Infectious disease surveillance needs for the United States: lessons from COVID-19 (preprint)

The COVID-19 pandemic has highlighted the need to upgrade systems for infectious disease surveillance and forecasting and modeling of the spread of infection, both of which inform evidence-based public health guidance and policies. Here, we discuss requirements for an effective surveillance system to support decision making during a pandemic, drawing on the lessons of COVID-19 in the U.S., while looking to jurisdictions in the U.S. and beyond to learn lessons about the value of specific data types. In this report, we define the range of decisions for which surveillance data are required, the data elements needed to inform these decisions and to calibrate inputs and outputs of transmission-dynamic models, and the types of data needed to inform decisions by state, territorial, local, and tribal health authorities. We define actions needed to ensure that such data will be available and consider the contribution of such efforts to improving health equity.

Combining models to generate consensus medium-term projections of hospital admissions, occupancy and deaths relating to COVID-19 in England (preprint)

Mathematical modelling has played an important role in offering informed advice during the COVID-19 pandemic. In England, a cross government and academia collaboration generated Medium-Term Projections (MTPs) of possible epidemic trajectories over the future 4-6 weeks from a collection of epidemiological models.In this paper we outline this collaborative modelling approach and evaluate the accuracy of the combined and individual model projections against the data over the period November 2021-December 2022 when various Omicron subvariants were spreading across England. Using a number of statistical methods, we quantify the predictive performance of the model projections for both the combined and individual MTPs, by evaluating the point and probabilistic accuracy. Our results illustrate that the combined MTPs, produced from an ensemble of heterogeneous epidemiological models, were a closer fit to the data than the individual models during the periods of epidemic growth or decline, with the 90% confidence intervals widest around the epidemic peaks. We also show that the combined MTPs increase the robustness and reduce the biases associated with a single model projection. Learning from our experience of ensemble modelling during the COVID-19 epidemic, our findings highlight the importance of developing cross-institutional multi-model infectious disease hubs for future outbreak control.

Forecasting regional-level COVID-19 hospitalisation in England as an ordinal variable using the machine learning method (preprint)

Background COVID-19 causes substantial pressure on healthcare, with many healthcare systems now needing to prepare for and mitigate the consequences of surges in demand caused by multiple overlapping waves of infections. Therefore, public health agencies and health system managers also now benefit from short-term forecasts for respiratory infections that allow them to manage services better. However, the availability of easily implemented effective tools for generating precise forecasts at the individual regional level still needs to be improved. Methods We extended prior work on influenza to forecast regional COVID-19 hospitalisations in England for the period from 19th March 2020 to 31st December 2022, treating the number of hospital admissions in each region as an ordinal variable. We further developed the XGBoost model used previously to forecast influenza to enable it to exploit the ordering information in ordinal hospital admission levels. We incorporated different types of data as predictors: epidemiological data including weekly region COVID-19 cases and hospital admissions, weather conditions and mobility data for multiple categories of locations (e.g., parks, workplaces, etc). The impact of different discretisation methods and the number of ordinal levels was also considered. Results We find that the inclusion of weather data consistently increases the accuracy of our forecasts compared with models that rely only on the intrinsic epidemiological data, but only by a small amount. Mobility data brings about a more substantial increase in our forecasts. When both weather and mobility data are used in addition to the epidemiological data, the results are very similar to the model with only epidemiological data and mobility data. Conclusion Accurate ordinal forecasts of COVID-19 hospitalisations can be obtained using XGBoost and mobility data. While uniform ordinal levels show higher apparent accuracy, we recommend N-tile ordinal levels which contain far richer information.

