Healthcare utilisation in people with long COVID: an OpenSAFELY cohort study (preprint)

BackgroundLong COVID, characterised by various symptoms and complications, potentially increases healthcare utilisation and costs. However, its impact on the NHS remains to be determined. ObjectiveThis study aims to assess the healthcare utilisation of individuals with long COVID. MethodsWith the approval of NHS England, we conducted a matched cohort study using primary and secondary care data via OpenSAFELY, a platform for analysing anonymous electronic health records. The long COVID exposure group, defined by diagnostic codes, was matched with five comparators without long COVID between Nov 2020 and Jan 2023. We compared their total healthcare utilisation from GP consultations, prescriptions, hospital admissions, A&E visits, and outpatient appointments. Healthcare utilisation and costs were evaluated using a two-part model adjusting for covariates. Using a difference-in-difference model, we also compared healthcare utilisation after long COVID with pre-pandemic records. ResultsWe identified 52,988 individuals with a long COVID diagnosis, matched to 264,867 comparators without a diagnosis. In the 12 months post-diagnosis, there was strong evidence that those with long COVID were more likely to use healthcare resources (OR: 8.07, 95% CI: 7.54 - 8.64), and have 49% more healthcare utilisation (RR: 1.49, 95% CI: 1.47 - 1.50). Our model estimated that the long COVID group had 30 healthcare visits per year (predicted mean: 29.23, 95% CI: 28.58 - 29.92), compared to 16 in the comparator group (predicted mean visits: 16.04, 95% CI: 15.73 - 16.36). Individuals with long COVID were more likely to have non-zero healthcare expenditures (OR = 7.47, 95% CI = 7.02 - 7.95), with costs being 43% higher than the comparator group (cost ratio = 1.43, 95% CI: 1.38 - 1.49). The long COVID group costs approximately {pound}2,500 per person per year (predicted mean cost: {pound}2,562.50, 95% CI: {pound}2,335.60 - {pound}2,819.22), and the comparator group costs {pound}1,500 (predicted mean cost: {pound}1,527.43, 95% CI: {pound}1,404.33 - 1,664.45.) Historically, individuals with long COVID utilised healthcare resources more frequently, but their average healthcare utilisation increased more after being diagnosed with long COVID, compared to the comparator group. ConclusionLong COVID increases healthcare utilisation and costs. Public health policies should allocate more resources towards preventing, treating, and supporting individuals with long COVID.

The impact of Long COVID on Health-Related Quality-of-life using OpenPROMPT (preprint)

Background: Long COVID is a major problem affecting patient health, the health service, and the workforce. To optimise the design of future interventions against COVID-19, and to better plan and allocate health resources, it is critical to quantify the health and economic burden of this novel condition. Methods With the approval of NHS England, we developed OpenPROMPT, a UK cohort study measuring the impact of long COVID on health-related quality-of-life (HRQoL). OpenPROMPT invited responses to Patient Reported Outcome Measures (PROMs) using a smartphone application and recruited between November 2022 and October 2023. We used the validated EuroQol EQ-5D questionnaire with the UK Value Set to develop disutility scores (1-utility) for respondents with and without Long COVID using linear mixed models, and we calculated subsequent Quality-Adjusted Life-Months (QALMs) for long COVID. Results We used data from 6,070 participants where 24.7% self-reported long COVID. In multivariable regressions, long COVID had a consistent impact on HRQoL, showing a high probability of reporting loss in quality-of-life (OR: 22, 95% CI:12.35-39.29) compared with people who did not report long COVID. Reporting a disability was the largest predictor of losses of HRQoL (OR: 60.2, 95% CI: 27.79-130.57) across survey responses. Self-reported long COVID was associated with an 0.37 QALM loss. Conclusions We found substantial impacts on quality-of-life due to long COVID, representing a major burden on patients and the health service. We highlight the need for continued support and research for long COVID, as HRQoL scores compared unfavourably to patients with conditions such as multiple sclerosis, heart failure, and renal disease.

