Predicting subnational incidence of COVID-19 cases and deaths in EU countries.

BACKGROUND: Recurring COVID-19 waves highlight the need for tools able to quantify transmission risk, and identify geographical areas at risk of outbreaks. Local outbreak risk depends on complex immunity patterns resulting from previous infections, vaccination, waning and immune escape, alongside other factors (population density, social contact patterns). Immunity patterns are spatially and demographically heterogeneous, and are challenging to capture in country-level forecast models. METHODS: We used a spatiotemporal regression model to forecast subnational case and death counts and applied it to three EU countries as test cases: France, Czechia, and Italy. Cases in local regions arise from importations or local transmission. Our model produces age-stratified forecasts given age-stratified data, and links reported case counts to routinely collected covariates (e.g. test number, vaccine coverage). We assessed the predictive performance of our model up to four weeks ahead using proper scoring rules and compared it to the European COVID-19 Forecast Hub ensemble model. Using simulations, we evaluated the impact of variations in transmission on the forecasts. We developed an open-source RShiny App to visualise the forecasts and scenarios. RESULTS: At a national level, the median relative difference between our median weekly case forecasts and the data up to four weeks ahead was 25% (IQR: 12-50%) over the prediction period. The accuracy decreased as the forecast horizon increased (on average 24% increase in the median ranked probability score per added week), while the accuracy of death forecasts was more stable. Beyond two weeks, the model generated a narrow range of likely transmission dynamics. The median national case forecasts showed similar accuracy to forecasts from the European COVID-19 Forecast Hub ensemble model, but the prediction interval was narrower in our model. Generating forecasts under alternative transmission scenarios was therefore key to capturing the range of possible short-term transmission dynamics. DISCUSSION: Our model captures changes in local COVID-19 outbreak dynamics, and enables quantification of short-term transmission risk at a subnational level. The outputs of the model improve our ability to identify areas where outbreaks are most likely, and are available to a wide range of public health professionals through the Shiny App we developed.

What are economic costs and when should they be used in health economic studies?

Economic analyses of healthcare interventions are an important consideration in evidence-based policymaking. A key component of such analyses is the costs of interventions, for which most are familiar with using budgets and expenditures. However, economic theory states that the true value of a good/service is the value of the next best alternative forgone as a result of using the resource and therefore observed prices or charges do not necessarily reflect the true economic value of resources. To address this, economic costs are a fundamental concept within (health) economics. Crucially, they are intended to reflect the resources' opportunity costs (the forgone opportunity to use those resources for another purpose) and they are based on the value of the resource's next-best alternative use that has been forgone. This is a broader conceptualization of a resource's value than its financial cost and recognizes that resources can have a value that may not be fully captured by their market price and that by using a resource it makes it unavailable for productive use elsewhere. Importantly, economic costs are preferred over financial costs for any health economic analyses aimed at informing decisions regarding the optimum allocation of the limited/competing resources available for healthcare (such as health economic evaluations), and they are also important when considering the replicability and sustainability of healthcare interventions. However, despite this, economic costs and the reasons why they are used is an area that can be misunderstood by professionals without an economic background. In this paper, we outline to a broader audience the principles behind economic costs and when and why they should be used within health economic analyses. We highlight that the difference between financial and economic costs and what adjustments are needed within cost calculations will be influenced by the context of the study, the perspective, and the objective.

Seasonality of acute kidney injury phenotypes in England: an unsupervised machine learning classification study of electronic health records.