Health inequalities in SARS-CoV-2 infection during the second wave in England: REACT-1 study (preprint)

Abstract Background The rapid spread of SARS-CoV-2 infection caused high levels of hospitalisation and deaths in late 2020 and early 2021 during the second wave in England. Severe disease during this period was associated with marked health inequalities across ethnic and sociodemographic subgroups. Methods We analysed risk factors for test-positivity for SARS-CoV-2, based on self-administered throat and nose swabs in the community during rounds 5 to 10 of the REal-time Assessment of Community Transmission-1 (REACT-1) study between 18 September 2020 and 30 March 2021. Results Compared to white ethnicity, people of Asian and black ethnicity had a higher risk of infection during rounds 5 to 10, with odds of 1.46 (1.27, 1.69) and 1.35 (1.11, 1.64) respectively. Among ethnic subgroups, the highest and the second-highest odds were found in Bangladeshi and Pakistan participants at 3.29 (2.23, 4.86) and 2.15 (1.73, 2.68) respectively when compared to British whites. People in larger (compared to smaller) households had higher odds of infection. Health care workers with direct patient contact and care home workers showed higher odds of infection compared to other essential/key workers. Additionally, the odds of infection among participants in public-facing activities or settings were greater than among those not working in those activities or settings. Interpretation Planning for future severe waves of respiratory pathogens should include policies to reduce inequality in risk of infection by ethnicity, household size, and occupational activity.

Key challenges for the surveillance of respiratory viruses: transitioning out of the acute phase of the SARS-CoV-2 pandemic (preprint)

To support the ongoing management of viral respiratory diseases, many countries are moving towards an integrated model of surveillance for SARS-CoV-2, influenza, and other respiratory pathogens. While many surveillance approaches catalysed by the COVID-19 pandemic provide novel epidemiological insight, continuing them as implemented during the pandemic is unlikely to be feasible for non-emergency surveillance, and many have already been scaled back. Furthermore, given anticipated co-circulation of SARS-CoV-2 and influenza, surveillance activities in place prior to the pandemic require review and adjustment to ensure their ongoing value for public health. In this perspective, we highlight key challenges for the development of integrated models of surveillance. We discuss the relative strengths and limitations of different surveillance practices and studies, their contribution to epidemiological assessment, forecasting, and public health decision making.

Global surveillance of novel SARS-CoV-2 variants (preprint)

Earlier global detection of novel SARS-CoV-2 variants gives governments more time to respond. However, few countries can implement timely national surveillance resulting in gaps in monitoring. The UK implemented large-scale community and hospital surveillance, but experience suggests it may be faster to detect new variants through testing UK arrivals for surveillance. We developed simulations of the emergence and importation of novel variants with a range of infection hospitalisation rates (IHR) to the UK. We compared time taken to detect the variant though testing arrivals at UK borders, hospital admissions, and the general community. We found that sampling 10 to 50% of arrivals at UK borders could confer a speed advantage of 3.5 to 6 weeks over existing community surveillance, and 1.5 to 5 weeks (depending on IHR) over hospital testing. We conclude that directing limited global capacity for surveillance to highly connected ports could speed up global detection of novel SARS-CoV-2 variants.

Combining models to generate a consensus effective reproduction number R for the COVID-19 epidemic status in England (preprint)

The effective reproduction number R was widely accepted as a key indicator during the early stages of the COVID-19 pandemic. In the UK, the R value published on the UK Government Dashboard has been generated as a combined value from an ensemble of fourteen epidemiological models via a collaborative initiative between academia and government. In this paper we outline this collaborative modelling approach and illustrate how, by using an established combination method, a combined R estimate can be generated from an ensemble of epidemiological models. We show that this R is robust to different model weighting methods and ensemble size and that using heterogeneous data sources for validation increases its robustness and reduces the biases and limitations associated with a single source of data. We discuss how R can be generated from different data sources and is therefore a good summary indicator of the current dynamics in an epidemic.

Epidemiological impact of a large number of incorrect negative SARS-CoV-2 test results in South West England during September and October 2021 (preprint)