Clinical coding of long COVID in primary care 2020-2023 in a cohort of 19 million adults: an OpenSAFELY analysis (preprint)

Background Long COVID is the patient-coined term for the persistent symptoms of COVID-19 illness for weeks, months or years following the acute infection. There is a large burden of long COVID globally from self-reported data, but the epidemiology, causes and treatments remain poorly understood. Primary care is used to help identify and treat patients with long COVID and therefore Electronic Health Records (EHRs) of past COVID-19 patients could be used to help fill these knowledge gaps. We aimed to describe those with long COVID in primary care records in England. Methods With the approval of NHS England we used routine clinical data from over 19 million adults in England linked to SARS-COV-2 test result, hospitalisation and vaccination data to describe trends in the recording of 16 clinical codes related to long COVID between November 2020 and January 2023. We calculated rates per 100,000 person-years and plotted how these changed over time. We compared crude and minimally adjusted rates of recorded long COVID in patient records between different key demographic and vaccination characteristics using negative binomial models. Findings We identified a total of 55,465 people recorded to have long COVID over the study period, with incidence of new long COVID records increasing steadily over 2021, and declining over 2022. The overall rate per 100,000 person-years was 177.5 cases in women (95% CI: 175.5-179) and 100.5 men (99.5-102). In terms of vaccination against COVID-19, the lowest rates were observed in those with 3+ vaccine doses (103.5 [95% CI: 101.5-105]). Finally, the majority of those with a long COVID record did not have a recorded positive SARS-COV-2 test 12 weeks before the long COVID record. Interpretation EHR recorded long COVID remains very low compared and incident records of long COVID declined over 2022. We found the lowest rates of recorded long COVID in people with 3 or more vaccine doses. We summarised several sources of possible bias for researchers using EHRs to study long COVID. Funding This research was supported by the National Institute for Health and Care Research (NIHR) (OpenPROMPT: COV-LT2-0073))

The macroeconomic and epidemiological impacts of Covid-19 in Pakistan. (preprint)

'Coronavirus Disease 2019' (C19) is a respiratory illness caused by 'new Coronavirus' SARS-CoV-2. The C19 pandemic, which engulfed the world in 2021, also caused a national C19 epidemic in Pakistan, who responded with initial forced lockdowns (15-30 March 2020) and a subsequent switch to a smart lockdown strategy, and, by 31 December 2020, Pakistan had managed to limit confirmed cases and case fatalities to 482,506 (456 per 100,000) and 10,176 (4.8 per 100,000). The early switch to a smart lockdown strategy, and successful follow-up move to central coordination and effective communication and enforcement of Standard Operating Procedures, was motivated by a concern over how broad-based forced lockdowns would affect poor households and day-labour. The current study aims to investigate how the national Pakistan C19 epidemic would have unfolded under an uncontrolled baseline scenario and an alternative set of controlled non-pharmaceutical intervention (NPI) policy lockdown scenarios, including health and macroeconomic outcomes. We employ a dynamically-recursive version of the IFPRI Standard Computable General Equilibrium model framework (Lofgren, Lee Harris and Robinson 2002), and a, by now, well-established epidemiological transmission-dynamic model framework (Davies, Klepac et al 2020) using Pakistan-specific 5-year age-group contact matrices on four types of contact rates, including at home, at work, at school, and at other locations (Prem, Cook & Jit 2017), to characterize an uncontrolled spread of disease. Our simulation results indicate that an uncontrolled C19 epidemic, by itself, would have led to a 0.12% reduction in Pakistani GDP (-721mn USD), and a total of 0.65mn critically ill and 1.52mn severely ill C19 patients during 2020-21, while 405,000 Pakistani citizens would have lost their lives. Since the majority of case fatalities and symptomatic cases, respectively 345,000 and 35.9mn, would have occurred in 2020, the case fatality and confirmed case numbers, observed by 31. December 2020 represents an outcome which is far better than the alternative. Case fatalities by 31. December 2020 could possibly have been somewhat improved either via a more prolonged one-off 10 week forced lockdown (66% reduction) or a 1-month forced lockdown/2-months opening intermittent lockdown strategy (33% reduction), but both sets of strategies would have carried significant GDP costs in the order of 2.2%-6.2% of real GDP.