BACKGROUND: Acute Kidney Injury (AKI) is a multifactorial condition which presents a substantial burden to healthcare systems. There is limited evidence on whether it is seasonal. We sought to investigate the seasonality of AKI hospitalisations in England and use unsupervised machine learning to explore clustering of underlying comorbidities, to gain insights for future intervention. METHODS: We used Hospital Episodes Statistics linked to the Clinical Practice Research Datalink to describe the overall incidence of AKI admissions between 2015 and 2019 weekly by demographic and admission characteristics. We carried out dimension reduction on 850 diagnosis codes using multiple correspondence analysis and applied k-means clustering to classify patients. We phenotype each group based on the dominant characteristics and describe the seasonality of AKI admissions by these different phenotypes. RESULTS: Between 2015 and 2019, weekly AKI admissions peaked in winter, with additional summer peaks related to periods of extreme heat. Winter seasonality was more evident in those diagnosed with AKI on admission. From the cluster classification we describe six phenotypes of people admitted to hospital with AKI. Among these, seasonality of AKI admissions was observed among people who we described as having a multimorbid phenotype, established risk factor phenotype, and general AKI phenotype. CONCLUSION: We demonstrate winter seasonality of AKI admissions in England, particularly among those with AKI diagnosed on admission, suggestive of community triggers. Differences in seasonality between phenotypes suggests some groups may be more likely to develop AKI as a result of these factors. This may be driven by underlying comorbidity profiles or reflect differences in uptake of seasonal interventions such as vaccines.

The influence of COVID-19 risk perception and vaccination status on the number of social contacts across Europe: insights from the CoMix study.

BACKGROUND: The SARS-CoV-2 transmission dynamics have been greatly modulated by human contact behaviour. To curb the spread of the virus, global efforts focused on implementing both Non-Pharmaceutical Interventions (NPIs) and pharmaceutical interventions such as vaccination. This study was conducted to explore the influence of COVID-19 vaccination status and risk perceptions related to SARS-CoV-2 on the number of social contacts of individuals in 16 European countries. METHODS: We used data from longitudinal surveys conducted in the 16 European countries to measure social contact behaviour in the course of the pandemic. The data consisted of representative panels of participants in terms of gender, age and region of residence in each country. The surveys were conducted in several rounds between December 2020 and September 2021 and comprised of 29,292 participants providing a total of 111,103 completed surveys. We employed a multilevel generalized linear mixed effects model to explore the influence of risk perceptions and COVID-19 vaccination status on the number of social contacts of individuals. RESULTS: The results indicated that perceived severity played a significant role in social contact behaviour during the pandemic after controlling for other variables (p-value < 0.001). More specifically, participants who had low or neutral levels of perceived severity reported 1.25 (95% Confidence intervals (CI) 1.13 - 1.37) and 1.10 (95% CI 1.00 - 1.21) times more contacts compared to those who perceived COVID-19 to be a serious illness, respectively. Additionally, vaccination status was also a significant predictor of contacts (p-value < 0.001), with vaccinated individuals reporting 1.31 (95% CI 1.23 - 1.39) times higher number of contacts than the non-vaccinated. Furthermore, individual-level factors played a more substantial role in influencing contact behaviour than country-level factors. CONCLUSION: Our multi-country study yields significant insights on the importance of risk perceptions and vaccination in behavioral changes during a pandemic emergency. The apparent increase in social contact behaviour following vaccination would require urgent intervention in the event of emergence of an immune escaping variant.

Long-Term Health-Related Quality of Life in Non-Hospitalized Coronavirus Disease 2019 (COVID-19) Cases With Confirmed Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection in England: Longitudinal Analysis and Cross-Sectional Comparison With Controls.

BACKGROUND: We aimed to quantify the unknown losses in health-related quality of life of coronavirus disease 2019 (COVID-19) cases using quality-adjusted lifedays (QALDs) and the recommended EQ-5D instrument in England. METHODS: Prospective cohort study of nonhospitalized, polymerase chain reaction (PCR)-confirmed severe acute respiratory syndrome coronavirus 2-positive (SARS-CoV-2-positive) cases aged 12-85 years and followed up for 6 months from 1 December 2020, with cross-sectional comparison to SARS-CoV-2-negative controls. Main outcomes were QALD losses; physical symptoms; and COVID-19-related private expenditures. We analyzed results using multivariable regressions with post hoc weighting by age and sex, and conditional logistic regressions for the association of each symptom and EQ-5D limitation on cases and controls. RESULTS: Of 548 cases (mean age 41.1 years; 61.5% female), 16.8% reported physical symptoms at month 6 (most frequently extreme tiredness, headache, loss of taste and/or smell, and shortness of breath). Cases reported more limitations with doing usual activities than controls. Almost half of cases spent a mean of £18.1 on nonprescription drugs (median: £10.0), and 52.7% missed work or school for a mean of 12 days (median: 10). On average, all cases lost 13.7 (95% confidence interval [CI]: 9.7, 17.7) QALDs, whereas those reporting symptoms at month 6 lost 32.9 (95% CI: 24.5, 37.6) QALDs. Losses also increased with older age. Cumulatively, the health loss from morbidity contributes at least 18% of the total COVID-19-related disease burden in the England. CONCLUSIONS: One in 6 cases report ongoing symptoms at 6 months, and 10% report prolonged loss of function compared to pre-COVID-19 baselines. A marked health burden was observed among older COVID-19 cases and those with persistent physical symptoms.