Background In England, free testing for COVID-19 was widely available from early in the pandemic until 1 April 2022. Based on apparent differences in the rate of positive PCR tests at a single laboratory compared to the rest of the laboratory network, we hypothesised that a substantial number of UK PCR tests processed during September and October 2021 may have been incorrectly reported as negative, compared with the rest of the laboratory network. We investigate the epidemiological impact of this incident. Methods We estimate the additional number of COVID-19 cases that would have been reported had the sensitivity of the laboratory test procedure not dropped for the period 2 September to 12 October. In addition, by making comparisons between the most affected local areas and comparator populations, we estimate the number of additional infections, cases, hospitalisations and deaths that could have occurred as a result of increased transmission due to the misclassification of tests. Results We estimate that around 39,000 tests may have been incorrectly classified during this period and, as a direct result of this incident, the most affected areas in the South West could have experienced between 6,000 and 34,000 additional reportable cases, with a central estimate of around 24,000 additional reportable cases. Using modelled relationships between key variables, we estimate that this central estimate could have translated to approximately 55,000 additional infections, which means that each incorrect negative test likely led to just over two additional infections. In those same geographical areas, our results also suggest an increased number of admissions and deaths. Conclusion The incident is likely to have had a measurable impact on cases and infections in the affected areas in the South West of England.

Quantifying changes in the IFR and IHR over 23 months of the SARS-CoV-2 pandemic in England (preprint)

BackgroundThe relationship between prevalence of infection and severe outcomes such as hospitalisation and death changed over the course of the COVID-19 pandemic. The REal-time Assessment of Community Transmission-1 (REACT-1) study estimated swab positivity in England approximately monthly from May 2020 to 31 March 2022. This period covers widespread circulation of the original strain, the emergence of the Alpha, Delta and Omicron variants and the rollout of Englands mass vaccination campaign. MethodsHere, we explore this changing relationship between prevalence of swab positivity and the infection fatality rate (IFR) and infection hospitalisation rate (IHR) over 23 months of the pandemic in England, using publicly available data for the daily number of deaths and hospitalisations, REACT-1 swab positivity data, time-delay models and Bayesian P-spline models. We analyse data for all age groups together, as well as in two sub-groups: those aged 65 and over and those aged 64 and under. ResultsDuring 2020, we estimated the IFR to be 0.67% and the IHR to be 2.6%. By late-2021/early-2022 the IFR and IHR had both decreased to 0.097% and 0.76% respectively. Continuous estimates of the IFR and IHR of the virus were observed to increase during the periods of Alpha and Deltas emergence. During periods of vaccination rollout, and the emergence of the Omicron variant, the IFR and IHR of the virus decreased. During 2020, we estimated a time-lag of 19 days between hospitalisation and swab positivity, and 26 days between deaths and swab positivity. By late-2021/early-2022 these time-lags had decreased to 7 days for hospitalisations, and 18 days for deaths. ConclusionEven though many populations have high levels of immunity to SARS-CoV-2 from vaccination and natural infection, waning of immunity and variant emergence will continue to be an upwards pressure on IHR and IFR. As investments in community surveillance are scaled back, alternative methods should be developed to accurately track the ever changing relationship between infection, hospitalisation and death.

Validity of self-testing at home with rapid SARS-CoV-2 antibody detection by lateral flow immunoassay (preprint)

Background: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibody lateral flow immunoassays (LFIA) can be carried out in the home and have been used as an affordable and practical approach to large-scale antibody prevalence studies. However, assay performance differs from that of high-throughput laboratory-based assays which can be highly sensitive. We explore LFIA performance under field conditions compared to laboratory-based ELISA and assess the potential of LFIAs to identify people who lack functional antibodies following infection or vaccination. Methods: Field evaluation of a self-administered LFIA test (Fortress, NI) among 3758 participants from the REal-time Assessment of Community Transmission-2 (REACT-2) study in England selected based on vaccination history and previous LFIA result to ensure a range of antibody titres. In July 2021, participants performed, at home, a self-administered LFIA on finger-prick blood, reported and submitted a photograph of the result, and provided a self-collected capillary blood sample (Tasso-SST) for serological assessment of IgG antibodies to the spike protein using the Roche Elecsys Anti-SARS-CoV-2 assay. We compared the self-administered and reported LFIA result to the quantitative Roche assay and checked the reading of the LFIA result with an automated image analysis (ALFA). In a subsample of 250 participants, we compared the results to live virus neutralisation. Results: Almost all participants (3593/3758, 95.6%) had been vaccinated or reported prior infection, with most having received one (862, 22.9%) or two (2430, 64.7%) COVID-19 vaccine doses. Overall, 2777/3758 (73.9%) were positive on self-reported LFIA, 2811/3457 (81.3%) positive by LFIA when ALFA-reported, and 3622/3758 (96.4%) positive on Roche anti-S (using the manufacturer reference standard threshold for positivity of 0.8 U ml-1). Live virus neutralisation was detected in 169 of 250 randomly selected samples (67.6%); 133/169 were positive with self-reported LFIA (sensitivity 78.7%; 95% CI 71.8, 84.6), 142/155 (91.6%; 86.1, 95.5) with ALFA, and 169 (100%; 97.8, 100.0) with Roche anti-S. There were 81 samples with no detectable virus neutralisation; 47/81 were negative with self-reported LFIA (specificity 58.0%; 95% CI 46.5, 68.9), 34/75 (45.3%; 33.8, 57.3) with ALFA, and 0/81 (0%; 0.0, 4.5) with Roche anti-S. All 250 samples remained positive with Roche anti-S when the threshold was increased to 1000U ml-1. Conclusions: Self-administered LFIA can provide insights into population patterns of infection and vaccine response, and sensitivity can be improved with automated reading of the result. The LFIA is less sensitive than a quantitative antibody test, but the positivity in LFIA correlates better than the quantitative ELISA with virus neutralisation.