Global spatiotemporal trends and determinants of COVID-19 vaccine acceptance on Twitter: a multilingual deep learning study in 135 countries and territories (preprint)

Background: COVID-19 vaccination has faced a range of challenges from supply-side barriers such as insufficient vaccine supply and negative information environment and demand-side barriers centring on public acceptance and confidence in vaccines. This study assessed global spatiotemporal trends in demand- and supply-side barriers to vaccine uptake using COVID-19-related social media data and explored the country-level determinants of vaccine acceptance. Methods: We accessed a total of 13,093,406 tweets sent between November 2020 and March 2022 about the COVID-19 vaccine in 90 languages from 135 countries using Meltwater (a social listening platform). Based on 8,125 manually-annotated tweets, we fine-tuned multilingual deep learning models to automatically annotate all 13,093,406 tweets. We present spatial and temporal trends in four key spheres: (1) COVID-19 vaccine acceptance; (2) confidence in COVID-19 vaccines; (3) the online information environment regarding the COVID-19 vaccine; and (4) perceived supply-side barriers to COVID-19 vaccination. Using univariate and multilevel regressions, we evaluated the association between COVID-19 vaccine acceptance on Twitter and (1) country-level characteristics regarding governance, pandemic preparedness, trust, culture, social development, and population demographics; (2) country-level COVID-19 vaccine coverage; and (3) Google search trends on adverse vaccine events. Findings: COVID-19 vaccine acceptance was high among Twitter users in Southeast Asian, Eastern Mediterranean, and Western Pacific countries, including India, Indonesia, and Pakistan. In contrast, acceptance was relatively low in high-income nations like South Korea, Japan, and the Netherlands. Spatial variations were correlated with country-level governance, pandemic preparedness, public trust, culture, social development, and demographic determinants. At the country level, vaccine acceptance sentiments expressed on Twitter predicted higher vaccine coverage. We noted the declining trend of COVID-19 vaccine acceptance among global Twitter users since March 2021, which was associated with increased searches for adverse vaccine events. Interpretation: In future pandemics, new vaccines may face the potential low-level and declining trend in acceptance, like COVID-19 vaccines, and early responses are needed. Social media mining represents a promising surveillance approach to monitor vaccine acceptance and can be validated against real-world vaccine uptake data. Keywords: COVID-19, vaccine confidence, vaccine acceptance, vaccine hesitancy, social media, machine learning

Is there a role for RDTs as we live with COVID? An assessment of different strategies (preprint)

Introduction: By 2022, high levels of past COVID-19 infections, combined with substantial levels of vaccination and the development of Omicron have shifted country strategies toward burden reduction policies. SARS-CoV-2 rapid antigen tests (RDTs) could contribute to these policies by helping rapidly detect, isolate and/or treat infections in different settings. However, the evidence to inform RDT policy choices in LMICs is limited. Method: We provide an overview of the potential impact of several RDT use cases (surveillance; testing, tracing and isolation without and with surveillance; hospital-based screening to reduce nosocomial COVID; and testing to enable earlier/expanded treatment) for a range of country settings. We use conceptual models and literature review to identify which use cases are likely to bring benefits and how these may change with outbreak characteristics. Impacts are measured through multiple outcomes related to gaining time, reducing the burden on the health system, and reducing deaths. Results: In an optimal scenario in terms of resources and capacity and with baseline parameters, we find marginal time gains of at least a week through surveillance and testing tracing and isolation with surveillance, a reduction in peak ICU or ICU admissions by 6% or more (hospital-based screening; testing, tracing and isolation), and reductions in COVID deaths by over 6% (hospital-based screening; test and treat). Time gains may be used to strengthen ICU capacity and/or boost vulnerable individuals, though only a small minority of at-risk individuals could be reached in the time available. The impact of RDTs declines with lower country resources and capacity, more transmissible or immune-escaping variants and reduced test sensitivity. Conclusion: RDTs alone are unlikely to dramatically reduce the burden of COVID-19 in LMICs, though they may have an important role alongside other interventions such as vaccination, therapeutic drugs, improved healthcare capacity and non-pharmaceutical measures.

