS-CoV-2. This can leave people waiting anywhere from 48 hours to more than 10 days for results, impeding rapid identification and isolation efforts.

It is believed that RDTs, with their turnaround time of up to half an hour, play a critical role in cutting this delay and limiting the transmission of Covid-19. The cheap cost and ease-of-use mean they are being considered as an effective method of testing alongside current PCR tests.

Read the full version of this article on Health Policy Watch