Trends in SARS-CoV-2 infection prevalence during England's roadmap out of lockdown, January to July 2021 (preprint)

BackgroundFollowing rapidly rising COVID-19 case numbers, England entered a national lockdown on 6 January 2021, with staged relaxations of restrictions from 8 March 2021 onwards. AimWe characterise how the lockdown and subsequent easing of restrictions affected trends in SARS-CoV-2 infection prevalence. MethodsOn average, risk of infection is proportional to infection prevalence. The REal-time Assessment of Community Transmission-1 (REACT-1) study is a repeat cross-sectional study of over 98,000 people every round (rounds approximately monthly) that estimates infection prevalence in England. We used Bayesian P-splines to estimate prevalence and the time-varying reproduction number (Rt) nationally, regionally and by age group from round 8 (beginning 6 January 2021) to round 13 (ending 12 July 2021) of REACT-1. As a comparator, a separate segmented-exponential model was used to quantify the impact on Rt of each relaxation of restrictions. ResultsFollowing an initial plateau of 1.54% until mid-January, infection prevalence decreased until 13 May when it reached a minimum of 0.09%, before increasing until the end of the study to 0.76%. Following the first easing of restrictions, which included schools reopening, the reproduction number Rt increased by 82% (55%, 108%), but then decreased by 61% (82%, 53%) at the second easing of restrictions, which was timed to match the Easter school holidays. Following further relaxations of restrictions, the observed Rt increased steadily, though the increase due to these restrictions being relaxed was masked by the effects of vaccination and the rapid rise of Delta. There was a high degree of synchrony in the temporal patterns of prevalence between regions and age groups. ConclusionHigh-resolution prevalence data fitted to P-splines allowed us to show that the lockdown was highly effective at reducing risk of infection with school holidays/closures playing a significant part.

The new normal? Dynamics and scale of the SARS-CoV-2 variant Omicron epidemic in England (preprint)

Summary The SARS-CoV-2 pandemic has been characterised by the regular emergence of genomic variants which have led to substantial changes in the epidemiology of the virus. With natural and vaccine-induced population immunity at high levels, evolutionary pressure favours variants better able to evade SARS-CoV-2 neutralising antibodies. The Omicron variant was first detected in late November 2021 and exhibited a high degree of immune evasion, leading to increased infection rates in many countries. However, estimates of the magnitude of the Omicron wave have relied mainly on routine testing data, which are prone to several biases. Here we infer the dynamics of the Omicron wave in England using PCR testing and genomic sequencing obtained by the REal-time Assessment of Community Transmission-1 (REACT-1) study, a series of cross-sectional surveys testing random samples of the population of England. We estimate an initial peak in national Omicron prevalence of 6.89% (5.34%, 10.61%) during January 2022, followed by a resurgence in SARS-CoV-2 infections in England during February-March 2022 as the more transmissible Omicron sub-lineage, BA.2 replaced BA.1 and BA.1.1. Assuming the emergence of further distinct genomic variants, intermittent epidemics of similar magnitude as the Omicron wave may become the ‘new normal’.