Optimal Respiratory Syncytial Virus intervention programmes using Nirsevimab in England and Wales (preprint)

Introduction Respiratory Syncytial Virus (RSV) is a major cause of acute lower respiratory tract infections (ALRI) in infants. There are no licensed vaccines and only one monoclonal antibody available to protect infants from disease. A new and potentially longer-lasting monoclonal antibody, Nirsevimab, showed promising results in phase IIb/III trials. We evaluate the cost-effectiveness of Nirsevimab intervention programmes in England and Wales Methods We used a dynamic model for RSV transmission, calibrated to data from England and Wales. We considered a suite of potential Nirsevimab programmes, including administration to all neonates (year-round); only neonates born during the RSV season (seasonal); or neonates born during the RSV season plus infants less than six months old before the start of the RSV season (seasonal + catch-up). Results If administered seasonally to all infants at birth, we found that Nirsevimab would have to be priced less than £63 per dose for at least 50% certainty that it could cost-effectively replace the current Palivizumab programme, using an ICER threshold of £20,000/QALY. An extended seasonal programme which includes a pre-season catch-up becomes the optimal strategy below a £34/dose purchasing price for at least 50% certainty. At a purchasing price per dose of £5-32, the annual implementation costs of a seasonal programme could be as high as £2 million before a switch to a year-round strategy would be optimal. Discussion Nirsevimab has the potential to be cost-effective in England and Wales not only for use in high-risk infants.

Influenza vaccine uptake among children and older adults in China: A secondary analysis of a quasi-experimental study (preprint)

Background: Influenza vaccination is the key to reducing the influenza-related disease burden, especially among high-risk populations. However, influenza vaccine uptake is low in China. This secondary analysis of a quasi-experimental trial in Guangdong Province aimed to understand factors associated with influenza vaccine uptake among children and older people stratified by funding context. Methods: : A total of 225 children (aged 0.5-8 years) and 225 older people (aged 60 years or above) were recruited from three clinics (rural, suburban, urban). Participants were allocated into two groups based on different funding contexts: self-paid group (N=150, including 75 children and 75 older adults) in which participants paid full market price for their vaccination; and subsidized group (N=300, including 150 children and 150 older adults) in which varying levels of financial support was provided. Univariable and multivariable logistic regressions were conducted stratified by funding contexts. Results: : Overall, 75.0% (225/300) of participants in the subsidized group and 36.7% (55/150) in the self-paid group got vaccinated. Older adults had lower vaccination rates than children in both funding groups, while both age groups showed much higher uptake in the subsidized group than in the self-paid group (86.7% vs 53.3% among children; 63.3% vs 20.0% among older people). In the self-paid group, participants living with children (aOR:2.61, 95%CI: 1.06-6.42) or older people (aOR:4.76, 95%CI: 1.08-20.90) having prior influenza vaccination in the same household were more likely to be vaccinated; trust in providers’ advice (aOR=4.95, 95%CI:1.99, 12.43) or effectiveness of the vaccine (aOR: 12.18, 95%CI: 5.21-28.50), and experienced influenza-like illnesses in the family (aOR=46.52, 4.10, 533.78) were associated with higher vaccine uptake in the subsidized group. Conclusions: Older people had suboptimal vaccine uptake compared to children in both contexts and need more attention in future efforts to enhance influenza vaccination. Tailoring interventions to different vaccine funding contexts may help improve influenza vaccine uptake: In self-paid context, measures to motivate people to accept their first ever influenza vaccination may be a promising strategy. In subsidized context, strategies to improve public confidence in vaccine effectiveness and providers’ advice would be useful. Trial registration: ChiCTR2000040048. Registered on November 19, 2020.

