On 21 January 2021, the Office of National Statistics (ONS) published its latest experimental estimates of long COVID symptom prevalence in the general population, using data from 9,063 respondents to the UK Coronavirus Infection Survey (CIS) testing positive for COVID-19 to 14 December 2020.
The CIS is a survey sample of respondents randomly selected from the UK population (excluding communal establishments) who are followed-up weekly for the first month from enrolment, followed by monthly for a up to a year. At each visit, respondents are swab tested for COVID-19 and describe their current symptoms (from a list of 12 common COVID-19 symptoms) to the interviewer.
The ONS estimated time-to-symptom-discontinuation using survival analysis techniques. Discontinuation was defined as the first post-infection occurrence of the respondent not reporting any symptoms for two consecutive visits (that is, the visit defining the date of discontinuation plus the next one). To allow time from infection to symptom onset, the ONS tracked symptoms reported at the visit where the respondent tested positive for COVID-19 or within the next month.