Consistent Pattern of Epidemic Slowing Across Many Geographies Led to Longer, Flatter Initial Waves of the COVID-19 Pandemic (preprint)

To define appropriate planning scenarios for future pandemics of respiratory pathogens, it is important to understand the initial transmission dynamics of COVID-19 during 2020. Here, we fit an age-stratified compartmental model with a flexible underlying transmission term to daily COVID-19 death data from states in the contiguous U.S. and to national and sub-national data from around the world. The daily death data of the first months of the COVID-19 pandemic was categorized into one of four main types: "spring single-peak profile", "summer single-peak profile", "spring/summer two-peak profile" and "broad with shoulder profile". We estimated a reproduction number R as a function of calendar time tc and as a function of time since the first death reported in that population (local pandemic time, tp). Contrary to the multiple categories and range of magnitudes in death incidence profiles, the R(tp) profiles were much more homogeneous. We find that in both the contiguous U.S. and globally, the initial value of both R(tc) and R(tp) was substantial: at or above two. However, during the early months, pandemic time R(tp) decreased exponentially to a value that hovered around one. This decrease was accompanied by a reduction in the variance of R(tp). For calendar time R(tc), the decrease in magnitude was slower and non-exponential, with a smaller reduction in variance. Intriguingly, similar trends of exponential decrease and reduced variance were not observed in raw death data. Our findings suggest that the combination of specific government responses and spontaneous changes in behaviour ensured that transmissibility dropped, rather than remaining constant, during the initial phases of a pandemic. Future pandemic planning scenarios should be based on models that assume similar decreases in transmissibility, which lead to longer epidemics with lower peaks when compared with models based on constant transmissibility.

Appropriately smoothing prevalence data to inform estimates of growth rate and reproduction number (preprint)

The time-varying reproduction number ( R t ) can change rapidly over the course of a pandemic due to changing restrictions, behaviours, and levels of population immunity. Many methods exist that allow the estimation of R t from case data. However, these are not easily adapted to point prevalence data nor can they infer R t across periods of missing data. We developed a Bayesian P-spline model suitable for fitting to a wide range of epidemic time-series, including point-prevalence data. We demonstrate the utility of the model by fitting to periodic daily SARS-CoV-2 swab-positivity data in England from the first 7 rounds (May 2020 – December 2020) of the REal-time Assessment of Community Transmission-1 (REACT-1) study. Estimates of R t over the period of two subsequent rounds (6-8 weeks) and single rounds (2-3 weeks) inferred using the Bayesian P-spline model were broadly consistent with estimates from a simple exponential model, with overlapping credible intervals. However, there were sometimes substantial differences in point estimates. The Bayesian P-spline model was further able to infer changes in R t over shorter periods tracking a temporary increase above one during late-May 2020, a gradual increase in R t over the summer of 2020 as restrictions were eased, and a reduction in R t during England’ s second national lockdown followed by an increase as the Alpha variant surged. The model is robust against both under-fitting and over-fitting and is able to interpolate between periods of available data; it is a particularly versatile model when growth rate can change over small timescales, as in the current SARS-CoV-2 pandemic. This work highlights the importance of pairing robust methods with representative samples to track pandemics.

SARS-CoV-2 lineage dynamics in England from September to November 2021: high diversity of Delta sub-lineages and increased transmissibility of AY.4.2 (preprint)