Estimating RSV seasonality from pandemic disruptions: a modelling study (preprint)

Background: Respiratory syncytial virus (RSV) is a leading cause of respiratory tract infections and bronchiolitis in young children. The seasonal pattern of RSV is shaped by short-lived immunity, seasonally varying contact rates and pathogen viability. The magnitude of each of these parameters is not fully clear. The disruption of the regular seasonality of RSV during the COVID pandemic in 2020 due to control measures, and the ensuing delayed surge in RSV cases provides an opportunity to disentangle these factors and to understand the implication for vaccination strategies. A better understanding of the drivers of RSV seasonality is key for developing future vaccination strategies. Methods: We developed a mathematical model of RSV transmission, which simulates the sequential re-infection (SEIRRS4) and uses a flexible Von Mises function to model the seasonal forcing. Using MCMC we fit the model to laboratory confirmed RSV data from 2010-2022 from NSW while accounting for the reduced contact rates during the pandemic with Google mobility data. We estimated the baseline transmission rate, its amplitude and shape during RSV season as well as the duration of immunity. The resulting parameter estimates were compared to a fit to pre-pandemic data only, and to a fit with a cosine forcing function. We then simulated the expected shifts in peak timing and amplitude under two vaccination strategies: continuous and seasonal vaccination. Results: We estimate that RSV dynamics in NSW can be best explained by a high effective baseline transmission rate (2.94/d, 95% CrI 2.72-3.19) and a narrow peak with a maximum 13% increase compared to the baseline transmission rate. We also estimate the duration of post infection temporary but sterilizing immunity to be 412 days (95% CrI 391-434). A cosine forcing resulted in a similar fit and posterior estimates. Excluding the data from the pandemic period in the fit increased parameter correlation and yielded less informative posterior distributions. The continuous vaccination strategy led to more extreme seasonal incidence with a delay in the peak timing and a higher amplitude whereas seasonal vaccination flattened the incidence curves. Conclusion: Quantifying the parameters that govern RSV seasonality is key in determining potential indirect effects from immunization strategies as those are being rolled out in the next few years.

Assessing the impacts of timing on the health benefits, cost-effectiveness and relative affordability of COVID-19 vaccination programmes in 27 African Countries (preprint)

Background: The COVID-19 vaccine supply shortage in 2021 constrained rollout efforts in Africa while populations experienced waves of epidemics. As supply picks up, a key question becomes if vaccination remains an impactful and cost-effective strategy given changes in the timing of implementation. Methods: We assessed the impact of timing using an epidemiological and economic model. We fitted our mathematical epidemiological model to reported COVID-19 deaths in 27 African countries to estimate the existing immunity (resulting from infection) before substantial vaccine rollout. We then projected health outcomes for different programme start dates (2021-01-01 to 2021-12-01, n = 12) and roll-out rates (slow, medium, fast; 275, 826, and 2066 doses/ million population-day, respectively) for viral vector and mRNA vaccines. Rollout rates used were derived from observed uptake trajectories. We collected data on vaccine delivery costs by country income group. Lastly, we calculated incremental cost-effectiveness ratios and relative affordability. Findings: Vaccination programmes with early start dates incur the most health benefits and are most cost-effective. While incurring the most health benefits, fast vaccine roll-outs are not always the most cost-effective. At a willingness-to-pay threshold of 0.5xGDP per capita, vaccine programmes starting in August 2021 using mRNA and viral vector vaccines were cost-effective in 6-10 and 17-18 of 27 countries, respectively. Interpretation: African countries with large proportions of their populations unvaccinated by late 2021 may find vaccination programmes less cost-effective than they could have been earlier in 2021. Lower vaccine purchasing costs and/or the emergence of new variants may improve cost-effectiveness. Funding: Bill and Melinda Gates Foundation, World Health Organization, National Institute of Health Research (UK), Health Data Research (UK)

