Good serological tests are needed to determine the state of immunity to covid-19 in the population. However, as is known, not all of these tests are equally reliable. The outcome of a study by Mayara Lisboa Bastos et al., Published in The BMJ, further underlines this. In particular, the lateral flow tests, which are widely used internationally as point of care tests, have to suffer.
Lisboa Bastos et al. Collected studies on serological tests published between January 1 and April 30 this year. 70 percent came from China, the rest from the UK, the US, Spain, Japan, Germany and Sweden. About half of these studies were published in the form of preprints, ie without peer review. In addition, much research was found to be negatively influenced by one or more forms of bias. Correcting for these shortcomings and ignoring a lot of research of too low quality, Lisboa Bastos et al. Arrive at a pooled sensitivity of all the tests examined. This varied between 66 and almost 98 percent. Specificity was better, moving between 96.6 and almost 100 percent.
In particular ELISA (the Enzyme-linked ImmunoSorbent Assay, also called EIA: Enzyme Immuno Assay), a test in which an antibody reacts to an antigen bound to a plastic surface, and CLIA (het chemiluminescent immunoassay) did a lot better than the lateral flow-tests.
The latter are unsuitable for epidemiological or medical decision-making, the researchers said. Therefore, governments considering introducing an immunity passport cannot rely on the results of such tests in providing it. The researchers calculate by way of illustration that in a population with a covid-19 prevalence of 10 percent, out of every thousand people tested with it, 31 will be wrongly told they are immune when they have never had covid-19 (the may well have been through another viral infection), and 34 individuals will be falsely reported to have never been infected with SARS-CoV-2. So don’t use those lateral flow tests, according to Lisboa Bartos et al
Doi: 10.1136 bmj.m2516