Since the emergence of SARS-CoV-2, evolutionary pressure has driven large increases in the transmissibility of the virus. However, with increasing levels of immunity through vaccination and natural infection the evolutionary pressure will switch towards immune escape. Here we present phylogenetic relationships and lineage dynamics within England (a country with high levels of immunity), as inferred from a random community sample of individuals who provided a self-administered throat and nose swab for rt-PCR testing as part of the REal-time Assessment of Community Transmission-1 (REACT-1) study. From 9 to 27 September 2021 (round 14) and 19 October to 5 November 2021 (round 15), all lineages sequenced within REACT-1 were Delta or a Delta sub-lineage with 44 unique lineages identified. The proportion of the original Delta variant (B.1.617.2) was found to be increasing between September and November 2021, which may reflect an increasing number of sub-lineages which have yet to be identified. The proportion of B.1.617.2 was greatest in London, which was further identified as a region with an increased level of genetic diversity. The Delta sub-lineage AY.4.2 was found to be robustly increasing in proportion, with a reproduction number 15% (8%, 23%) greater than its parent and most prevalent lineage, AY.4. Both AY.4.2 and AY.4 were found to be geographically clustered in September but this was no longer the case by late October/early November, with only the lineage AY.6 exhibiting clustering towards the South of England. Though no difference in the viral load based on cycle threshold (Ct) values was identified, a lower proportion of those infected with AY.4.2 had symptoms for which testing is usually recommend (loss or change of sense of taste, loss or change of sense of smell, new persistent cough, fever), compared to AY.4 (p = 0.026). The evolutionary rate of SARS-CoV-2, as measured by the mutation rate, was found to be slowing down during the study period, with AY.4.2 further found to have a reduced mutation rate relative to AY.4. As SARS-CoV-2 moves towards endemicity and new variants emerge, genomic data obtained from random community samples can augment routine surveillance data without the potential biases introduced due to higher sampling rates of symptomatic individuals.

REACT-1 study round 14: High and increasing prevalence of SARS-CoV-2 infection among school-aged children during September 2021 and vaccine effectiveness against infection in England (preprint)

Background: England experienced a third wave of the COVID-19 epidemic from end May 2021 coinciding with the rapid spread of Delta variant. Since then, the population eligible for vaccination against COVID-19 has been extended to include all 12-15-year-olds, and a booster programme has been initiated among adults aged 50 years and over, health care and care home workers, and immunocompromised people. Meanwhile, schoolchildren have returned to school often with few COVID-19-related precautions in place. Methods: In the REal-time Assessment of Community Transmission-1 (REACT-1) study, throat and nose swabs were sent to non-overlapping random samples of the population aged 5 years and over in England. We analysed prevalence of SARS-CoV-2 using reverse transcription-polymerase chain reaction (RT-PCR) swab-positivity data from REACT-1 round 14 (between 9 and 27 September 2021). We combined results for round 14 with round 13 (between 24 June and 12 July 2021) and estimated vaccine effectiveness and prevalence of swab-positivity among double-vaccinated individuals. Unlike all previous rounds, in round 14, we switched from dry swabs transported by courier on a cold chain to wet swabs using saline. Also, at random, 50% of swabs (not chilled until they reached the depot) were transported by courier and 50% were sent through the priority COVID-19 postal service. Results: We observed stable or rising prevalence (with an R of 1.03 (0.94, 1.14) overall) during round 14 with a weighted prevalence of 0.83% (0.76%, 0.89%). The highest weighted prevalence was found in children aged 5 to 12 years at 2.32% (1.96%, 2.73%) and 13 to 17 years at 2.55% (2.11%, 3.08%). All positive virus samples analysed correspond to the Delta variant or sub-lineages of Delta with one instance of the E484K escape mutation detected. The epidemic was growing in those aged 17 years and under with an R of 1.18 (1.03, 1.34), but decreasing in those aged 18 to 54 years with an R of 0.81 (0.68, 0.97). For all participants and all vaccines combined, vaccine effectiveness against infection (rounds 13 and 14 combined) was estimated to be 62.8% (49.3%, 72.7%) after two doses compared to unvaccinated people when adjusted for round, age, sex, index of multiple deprivation, region and ethnicity; the adjusted estimate was 44.8% (22.5%, 60.7%) for AstraZeneca and 71.3% (56.6%, 81.0%) for Pfizer-BioNTech, and for all vaccines combined it was 66.4% (49.6%, 77.6%) against symptomatic infection (one or more of 26 surveyed symptoms in month prior). Across rounds 13 and 14, weighted prevalence of swab-positivity was 0.55% (0.50%, 0.61%) for those who received their second dose 3-6 months before their swab compared to 0.35% (0.31%, 0.40%) for those whose second dose was within 3 months of their swab. However, the prevalence was lower in those with one or two doses of vaccine than in unvaccinated individuals at 1.76% (1.60%, 1.95%). In round 14, age group, region, key worker status, and household size jointly contributed to the risk of higher prevalence of swab-positivity. Discussion: In September 2021 infections were increasing exponentially in the 5-to-17-year age group coinciding with the start of the autumn school term in England. Relatively few schoolchildren aged 5 to 17 years have been vaccinated in the UK though single doses are now being offered to those aged 12 years and over. In adults, the higher prevalence of swab-positivity following two doses of vaccine within 3 to 6 months supports the use of a booster vaccine. It is important that the vaccination programme maintains high coverage and reaches children and unvaccinated or partially vaccinated adults to reduce transmission and associated disruptions to work and education.

