Tracking COVID-19 in England and Wales: Insights from Virus Watch - a prospective community cohort study (preprint)

BackgroundVirus Watch is a prospective community cohort study of COVID-19 of 28{square},527 households in England and Wales designed to estimate the incidence of PCR-confirmed COVID-19 in those with respiratory presentations and examine symptom profiles and transmission of COVID-19 in relation to population movement and behaviour. The Office for National Statistics (ONS) COVID-19 infection survey (CIS) was the largest regular survey of COVID-19 infections and antibodies in the UK and included 227,797 households. In this analysis, we aimed to compare incidence rate estimates from the two studies to understand differences in estimates from the two study designs. MethodsWe used the Virus Watch prospective community cohort study to estimate the overall SARS-CoV-2 incidence rate and incidence rate by age in England and Wales from June 2020 to February 2023. Virus Watch data consisted of self-reported laboratory COVID-19 test results and linkage to the Second Generation Surveillance System, the UK national database for COVID-19 testing. We compared our findings with modelled incidence rates from ONS CIS using 3-day rolling Pearsons correlation to measure synchrony. Results58,628 participants were recruited into the Virus Watch study between June 2020 and March 2022, of whom 52,526 (90%) were reported to be living in England and 1,532 (2.6%) in Wales. COVID-19 incidence rates were initially similar across age groups until the Delta wave when rates increased at different magnitudes. During the Omicron BA.1, the 0-14 age group had the highest incidence rates, which shifted to the 25-44 age group with Omicron BA.2, 4, and 5 dominance. We found strong synchrony between Virus Watch and ONS CIS COVID-19 incidence estimates for England and Wales, both with and without the incorporation of linked national testing data into the Virus Watch study. In particular, the magnitude and trend of Virus Watch- and ONS-estimated rates for England were generally consistent, although Virus Watch-estimated peaks of infection during the Omicron BA.1 and 2 waves were found to be lower than estimates from the ONS. ConclusionOur findings suggest that the Virus Watch research approach is a low-cost and effective method for on-going surveillance of COVID-19 regardless of the availability of national testing in the UK. Similar approaches can also be utilised by low-resource settings to provide accurate incidence rate estimates to better monitor and respond to COVID-19 as well as other acute respiratory diseases in the future.

Comparative effectiveness of ChAdOx1 versus BNT162b2 vaccines against SARS-CoV-2 infections in England and Wales: A cohort analysis using trial emulation in the Virus Watch community data (preprint)

Introduction: Infections of SARS-CoV-2 in vaccinated individuals have been increasing globally. Understanding the associations between vaccine type and a post-vaccination infection could help prevent further COVID-19 waves. In this paper, we use trial emulation to understand the impact of a phased introduction of the vaccine in the UK driven by vulnerability and exposure status. We estimate the comparative effectiveness of COVID-19 vaccines (ChAdOx1 versus BNT162b2) against post-vaccination infections of SARS-CoV-2 in a community setting in England and Wales. Method: Trial emulation was conducted by pooling results from six cohorts whose recruitment was staggered between 1st January 2021 and 31st March 2021 and followed until 12th November 2021. Eligibility for each trial was based upon age (18+ at the time of vaccination), without prior signs of infection or an infection within the first 14 days of the first dose. Time from vaccination of ChAdOx1 or BNT162b2 until SARS-CoV-2 infection (positive polymerase chain reaction or lateral flow test after 14 of the vaccination) was modelled using Cox proportional hazards model for each cohort and adjusted for age at vaccination, gender, minority ethnic status, clinically vulnerable status and index of multiple deprivation quintile. For those without SARS-CoV-2 infection during the study period, follow-up was until loss-of-follow-up or end of study (12th November 2021). Pooled hazard ratios were generated using random-effects meta-analysis. Results: Across six cohorts, there were a total of 21,283 participants who were eligible and vaccinated with either ChAdOx1 (n = 13,813) or BNT162b2 (n = 7,470) with a median follow-up time of 266 days (IQR: 235 - 282). By November 12th 2021, 750 (5.4%) adults who had ChAdOx1 as their vaccine experienced a SARS-CoV-2 infection, compared to 296 (4.0%) who had BNT162b2. We found that people who received ChAdOx1 vaccinations had 10.54 per 1000 people higher cumulative incidence for SARS-CoV-2 infection compared to BNT162b2 for infections during a maximum of 315 days of follow-up. When adjusted for age at vaccination, sex, minority ethnic status, index of multiple deprivation, and clinical vulnerability status, we found a pooled adjusted hazard ratio of 1.35 [HR: 1.35, 95%CI: 1.15 - 1.58], demonstrating a 35% increase in SARS-CoV-2 infections in people who received ChAdOx1 compared to BNT162b2. Discussion: We found evidence of greater effectiveness of receiving BNT162b2 compared to ChAdOx1 vaccines against SARS-CoV-2 infection in England and Wales during a time period when Delta became the most prevalent variant of concern. Our findings demonstrate the importance of booster (third) doses to maintain protection and suggest that these should be prioritised to those who received ChAdOx1 as their primary course.