The impact of alternative delivery strategies for novel tuberculosis vaccines in low- and middle-income countries: a modelling study (preprint)

Background Tuberculosis is a leading infectious cause of death worldwide. Novel vaccines will be required to reach global targets and reverse setbacks from the COVID-19 pandemic. We estimated the impact of novel tuberculosis vaccines in low- and middle-income countries (LMICs), under alternative delivery scenarios. Methods We calibrated a tuberculosis model to 105 LMICs (93% of global incidence). Vaccine scenarios were implemented as Basecase routine vaccination of 9-year-olds and a one-time vaccination campaign for ages ≥10 with country-specific introduction between 2028–2047 and 5-year scale-up to target coverage; Accelerated Scale-up as Basecase , but all countries introducing in 2025 with instant scale-up; and Routine Only as Basecase , but routine vaccination only. Vaccines protected against disease for 10-years, with 50% efficacy. Findings The Basecase scenario reduced tuberculosis incidence (19.5% [95% uncertainty range=18.3– 21.6%]) and mortality (20.6% [19.2–23.4%]) rates in 2050 and prevented 3.6 (3.3–3.9) million deaths before 2050, including 1.6 million in the WHO South-East Asian region. The Accelerated Scale-up scenario reduced tuberculosis incidence (25.2% [23.9–27.5%]), mortality (26.7% [25.2–29.9%]), and prevented 7.9 (7.3–8.5) million deaths. The Routine Only scenario reduced tuberculosis incidence (9.9% [9.0–11.6%]), mortality (9.9% [8.9–12.3%]), and prevented 1.1 (0.9–1.2) million deaths. Interpretations Novel tuberculosis vaccines could have substantial impact, which will vary depending on delivery strategy. Including a campaign will be crucial for rapid impact. Accelerated introduction, similar to the pace of COVID-19 vaccines, could approximately double the lives saved before 2050. Investment is required to support vaccine development, manufacturing, prompt introduction and scale-up. Funding WHO (2020/985800-0)

COVID-19 impact on routine immunisations for vaccine-preventable diseases: Projecting the effect of different routes to recovery (preprint)

Introduction: Over the past two decades, vaccination programmes for vaccine-preventable diseases (VPDs) have expanded across low- and middle-income countries (LMICs). However, the rise of COVID-19 resulted in global disruption to routine immunisation (RI) activities. Such disruptions could have a detrimental effect on public health, leading to more deaths from VPDs, particularly without mitigation efforts. Hence, as RIs resume, it is important to estimate the effectiveness of different approaches for recovery. Methods: We apply an impact extrapolation method developed by the Vaccine Impact Modelling Consortium to estimate the impact of COVID-19-related disruptions with different recovery scenarios for ten VPDs across 112 LMICs. We focus on deaths averted due to RIs occurring in the years 2020- 2030 and investigate two recovery scenarios relative to a no-COVID-19 scenario. In the recovery scenarios, we assume a 10% COVID-19-related drop in RI coverage in the year 2020. We then linearly interpolate coverage to the year 2030 to investigate two routes to recovery, whereby the immunization agenda (IA2030) targets are reached by 2030 or fall short by 10%. Results: We estimate that falling short of the IA2030 targets by 10% leads to 11.26% fewer fully vaccinated persons (FVPs) and 11.34% more deaths over the years 2020-2030 relative to the noCOVID-19 scenario, whereas, reaching the IA2030 targets reduces these proportions to 5% fewer FVPs and 5.22% more deaths. The impact of the disruption varies across the VPDs with diseases where coverage expands drastically in future years facing a smaller detrimental effect. Conclusion: Overall, our results show that drops in RI coverage could result in more deaths due to VPDs. As the impact of COVID-19-related disruptions is dependent on the vaccination coverage that is achieved over the coming years, the continued efforts of building up coverage and addressing gaps in immunity are vital in the road to recovery.