Exponential growth, high prevalence of SARS-CoV-2 and vaccine effectiveness associated with Delta variant in England during May to July 2021 (preprint)

BackgroundThe prevalence of SARS-CoV-2 infection continues to drive rates of illness and hospitalisations despite high levels of vaccination, with the proportion of cases caused by the Delta lineage increasing in many populations. As vaccination programs roll out globally and social distancing is relaxed, future SARS-CoV-2 trends are uncertain. MethodsWe analysed prevalence trends and their drivers using reverse transcription-polymerase chain reaction (RT-PCR) swab-positivity data from round 12 (between 20 May and 7 June 2021) and round 13 (between 24 June and 12 July 2021) of the REal-time Assessment of Community Transmission-1 (REACT-1) study, with swabs sent to non-overlapping random samples of the population ages 5 years and over in England. ResultsWe observed sustained exponential growth with an average doubling time in round 13 of 25 days (lower Credible Interval of 15 days) and an increase in average prevalence from 0.15% (0.12%, 0.18%) in round 12 to 0.63% (0.57%, 0.18%) in round 13. The rapid growth across and within rounds appears to have been driven by complete replacement of Alpha variant by Delta, and by the high prevalence in younger less-vaccinated age groups, with a nine-fold increase between rounds 12 and 13 among those aged 13 to 17 years. Prevalence among those who reported being unvaccinated was three-fold higher than those who reported being fully vaccinated. However, in round 13, 44% of infections occurred in fully vaccinated individuals, reflecting imperfect vaccine effectiveness against infection despite high overall levels of vaccination. Using self-reported vaccination status, we estimated adjusted vaccine effectiveness against infection in round 13 of 49% (22%, 67%) among participants aged 18 to 64 years, which rose to 58% (33%, 73%) when considering only strong positives (Cycle threshold [Ct] values < 27); also, we estimated adjusted vaccine effectiveness against symptomatic infection of 59% (23%, 78%), with any one of three common COVID-19 symptoms reported in the month prior to swabbing. Sex (round 13 only), ethnicity, household size and local levels of deprivation jointly contributed to the risk of higher prevalence of swab-positivity. DiscussionFrom end May to beginning July 2021 in England, where there has been a highly successful vaccination campaign with high vaccine uptake, infections were increasing exponentially driven by the Delta variant and high infection prevalence among younger, unvaccinated individuals despite double vaccination continuing to effectively reduce transmission. Although slower growth or declining prevalence may be observed during the summer in the northern hemisphere, increased mixing during the autumn in the presence of the Delta variant may lead to renewed growth, even at high levels of vaccination.

Characterising the persistence of RT-PCR positivity and incidence in a community survey of SARS-CoV-2 (preprint)