The impact of COVID-19 vaccination in prisons in England and Wales: a metapopulation model.

BACKGROUND: High incidence of cases and deaths due to coronavirus disease 2019 (COVID-19) have been reported in prisons worldwide. This study aimed to evaluate the impact of different COVID-19 vaccination strategies in epidemiologically semi-enclosed settings such as prisons, where staff interact regularly with those incarcerated and the wider community. METHODS: We used a metapopulation transmission-dynamic model of a local prison in England and Wales. Two-dose vaccination strategies included no vaccination, vaccination of all individuals who are incarcerated and/or staff, and an age-based approach. Outcomes were quantified in terms of COVID-19-related symptomatic cases, losses in quality-adjusted life-years (QALYs), and deaths. RESULTS: Compared to no vaccination, vaccinating all people living and working in prison reduced cases, QALY loss and deaths over a one-year period by 41%, 32% and 36% respectively. However, if vaccine introduction was delayed until the start of an outbreak, the impact was negligible. Vaccinating individuals who are incarcerated and staff over 50 years old averted one death for every 104 vaccination courses administered. All-staff-only strategies reduced cases by up to 5%. Increasing coverage from 30 to 90% among those who are incarcerated reduced cases by around 30 percentage points. CONCLUSIONS: The impact of vaccination in prison settings was highly dependent on early and rapid vaccine delivery. If administered to both those living and working in prison prior to an outbreak occurring, vaccines could substantially reduce COVID-19-related morbidity and mortality in prison settings.

Public health considerations for transitioning beyond the acute phase of the COVID-19 pandemic in the EU/EEA.

Many countries, including some within the EU/EEA, are in the process of transitioning from the acute pandemic phase. During this transition, it is crucial that countries' strategies and activities remain guided by clear COVID-19 control objectives, which increasingly will focus on preventing and managing severe outcomes. Therefore, attention must be given to the groups that are particularly vulnerable to severe outcomes of SARS-CoV-2 infection, including individuals in congregate and healthcare settings. In this phase of pandemic management, a strong focus must remain on transitioning testing approaches and systems for targeted surveillance of COVID-19, capitalising on and strengthening existing systems for respiratory virus surveillance. Furthermore, it will be crucial to focus on lessons learned from the pandemic to enhance preparedness and to enact robust systems for the preparedness, detection, rapid investigation and assessment of new and emerging SARS-CoV-2 variants. Filling existing knowledge gaps, including behavioural insights, can help guide the response to future resurgences of SARS-CoV-2 and/or the emergence of other pandemics. Finally, 'vaccine agility' will be needed to respond to changes in people's behaviours, changes in the virus, and changes in population immunity, all the while addressing issues of global health equity.

How to Prevent Vaccines Falling Victim to Their Own Success: Intertemporal Dependency of Incidence Levels on Indirect Effects in Economic Reevaluations.