Projected epidemiological consequences of the Omicron SARS-CoV-2 variant in England, December 2021 to April 2022 (preprint)

The Omicron B.1.1.529 SARS-CoV-2 variant was first detected in late November 2021 and has since spread to multiple countries worldwide. We model the potential consequences of the Omicron variant on SARS-CoV-2 transmission and health outcomes in England between December 2021 and April 2022, using a deterministic compartmental model fitted to epidemiological data from March 2020 onwards. Because of uncertainty around the characteristics of Omicron, we explore scenarios varying the extent of Omicron's immune escape and the effectiveness of COVID-19 booster vaccinations against Omicron, assuming the level of Omicron's transmissibility relative to Delta to match the growth in observed S gene target failure data in England. We consider strategies for the re-introduction of control measures in response to projected surges in transmission, as well as scenarios varying the uptake and speed of COVID-19 booster vaccinations and the rate of Omicron's introduction into the population. These results suggest that Omicron has the potential to cause substantial surges in cases, hospital admissions and deaths in populations with high levels of immunity, including England. The reintroduction of additional non-pharmaceutical interventions may be required to prevent hospital admissions exceeding the levels seen in England during the previous peak in winter 2020-2021.

Determinants of RSV epidemiology following suppression through pandemic contact restrictions (preprint)

Introduction COVID-19 related non-pharmaceutical interventions (NPIs) led to a suppression of RSV circulation in winter 2020/21 throughout Europe and an off-season resurgence in Summer 2021 in several European countries. We explore how such temporary interruption may shape future RSV epidemiology and what factors drive the associated uncertainty. Methods We developed an age-structured dynamic transmission model to simulate pre-pandemic RSV infections and hospitalisations. We sampled parameters governing RSV seasonality, immunity acquisition and duration of post-infection immunity and retained those simulations that qualitatively fit the UK’s pre-pandemic epidemiology. From Spring 2020 to Summer 2021 we assumed a 50% reduced contact frequency, returning to pre-pandemic levels from mid-May 2021. We simulated transmission forwards until 2023 and evaluated the impact of the sampled parameters on the projected trajectories of RSV hospitalisations. Results Following a lifting of contact restrictions in summer 2021 the model replicated an out-of-season resurgence of RSV. If unmitigated, paediatric RSV hospitalisation incidence in the 2021/22 season was projected to increase by 32% to 67% compared to pre-pandemic levels. The size of the increase depended most on whether infection risk was primarily determined by immunity acquired from previous exposure or general immune maturation. While infants were less affected, the increase in seasonal hospitalisation incidence exceeded 100% in 1-2 year old children and 275% in 2-5 year old children, respectively, in some simulations where immunity from previous exposure dominated. Consequently, the average age of a case increased by 1 to 5 months, most markedly if there was strong immunity acquisition from previous exposure. If immunity to infection was largely determined by age rather than previous exposure, the 2021/22 season started earlier and lasted longer but with a peak incidence lower or similar to pre-pandemic levels. For subsequent seasons, simulations suggested a quick return to pre-pandemic epidemiology, with some slight oscillating behaviour possible depending on the strength of post-exposure immunity. Conclusion COVID-19 mitigation measures stopped RSV circulation in the 2020/21 season and generated immunity debt that will likely lead to a temporary increase in RSV burden in the season following the lifting of restrictions, particularly in 1 to 5 year old children. A more accurate understanding of immunity drivers for RSV is needed to better predict the size of such an increase and plan a potential expansion of pharmaceutical and non-pharmaceutical mitigation measures.