BackgroundCommunity surveys of SARS-CoV-2 RT-PCR swab-positivity provide prevalence estimates largely unaffected by biases from who presents for routine case testing. The REal-time Assessment of Community Transmission-1 (REACT-1) has estimated swab-positivity approximately monthly since May 2020 in England from RT-PCR testing of self-administered throat and nose swabs in random non-overlapping cross-sectional community samples. Estimating infection incidence from swab-positivity requires an understanding of the persistence of RT-PCR swab positivity in the community. MethodsDuring round 8 of REACT-1 from 6 January to 22 January 2021, of the 2,282 participants who tested RT-PCR positive, we recruited 896 (39%) from whom we collected up to two additional swabs for RT-PCR approximately 6 and 9 days after the initial swab. We estimated sensitivity and duration of positivity using an exponential model of positivity decay, for all participants and for subsets by initial N-gene cycle threshold (Ct) value, symptom status, lineage and age. Estimates of infection incidence were obtained for the entire duration of the REACT-1 study using P-splines. ResultsWe estimated the overall sensitivity of REACT-1 to detect virus on a single swab as 0.79 (0.77, 0.81) and median duration of positivity following a positive test as 9.7 (8.9, 10.6) days. We found greater median duration of positivity where there was a low N-gene Ct value, in those exhibiting symptoms, or for infection with the Alpha variant. The estimated proportion of positive individuals detected on first swab, P0, was found to be higher for those with an initially low N-gene Ct value and those who were pre-symptomatic. When compared to swab-positivity, estimates of infection incidence over the duration of REACT-1 included sharper features with evident transient increases around the time of key changes in social distancing measures. DiscussionHome self-swabbing for RT-PCR based on a single swab, as implemented in REACT-1, has high overall sensitivity. However, participants time-since-infection, symptom status and viral lineage affect the probability of detection and the duration of positivity. These results validate previous efforts to estimate incidence of SARS-CoV-2 from swab-positivity data, and provide a reliable means to obtain community infection estimates to inform policy response.

Global predictions of short- to medium-term COVID-19 transmission trends : a retrospective assessment (preprint)

From 8th March to 29th November 2020, we produced weekly estimates of SARS-CoV-2 transmissibility and forecasts of deaths due to COVID-19 for 81 countries with evidence of sustained transmission. We also developed a novel heuristic to combine weekly estimates of transmissibility to produce forecasts over a 4-week horizon. We evaluated the robustness of the forecasts using relative error, coverage probability, and comparisons with null models. During the 39-week period covered by this study, both the short- and medium-term forecasts captured well the epidemic trajectory across different waves of COVID-19 infections with small relative errors over the forecast horizon. The model was well calibrated with 56.3\% and 45.6\% of the observations lying in the 50\% Credible Interval in 1-week and 4-week ahead forecasts respectively. We could accurately characterise the overall phase of the epidemic up to 4-weeks ahead in 84.9\% of country-days. The medium-term forecasts can be used in conjunction with the short-term forecasts of COVID-19 mortality as a useful planning tool as countries continue to relax public health measures.

Global predictions of short- to medium-term COVID-19 transmission trends : a retrospective assessment (preprint)

BackgroundAs of July 2021, more than 180,000,000 cases of COVID-19 have been reported across the world, with more than 4 million deaths. Mathematical modelling and forecasting efforts have been widely used to inform policy-making and to create situational awareness. Methods and FindingsFrom 8th March to 29th November 2020, we produced weekly estimates of SARS-CoV-2 transmissibility and forecasts of deaths due to COVID-19 for countries with evidence of sustained transmission. The estimates and forecasts were based on an ensemble model comprising of three models that were calibrated using only the reported number of COVID-19 cases and deaths in each country. We also developed a novel heuristic to combine weekly estimates of transmissibility and potential changes in population immunity due to infection to produce forecasts over a 4-week horizon. We evaluated the robustness of the forecasts using relative error, coverage probability, and comparisons with null models. ConclusionsDuring the 39-week period covered by this study, we produced short- and medium-term forecasts for 81 countries. Both the short- and medium-term forecasts captured well the epidemic trajectory across different waves of COVID-19 infections with small relative errors over the forecast horizon. The model was well calibrated with 56.3% and 45.6% of the observations lying in the 50% Credible Interval in 1-week and 4-week ahead forecasts respectively. We could accurately characterise the overall phase of the epidemic up to 4-weeks ahead in 84.9% of country-days. The medium-term forecasts can be used in conjunction with the short-term forecasts of COVID-19 mortality as a useful planning tool as countries continue to relax stringent public health measures that were implemented to contain the pandemic.