OBJECTIVES: Incremental cost-effectiveness analyses may inform the optimal choice of healthcare interventions. Nevertheless, for many vaccines, benefits fluctuate with incidence levels over time. Reevaluating a vaccine after it has successfully decreased incidences may eventually cause a disease resurgence if switching to a vaccine with lower indirect benefits. Decisions may successively alternate between vaccines alongside repeated rises and falls in incidence and when indirect effects from historic use are ignored. Our suggested proposal aims to prevent suboptimal decision making. METHODS: We used a conceptual model of demand to illustrate alternating decisions between vaccines because of time-varying levels of indirect effects. Similar to the concept of subsidies, we propose internalizing the indirect effects achievable with vaccines. In a case study over 60 years, we simulated a hypothetical 10-year reevaluation of 2 oncogenic human papillomavirus vaccines, of which only 1 protects additionally against anogenital warts. RESULTS: Our case study showed that the vaccine with additional warts protection is initially valued higher than the vaccine without additional warts protection. After 10 years, this differential decreases because of declines in warts incidence, which supports switching to the nonwarts vaccine that causes a warts resurgence eventually. Instead, pricing the indirect effects separately supports continuing with the warts vaccine. CONCLUSIONS: Ignoring how the observed incidences depend on the indirect effects achieved with a particular vaccine may lead to repeated changes in vaccines at successive reevaluations, with unintended resurgences, economic inefficiencies, and eroding vaccine confidence. We propose internalizing indirect effects to prevent vaccines falling victim to their own success.

Optimizing Benefits of Testing Key Workers for Infection with SARS-CoV-2: A Mathematical Modeling Analysis.

BACKGROUND: Internationally, key workers such as healthcare staff are advised to stay at home if they or household members experience coronavirus disease 2019 (COVID-19)-like symptoms. This potentially isolates/quarantines many staff without SARS-CoV-2, while not preventing transmission from staff with asymptomatic infection. We explored the impact of testing staff on absence durations from work and transmission risks to others. METHODS: We used a decision-analytic model for 1000 key workers to compare the baseline strategy of (S0) no RT-PCR testing of workers to testing workers (S1) with COVID-19-like symptoms in isolation, (S2) without COVID-19-like symptoms but in household quarantine, and (S3) all staff. We explored confirmatory re-testing scenarios of repeating all initial tests, initially positive tests, initially negative tests, or no re-testing. We varied all parameters, including the infection rate (0.1-20%), proportion asymptomatic (10-80%), sensitivity (60-95%), and specificity (90-100%). RESULTS: Testing all staff (S3) changes the risk of workplace transmission by -56.9 to +1.0 workers/1000 tests (with reductions throughout at RT-PCR sensitivity ≥65%), and absences by -0.5 to +3.6 days/test but at heightened testing needs of 989.6-1995.9 tests/1000 workers. Testing workers in household quarantine (S2) reduces absences the most by 3.0-6.9 days/test (at 47.0-210.4 tests/1000 workers), while increasing risk of workplace transmission by 0.02-49.5 infected workers/1000 tests (which can be minimized when re-testing initially negative tests). CONCLUSIONS: Based on optimizing absence durations or transmission risk, our modeling suggests testing staff in household quarantine or all staff, depending on infection levels and testing capacities.

Infant Hospitalizations and Fatalities Averted by the Maternal Pertussis Vaccination Program in England, 2012-2017: Post-implementation Economic Evaluation.

In October 2012, a maternal pertussis vaccination program was implemented in England following an increased incidence and mortality in infants. We evaluated the cost-effectiveness of the program by comparing pertussis-related infant hospitalizations and deaths in 2012-2017 with nonvaccination scenarios. Despite considerable uncertainties, findings support the cost-effectiveness of the program.

Quantifying the economic cost of antibiotic resistance and the impact of related interventions: rapid methodological review, conceptual framework and recommendations for future studies.

BACKGROUND: Antibiotic resistance (ABR) poses a major threat to health and economic wellbeing worldwide. Reducing ABR will require government interventions to incentivise antibiotic development, prudent antibiotic use, infection control and deployment of partial substitutes such as rapid diagnostics and vaccines. The scale of such interventions needs to be calibrated to accurate and comprehensive estimates of the economic cost of ABR. METHODS: A conceptual framework for estimating costs attributable to ABR was developed based on previous literature highlighting methodological shortcomings in the field and additional deductive epidemiological and economic reasoning. The framework was supplemented by a rapid methodological review. RESULTS: The review identified 110 articles quantifying ABR costs. Most were based in high-income countries only (91/110), set in hospitals (95/110), used a healthcare provider or payer perspective (97/110), and used matched cohort approaches to compare costs of patients with antibiotic-resistant infections and antibiotic-susceptible infections (or no infection) (87/110). Better use of methods to correct biases and confounding when making this comparison is needed. Findings also need to be extended beyond their limitations in (1) time (projecting present costs into the future), (2) perspective (from the healthcare sector to entire societies and economies), (3) scope (from individuals to communities and ecosystems), and (4) space (from single sites to countries and the world). Analyses of the impact of interventions need to be extended to examine the impact of the intervention on ABR, rather than considering ABR as an exogeneous factor. CONCLUSIONS: Quantifying the economic cost of resistance will require greater rigour and innovation in the use of existing methods to design studies that accurately collect relevant outcomes and further research into new techniques for capturing broader economic outcomes.