Comparing human and model-based forecasts of COVID-19 in Germany and Poland (preprint)

Forecasts based on epidemiological modelling have played an important role in shaping public policy throughout the COVID-19 pandemic. This modelling combines knowledge about infectious disease dynamics with the subjective opinion of the researcher who develops and refines the model and often also adjusts model outputs. Developing a forecast model is difficult, resource- and time-consuming. It is therefore worth asking what modelling is able to add beyond the subjective opinion of the researcher alone. To investigate this, we analysed different real-time forecasts of cases of and deaths from COVID-19 in Germany and Poland over a 1-4 week horizon submitted to the German and Polish Forecast Hub. We compared crowd forecasts elicited from researchers and volunteers, against a) forecasts from two semi-mechanistic models based on common epidemiological assumptions and b) the ensemble of all other models submitted to the Forecast Hub. We found crowd forecasts, despite being overconfident, to outperform all other methods across all forecast horizons when forecasting cases (weighted interval score relative to the Hub ensemble 2 weeks ahead: 0.89). Forecasts based on computational models performed comparably better when predicting deaths (rel. WIS 1.26), suggesting that epidemiological modelling and human judgement can complement each other in important ways.

Dosing interval strategies for two-dose COVID-19 vaccination in 13 low- and middle-income countries of Europe: health impact modelling and benefit-risk analysis (preprint)

Background: In settings where the COVID-19 vaccine supply is constrained, extending the intervals between the first and second doses of the COVID-19 vaccine could let more people receive their first doses earlier. Our aim is to estimate the health impact of COVID-19 vaccination alongside benefit-risk assessment of different dosing intervals for low- and middle-income countries of Europe. Methods: We fitted a dynamic transmission model to country-level daily reported COVID-19 mortality in 13 low- and middle-income countries in the World Health Organization European Region (Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Georgia, Republic of Moldova, Russian Federation, Serbia, North Macedonia, Turkey, and Ukraine). A vaccine product with characteristics similar to the Oxford/AstraZeneca COVID-19 (AZD1222) vaccine was used in the base case scenario and was complemented by sensitivity analyses around efficacies related to other COVID-19 vaccines. Both fixed dosing intervals at 4, 8, 12, 16, and 20 weeks and dose-specific intervals that prioritise specific doses for certain age groups were tested. Optimal intervals minimise COVID-19 mortality between March 2021 and December 2022. We incorporated the emergence of variants of concern into the model, and also conducted a benefit-risk assessment to quantify the trade-off between health benefits versus adverse events following immunisation. Findings: In 12 of the 13 countries, optimal strategies are those that prioritise the first doses among older adults (60+ years) or adults (20-59 years). These strategies lead to dosing intervals longer than six months. In comparison, a four-week fixed dosing interval may incur 10.2% [range: 4.0% - 22.5%; n = 13 (countries)] more deaths. There is generally a negative association between dosing interval and COVID-19 mortality within the range we investigated. Assuming a shorter first dose waning duration of 120 days, as opposed to 360 days in the base case, led to shorter optimal dosing intervals of 8-12 weeks. Benefit-risk ratios were the highest for fixed dosing intervals of 8-12 weeks. Interpretation: We infer that longer dosing intervals of over six months, which are substantially longer than the current label recommendation for most vaccine products, could reduce COVID-19 mortality in low- and middle-income countries of WHO/Europe. Certain vaccine features, such as fast waning of first doses, significantly shorten the optimal dosing intervals.

Behaviour, booster vaccines and waning vaccine protection: modelling the medium-term dynamics of SARS-CoV-2 transmission in England (preprint)

England has experienced a heavy burden of COVID-19, with high infection levels observed throughout the summer months of 2021. Alongside the emergence of evidence suggesting that COVID-19 vaccine protection wanes over time, booster vaccinations began for individuals aged 50 and above in September 2021. Using a model fitted to 18 months of epidemiological data, we project potential dynamics of SARS-CoV-2 transmission in England to September 2022. We consider key uncertainties including behavioural change, waning vaccine protection, strategies for vaccination, and the reintroduction of public health and social measures. We project the current wave of transmission will peak in Autumn 2021, with low levels of transmission in early 2022. The extent to which SARS-CoV-2 transmission resurges in 2022 depends largely on assumptions around waning vaccine protection and booster vaccinations. Widespread booster vaccinations or the reimposition of mild public health and social measures such as work-from-home policies could largely mitigate the wave of COVID-19 transmission projected to occur in England in Spring/Summer 2022.