In this blog for General Practitioners, Robert Walton, Senior Fellow in General Practice at Cochrane UK, examines evidence from a new Cochrane Review on antibody testing for COVID-19 and looks at how useful these tests might be for GPs in the UK.
Editor’s note: The Cochrane Review was updated on 17 November 2022 and the blog will be revised to reflect the new evidence.
Facing a new wave of the coronavirus, GPs may be better prepared this time for the onslaught. Among our new tools are a range of antibody tests for COVID-19 infection offering the prospect of quick diagnosis from routine blood samples. NHS labs have made these tests available to primary care remarkably quickly but it is not completely clear how they are best used and what the benefits of antibody testing might be when we already have a pretty reliable route to diagnosis using polymerase chain reaction (PCR) on nasopharangeal swabs.
So what’s the evidence on antibody testing?
The Cochrane Review Antibody tests for identification of current and past infection with SARS‐CoV‐2 examined data from 54 study cohorts with 15,976 samples and 8,526 of these were from people with COVID-19 infection.
As might be expected, the sensitivity of antibody testing was poor in the first week of the illness – only detecting 30% of people with COVID-19. This means that there would be a high number of false negative tests if used in the early stages of the disease. However the sensitivity of the test rose surprisingly quickly to 72% for the combination of Immunoglobulin G (IgG) and Immunoglobulin M (IgM) antibodies in the second week and 90% in week three. Unfortunately there were insufficient data at later time points so we don’t know how rapidly the antibody levels fall off.
It is surprising that the ‘gold standard’ in most of the studies included in the review was PCR alone rather than a composite of PCR and clinical data. So there may have been patients included who were PCR negative but who clearly had COVID-19 infection on clinical grounds. Thus the number of false positives for antibody tests in the studies may have been overestimated.
And how useful are these tests in practice?
It seems clear that clinical assessment with a PCR test, if available, will remain the mainstay for diagnosis of COVID-19 infection in the first two weeks of the illness, given the low sensitivity of antibody testing. It is during this period that, at least in my experience, most people tend to present to GPs. So blood tests of any kind at this early stage seem to be irrelevant – indeed an unnecessary infection hazard for staff and other patients in the practice.
Since most people start to recover after the first week of infection the need for testing diminishes over time. But for those who remain unwell there may be uncertainty about the diagnosis. PCR testing at this stage may be less reliable so antibody testing could be of some use for those who did not have a PCR test initially or who were PCR negative. However many patients who are deteriorating clinically at this stage will need specialist assessment and antibody testing in primary care may simply introduce unnecessary delay.
People who remain unwell and who do not need specialist treatment will be managed in primary care and where there is diagnostic uncertainty then COVID-19 IgG and IgM antibody tests may be a useful addition to a routine panel of blood tests.
For people presenting later with ‘long COVID’ then antibody testing may still be of value in establishing whether the person has been infected. From the data in the review we don’t know how quickly levels fall after the first month. However it is likely that the IgM antibodies would fall more quickly whereas IgG antibodies would persist for longer. So for people presenting with persistent symptoms some weeks or months after the illness a positive antibody test may be helpful but a negative one would not rule out the diagnosis.
Similarly, since the studies in the Cochrane Review were mainly conducted in hospital patients who may have had more severe disease, we do not know how well the tests perform for people who have a milder illness and potentially lower antibody levels. In addition the authors had concerns about high risk of bias which means that the accuracy of tests in clinical care will be lower than reported in the review. Thus clinical assessment will continue to play an important role in diagnosis and management.
Take home points
Read more blogs about COVID-19 on Evidently Cochrane.
Join in the conversation on Twitter with @CochraneUK and @rtwalton123 or leave a comment on the blog. Please note, we cannot give medical advice and we will not publish comments that link to commercial sites or appear to endorse commercial products. We welcome diverse views and encourage discussion. However we ask that comments are respectful and reserve the right to not publish comments we consider offensive.
Deeks JJ, Dinnes J, Takwoingi Y, Davenport C, Spijker R, Taylor-Phillips S, Adriano A, Beese S, Dretzke J, Ferrante di Ruffano L, Harris IM, Price MJ, Dittrich S, Emperador D, Hooft L, Leeflang MMG, Van den Bruel A. 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.
Declaration of interests:
Robert Walton reports grants from NIHR Health Technology Asessment, grants from NIHR Programme Grants for Applied Research, personal fees and other from TTS Pharma, outside the submitted work; In addition, Dr. Walton has a patent WALTON R, MCKINNEY E, MARSHALL S, MURPHY M, WELSH K, others. GENETIC INDICATORS OF TOBACCO CONSUMPTION. Patent number: 2001038567. Filed date: 24 Nov 2000. Publication date: 01 Jun 2001 with royalties paid to gNostics, and a patent Tucker MR, Walton R, Matthews H, Miskin A. Method and Kit for Assessing a Patient’s Genetic Information, Lifestyle and Environment Conditions, and Providing a Tailored Therapeutic Regime. Patent number: US20110251243 A1. Application number: US 12/944,372. Filed date: 11 Nov 2010. Publication date: 13 Oct 2011 issued to None.