Relative perceived importance of different settings for SARS-CoV2 acquisition in England and Wales: Analysis of the Virus Watch Community Cohort (preprint)

We aimed to assess the relative importance of different settings for SARS-CoV2 transmission in a large community cohort. We demonstrate the importance of home, work and education as venues for transmission. In children, education was most important and in older adults essential shopping was of high importance. Our findings support public health messaging about infection control at home, advice on working from home and restrictions in different venues.

Symptom profiles and accuracy of clinical definitions for COVID-19 in the community. Results of the Virus Watch community cohort. (preprint)

Background: Understanding the symptomatology and accuracy of clinical case definitions for COVID-19 in the community is important for the initiation of Test, Trace and Isolate (TTI) and may, in future, be important for early prescription of antivirals. Methods: Virus Watch is a large community cohort with prospective daily recording of a wide range of symptoms and self-reporting of swab results (mainly undertaken through the UK TTI System). We compared frequency, severity, timing, and duration of symptoms in test positive and test negative cases. We compared the test performance of the current UK case definition used by TTI (any one of: new continuous cough, high temperature or loss of or change in sense of smell or taste) with a wider definition that also included muscle aches or chills or headache or loss of appetite. Findings: We included results from 8213 swabbed illnesses, 944 of which tested positive for COVID-19. All symptoms were more common in swab positive than swab negative illnesses and symptoms were also more severe and of longer duration. Common symptoms such as cough, headache, fatigue, muscle aches and loss of appetite occurred early in the course of illness but were also very common in test-negative illnesses. Rarer symptoms such as fever or loss or altered sense of smell or taste were often not present but were markedly more common in swab positive compared to swab negative cases. The current UK definition had a sensitivity and specificity of 81% and 47% respectively for symptomatic COVID-19 compared to 93% and 26% for the broader definition. On average cases met the broader case definition one day earlier than current definition. 1.7-fold more illnesses met the broader definition than the current case definition. Interpretation: COVID-19 is difficult to distinguish from other respiratory infections and common ailments on the basis of symptoms. Broadening the list of symptoms used to encourage engagement with TTI could moderately increase the number of infections identified and shorten delays but with a large increase in the number of tests needed and in the number of people and contacts who do not have COVID-19 but might need to self-isolate whilst awaiting results.

Serial interval of COVID-19 and the effect of Variant B.1.1.7: analyses from a prospective community cohort study (Virus Watch). (preprint)

Introduction Increased transmissibility of B.1.17 variant of concern (VOC) in the UK may explain its rapid emergence and global spread. We analysed data from putative household infector infectee pairs in the Virus Watch Community cohort study to assess the serial interval of COVID-19 and whether this was affected by emergence of the B.1.17 variant. Methods The Virus Watch study is an online, prospective, community cohort study following up entire households in England and Wales during the COVID-19 pandemic. Putative household infector infectee pairs were identified where more than one person in the household had a positive swab matched to an illness episode. Data on whether individual infections were caused by the B.1.1.7 variant were not available. We therefore developed a classification system based on the percentage of cases estimated to be due to B.1.17 in national surveillance data for different English regions and study weeks. Results Out of 24,887 illnesses reported, 915 tested positive for SARS-CoV-2 and 186 likely infector infectee pairs in 186 households amongst 372 individuals were identified. The mean COVID-19 serial interval was 3.18 (95%CI: 2.55;3.81) days. There was no significant difference (p=0.267) between the mean serial interval for Variants of Concern (VOC) hotspots (mean = 3.64 days, (95%CI: 2.55;4.73)) days and non-VOC hotspots, (mean = 2.72 days, (95%CI: 1.48;3.96)). Conclusions Our estimates of the average serial interval of COVID-19 are broadly similar to estimates from previous studies and we find no evidence that B.1.1.7 is associated with a change in serial intervals. Alternative explanations such as increased viral load, longer period of viral shedding or improved receptor binding may instead explain the increased transmissibility and rapid spread and should undergo further investigation.