Study population and setting
This report posted to the UK Office for National Statistics website detailed the long term symptom duration following a SARS-CoV-2 infection among 9,063 “Coronavirus Infection Survey” (CIS) respondents. The CIS randomly samples individuals from the UK (private households only) with a confirmed/suspected SARS-CoV-2 infection and follows them monthly for up to a year. Interviewers asked participants about 12 current symptoms. Time to symptom discontinuation, defined as the date of the first study visit in which the participant did not report any symptoms conditional on that individual remaining symptom free through their next visit, was estimated using Kaplan-Meier analysis. Study data were collected through December 14, 2020.
Summary of Main Findings
Of the 9,063 participants included, 22.1% (95% CI: 21.2-23.2) had at least one persistent symptom 5 weeks following SARS-CoV-2 infection, and 9.8% (95% CI: 7.4-13.1) had at least one symptom through 12 weeks. At 5 weeks, the most common symptoms were fatigue (12.7%), cough (12.4%), headache (11.1%), loss of taste/smell (10.4%), and myalgia (muscle aches; 8.8%). Symptom prevalence at 5 weeks was highest among individuals aged 25-69 years, but it is notable that more than 10% of children aged 2-11 years and about 15% of children aged 12-16 years had at least one persistent symptom. Using the symptom prevalence reported in this study and UK COVID-19 incidence rates, it was estimated that as of December 27, 2020 about 301,000 people in England had persistent COVID-19 symptoms (95% CI: 274,000-329,000).
Data were collected from a national, prospective survey.
This is a preliminary report with limited methodologic content. The study population was not well characterized, especially in terms of severity of initial SARS-CoV-2 infection, age distribution, prevalence of comorbidities or pre-existing symptoms, socio-economic status, and duration of follow-up, and, thus, we cannot speak to the generalizability of the findings. Additionally, though the data were collected from a randomly selected population in the UK, it is not clear how many people declined to participate or were lost to follow-up during the study, which is essential to estimate potential selection bias – the degree to which the study population differs from the target population. Further, the data are unweighted and do not take the sample design into account, so estimates should not be interpreted as population-level estimates. There is no data on severity of symptoms, only on presence of them, so it is not clear whether symptoms improved over time, despite never being completely resolved.
This is one of the largest studies to-date, summarizing prospectively collected data on long-COVID symptoms, and the study will continue to provide important insight into the long-term effects of COVID-19.
This review was posted on: 5 March 2021