September 2, 2020

expert reaction to study from Italy looking at viral clearance and whether retesting weeks after symptom onset is a good idea

A study, published in BMJ Open, looked at viral clearance and whether retesting weeks after symptom onset is a good idea.

Prof Peter Openshaw, Professor of Experimental Medicine at the National Heart & Lung Institute, Imperial College London, said:

“It is important to recognise the great sensitivity of PCR as a method of detecting residual viral genetic material.  This should not be described as ‘virus shedding’: first, it’s not intact live virus; second it’s not shedding, in that the sample is obtained by pushing a swab deep into the nose and moving it around.  This is not evidence of infectivity, or that anyone in the vicinity of a person with a positive PCR result is at risk of being infected.

“The authors say: “Until there is information on whether patients who are clinically recovered but still swab-positive are contagious, and given the low sensitivity of viral tests, patients must remain isolated until complete viral clearance is achieved as confirmed by RT-PCR for SARS-CoV-2 on two consecutive negative swabs”. This is not in agreement with clinical consensus, which is that patients are very unlikely to be infectious beyond day 10 of disease.

“It is possible to estimate viral load and to measure the immune response by testing for antibody.  There are excellent studies that have examined infectivity, for example: https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2020.25.32.2001483#abstract_content; https://doi.org/10.1038/s41586-020-2196-x; https://doi.org/10.1038/s41591-020-0869-5 https://www.medrxiv.org/content/10.1101/2020.06.08.20125310v1. If the viral load is low and antibody is present, transmission is very unlikely indeed.  Infectivity is maximal at the time of symptom onset (or possibly just before).  Infectivity diminishes over time and is very unlikely in the later stages of disease.

“Publications of this sort can distract the public from a focus on known high-risk transmission events.  There is no need to be alarmed.  This manuscript should not be used as a source of guidance to change policy.  There is no evidence that people who have recovered from COVID-19 are a danger to others.”

Dr Alexander Edwards, Associate Professor in Biomedical Technology, Reading School of Pharmacy, University of Reading, said:

“This is a detailed and valuable study that provides us with more information about how long virus can be detected after COVID-19 infection.  It complements other, smaller studies from different groups, as well as many other reports with different methodology.

“The main findings relate to the range of time it can take after symptoms start before a swab test no longer gives positive result.  The study also provides additional data about the real-world accuracy of swab testing, showing that even when COVID-19 patients have a negative test soon after symptoms clear, 1 out of 5 of these negatives will test positive again 2 days later, showing that false negatives still occur even with the standard swab testing procedures.

“In the UK and across the globe, we are relying on this type of RT-PCR test from swabs very extensively as they remain the most established way to detect virus in a patient who is infected.  But we should remember that most tests including this one have some limitations.  The RT-PCR tests can detect very tiny amounts of virus which is a strength when you want to avoid false positives, but this can clearly result in people testing positive even >1 month after they are ill, as the virus can still be detected.  We still cannot rely on the test to “prove” that someone is either “infectious” or “safe”.  Instead, tests must always be used as part of a wider strategy, with a clear understanding of any limitations in performance.

“An obvious question remains (which is discussed clearly within this publication): if many COVID-19 patients are still swab test positive for virus 36 days after symptoms first occur, how does this fit with self-isolation guidance?  For example some guidance suggests people can stop self-isolating “once you have had symptoms for 10 days and your temperature has returned to normal”?  If this is under 30 days might they still be infectious?

“The answer to this question lies in the difference between the highest amount of virus present – which is seen for around 8 days from when symptoms start (and possibly even a few days before symptoms first appear) – vs being able to detect virus by RT-PCR from a swab, which is what is being studied in this report.  Even though the virus can still be detected much later, we still believe the most infectious period that is within the first week or so of the infection.

“However, we also believe a significant number of people can be infected but remain asymptomatic – in which case guidance for a safe self-isolation period around symptoms become even harder.  We have even less information about virus levels, how long people test positive, and how infectious people are, for asymptomatic persons as it’s even harder to study.

“What is vital therefore, is detailed use of case tracking to understand better how most people get infected.  Do they get infected mainly from people who are (or were very recently) symptomatic, or do some people get infected from an individual 30 days after they had symptoms?  If the majority of infections arise from people within ~14 days of symptom onset, it’s fine to use the current recommended self-isolation period.  The current guidance remains rational and evidence-based, but further information will still allow us to refine it further and both reduce levels of disruption, whilst continuing to slow spread.”

‘Temporal profile and determinants of viral shedding and of viral clearance confirmation on nasopharyngeal swabs from SARS-CoV-2-positive subjects: a population-based prospective cohort study in Reggio Emilia, Italy’ by Pamela Mancuso et al. was published in the BMJ Open at 23:30 UK time on Wednesday 2 September 2020. 

Declared interests 

None received.