Resurgence of SARS-CoV-2 in England: detection by community antigen surveillance
REACT Study Investigators: None., Riley S., Ainslie K., Eales O., Walters C., Wang H., Atchison C., Fronterre C., Diggle P., Ashby D., Donnelly C., Cooke G., Barclay W., Ward H., Darzi A., Elliott P.
Summary Background Based on cases and deaths, transmission of SARS-CoV-2 in England peaked in late March and early April 2020 and then declined until the end of June. Since the start of July, cases have increased, while deaths have continued to decrease. Methods We report results from 594,000 swabs tested for SARS-CoV-2 virus obtained from a representative sample of people in England over four rounds collected regardless of symptoms, starting in May 2020 and finishing at the beginning of September 2020. Swabs for the most recent two rounds were taken between 24th July and 11th August and for round 4 between 22nd August and 7th September. We estimate weighted overall prevalence, doubling times between and within rounds and associated reproduction numbers. We obtained unweighted prevalence estimates by sub-groups: age, sex, region, ethnicity, key worker status, household size, for which we also estimated odds of infection. We identified clusters of swab-positive participants who were closer, on average, to other swab-positive participants than would be expected. Findings Over all four rounds of the study, we found that 72% (67%, 76%) of swab-positive individuals were asymptomatic at the time of swab and in the week prior. The epidemic declined between rounds 1 and 2, and rounds 2 and 3. However, the epidemic was increasing between rounds 3 and 4, with a doubling time of 17 (13, 23) days corresponding to an R value of 1.3 (1.2, 1.4). When analysing round 3 alone, we found that the epidemic had started to grow again with 93% probability. Using only the most recent round 4 data, we estimated a doubling time of 7.7 (5.5, 12.7) days, corresponding to an R value of 1.7 (1.4, 2.0). Cycle threshold values were lower (viral loads were higher) for rounds 1 and 4 than they were for rounds 2 and 3. In round 4, we observed the highest prevalence in participants aged 18 to 24 years at 0.25% (0.16%, 0.41%), increasing from 0.08% (0.04%, 0.18%) in round 3. We observed the lowest prevalence in those aged 65 and older at 0.04% (0.02%, 0.06%) which was stable compared with round 3. Participants of Asian ethnicity had elevated odds of infection. We identified clusters in and around London, transient clusters in the Midlands, and an expanding area of clustering in the North West and more recently in Yorkshire and the Humber. Interpretation Although low levels of transmission persisted in England through to mid-summer 2020, the prevalence of SARS-CoV-2 is now increasing. We found evidence of accelerating transmission at the end of August and beginning of September. Representative community antigen sampling can increase situational awareness and help improve public health decision making even at low prevalence.