select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to ONS release on estimating the prevalence of long COVID symptoms and COVID-19 complications

The Office for National Statistics (ONS) have released estimates of the prevalence and duration of long-COVID symptoms and COVID-19 complications.


Dr Amitava Banerjee, Associate Professor in Clinical Data Science and Honorary Consultant Cardiologist, Institute of Health Informatics, University College London, said:

“Until now, our focus has been on effects of the COVID-19 pandemic in acutely unwell patients in hospital, and those with underlying conditions, such as heart disease or diabetes. However, long COVID, affecting people whether or not they were hospitalised with COVID-19, is already having an impact in terms of quality of life and time off work for a substantial number of people in the UK and other countries. Knowing the prevalence of long COVID and its risk factors is essential for planning how to prevent and treat it.

“ONS data have already shown that about 10% of the UK population has been infected with coronavirus during the pandemic. Recent data from the REACT study suggest that the current rate of infection may be 1-2% in certain parts of the country. The ONS data released today suggest that 1 in 10 people infected may have symptoms lasting for more than 12 weeks, i.e. long COVID. Although post-viral symptoms are common, whether after flu or other coronaviruses, the scale of this pandemic and the potential number of infected people with long COVID is unprecedented, and poses yet another COVID-19-related public health challenge.

“The observed association between multi-organ impairment and being hospitalised with COVID-19 is a cause for concern and definitely warrants more investigation. Long COVID provides yet more solid evidence to support public health strategies that suppress the infection rate.”


Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“It’s good that ONS have begun to use data from their Infection Survey to give a national picture of what’s been called ‘long COVID’. Many other data sources can’t give good information on this, because it needs to involve following up people who have been infected for a long period of time, to see how long their symptoms last, how they might change over time, and to investigate which complications might arise. Looking at people who are still reporting symptoms 12 or more weeks after they were originally infected is in line with the work going on at NICE* in developing guidelines for managing the long-term effects of the virus. ONS should be in a good position to provide relevant information, because their infection survey does repeatedly return to people who were previously tested (positive or negative) and is in a position to collect the necessary follow-up data, and also because they have access to different sources of data that can be linked to provide a more complete picture.

“However, this first publication of data is limited, and, as ONS write in the Statement that accompanies the estimates, “the analysis is very much a work in progress”. Some of the details that I’d expect to have seen in reporting results like this, such as numbers of people still being followed up at various times after their initial infection, and numbers of patient-years at risk for the adverse events that are considered, are not published in this release. The plans for further work and further analysis, outlined in the Statement accompanying the numbers, look promising, but it’s much too early for me to judge how they will work out.

“The data comparing the rates of important health complications in patients with Covid-19 and matched control patients who do not have Covid-19 do, on the face of it, look worrying. But, as ONS point out, this is observational data and it’s impossible to know whether Covid-19 actually caused the higher rates in the Covid-19 patients. Some detail is not given on exactly how the patients were matched, and the matching may not have covered all the potential factors that affect the risk of these complications. A detailed statistical analysis of the comparison between the Covid-19 patients and controls has not been provided yet. And (as ONS point out) the estimates relate only to patients who ended up in hospital – further work aims to look at complications in patients who were not treated in hospital.”



Dr David Strain, Clinical senior lecturer and honorary consultant, Co-Chair BMA Medical Academic Staff Committee, University of Exeter Medical School, said:

“These preliminary data are very concerning, suggesting that 10% of people who have experienced COVID are left with residual symptoms after 3 months – more than twice the rate than we previously thought. It also highlights that the majority of people who contract COVID have symptoms for over a month. 

“These data also give an insight into the range of complications occurring, for example the risk of a heart attack, stroke or heart failure being nearly 12 times higher in COVID patients than in comparable patients hospitalised with other infections, and the rates of developing diabetes or kidney failure being ~9 and ~10 times higher respectively. These confirm what we have seen working on the COVID wards, that this virus is more than a simple respiratory infection, but is a multi-system disease. In some ways, however, people affected by these complications are less unfortunate, as there is a pre-existing clear knowledge base and there are effective treatment strategies for these conditions. Of greater concern, are those with the non-specific fatigue, myalgia (muscle pains) and fevers that we have no real understanding of the causes, the risk factors or how to treat it.

“It is important to appreciate that these data mirror our experience of evaluating people with long COVID within the NHS, that the risks are not necessarily the same as those for a “poor” outcome from the primary disease or ultimately dying. This is happening to younger people, more women, than men – basically the population that were suggested to be at lower vulnerability from the initial disease, and therefore have been taking roles with higher hazard of coming into contact with the virus. The long term consequences for these individuals, and for the population as a whole, could be potentially devastating, in terms of physical manifestations for the individuals but also the economic impact of these individuals being unable to work.

“As with all experimental data analyses there are limitations. In the first wave testing was not widely available, therefore the denominator may be artificially lower. Paradoxically, however, as there does not appear to be a correlation between severity of primary disease and extent of Long COVID, it is possible that there are many more individuals with the condition that have not, as yet been recognised. The percentages, however, are very similar to those that were found in the Doctors tracker survey conducted by the British Medical Association  towards the end of the 1st Wave.”



All our previous output on this subject can be seen at this weblink:



Declared interests

Prof Kevin McConway: “I am a Trustee of the SMC and a member of the Advisory Committee, but my quote above is in my capacity as a professional statistician.”

Dr David Strain: “I am on the NHS LongCovid Task force, I am the BMA Lead on Long COVID, a clinician that is part of the long-COVID service delivery team in the South West and am the Chief of investigator of a study looking into the genetics of Long COVID.”

None others received.

in this section

filter RoundUps by year

search by tag