The media has reported that blood antibody tests for COVID-19 may not give reliable results. There are now dozens of tests available, which may be used either for population surveillance or in the diagnosis of individuals.

A review in the British Medical Journal gathered the evidence from 40 studies, the majority from China, which had looked at the accuracy of blood antibody tests. The various tests did not wrongly indicate someone had had COVID-19 when they hadn’t, but they failed to detect up to a third who’d had confirmed infection. Point-of-care tests that give instant results and commercial kits tended to be less reliable than standard laboratory tests.

A large Cochrane review also looked at the topic, focusing on the time when the test was taken. Less than 30% of people would show antibodies within the first week of symptom onset, but almost all would test positive after 2-3 weeks.

However, the low quality of studies and variety of tests, make it difficult to apply this research directly to the varied clinical and surveillance circumstances they could be used in. The question remains whether they are accurate enough to be useful and in which settings.

Where did the story come from?

The Telegraph was among media outlets in the UK to report on the summary of research on antibody tests, published in June 2020 in the British Medical Journal. At around the same time the Cochrane Collaboration also summarised research on antibody tests, focusing on their accuracy according to the time since infection.

What is the basis for the claim?

The BMJ study looked at 40 studies, 28 of which were from China, and 32 used a ‘case control’ design, including people known to have had COVID-19 and comparison samples of those not known to be infected. Most studies used commercial test kits, and around half of studies used tests that gave rapid results at point-of-care, rather than sending for laboratory analysis.

The antibody tests were positive for between 66% and 98% of people who’d had COVID-19 depending on the test used. Point-of-care tests and commercial kits tended to be less accurate than non-commercial kits and standard laboratory analysis. Though the tests very rarely gave ‘false positive’ results, meaning they didn’t suggest someone had had the infection when they hadn’t.

The Cochrane review looked at 54 studies, most from Asia and comparing people known to have had the infection with those who hadn’t. They found that less than a third of people with COVID-19 would test positive for antibodies within the first week of infection. About three-quarters would test positive at 1-2 weeks, and 90% by 2-3 weeks.

However, both reviews highlighted the low quality of the evidence as a whole. Overall this makes it very hard to apply these results to specific tests. But they do highlight that test timing makes a difference when testing for antibodies, and question the value of, instant result, point-of-care tests. These reviews looked at studies published to the end April only, so the tests, and the evidence, may have developed since.

What do trusted sources say?

The UK Department of Health and Social Care has guidance on the use of antibody tests, which are being provided to all NHS and care staff, and also may be requested by doctors for testing individuals in hospital and the community. They say that a positive antibody result will show you have previously been infected, but won’t tell you whether you’re immune or not, or may be able to pass the infection to others. DHSC also note that, as of the end of May 2020, finger-prick tests for home use had not been validated.

Other organisations, including the US Centers for Disease Control and Prevention also advise on antibody tests, saying they should not be used to diagnose current COVID-19 infection, and that having antibodies does not necessarily indicate immunity.

Analysis by EIU Healthcare, supported by Reckitt Benckiser

Citation

  1. Mayara LB, et al. Diagnostic accuracy of serological tests for covid-19: systematic review and meta-analysis BMJ 2020; 370 :m2516 https://www.bmj.com/content/370/bmj.m2516
  2. Deeks JJ, et al. Antibody tests for identification of current and past infection with SARS‐CoV‐2. Cochrane Database of Systematic Reviews 2020, Issue 6. Art. No.: CD013652. DOI: 10.1002/14651858.CD013652. https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD013652/full

Reading list

  1. Department of Health and Social Care. Guidance. Coronovirus (COVID-19): antibody tests. Published 22 May 2020. https://www.gov.uk/government/publications/coronavirus-covid-19-antibody-tests/coronavirus-covid-19-antibody-tests
  2. Centers for Disease Control and Prevention. COVID-19 Frequently Asked Questions: What is antibody testing? https://www.cdc.gov/coronavirus/2019-ncov/faq.html#Testing (Accessed 7 July 2020)