Cost-Effectiveness Analysis of Screening for Persistent Hepatitis E Virus Infection in Solid Organ Transplant Patients in the United Kingdom: A Model-Based Economic Evaluation.

BACKGROUND: Despite potentially severe and fatal outcomes, recent studies of solid organ transplant (SOT) recipients in Europe suggest that hepatitis E virus (HEV) infection is underdiagnosed, with a prevalence of active infection of up to 4.4%. OBJECTIVES: To determine the cost-effectiveness of introducing routine screening for HEV infection in SOT recipients in the UK. METHODS: A Markov cohort model was developed to evaluate the cost-utility of 4 HEV screening options over the lifetime of 1000 SOT recipients. The current baseline of nonsystematic testing was compared with annual screening of all patients by polymerase chain reaction (PCR; strategy A) or HEV-antigen (HEV-Ag) detection (strategy B) and selective screening of patients who have a raised alanine aminotransferase (ALT) value by PCR (strategy C) or HEV-Ag (strategy D). The primary outcome was the incremental cost per quality-adjusted life-year (QALY). We adopted the National Health Service (NHS) perspective and discounted future costs and benefits at 3.5%. RESULTS: At a willingness-to-pay of £20 000/QALY gained, systematic screening of SOT patients by any method (strategy A-D) had a high probability (77.9%) of being cost-effective. Among screening strategies, strategy D is optimal and expected to be cost-saving to the NHS; if only PCR testing strategies are considered, then strategy C becomes cost-effective (£660/QALY). These findings were robust against a wide range of sensitivity and scenario analyses. CONCLUSIONS: Our model showed that routine screening for HEV in SOT patients is very likely to be cost-effective in the UK, particularly in patients presenting with an abnormal alanine aminotransferase.

Should NICE reconsider the 2016 UK guidelines on TB contact tracing? A cost-effectiveness analysis of contact investigations in London.

BACKGROUND: In January 2016, clinical TB guidance in the UK changed to no longer recommend screening contacts of non-pulmonary, non-laryngeal (ETB) index cases. However, no new evidence was cited for this change, and there is evidence that screening these contacts may be worthwhile. The objective of this study was to estimate the cost-effectiveness of screening contacts of adult ETB cases and adult pulmonary or laryngeal TB (PTB) cases in London, UK. METHODS: We carried out a cross-sectional analysis of data collected on TB index cases and contacts in the London TB register and an economic evaluation using a static model describing contact tracing outcomes. Incremental cost-effectiveness ratios (ICERs) were calculated using no screening as the baseline comparator. All adult TB cases (≥15 years old) in London from 2012 to 2015, and their contacts, were eligible (2465/5084 PTB and 2559/6090 ETB index cases were included). RESULTS: Assuming each contact with PTB infects one person/month, the ICER of screening contacts of ETB cases was £78 000/quality-adjusted life-years (QALY) (95% CI 39 000 to 140 000), and screening contacts of PTB cases was £30 000/QALY (95% CI 18 000 to 50 000). The ICER of screening contacts of ETB cases was £30 000/QALY if each contact with PTB infects 3.4 people/month. Limitations of this study include the use of self-reported symptomatic periods and lack of knowledge about onward transmission from PTB contacts. CONCLUSIONS: Screening contacts of ETB cases in London was almost certainly not cost-effective at any conventional willingness-to-pay threshold in England, supporting recent changes to National Institute for Health and Care Excellence national guidelines.

Estimating the Hospital Burden of Norovirus-Associated Gastroenteritis in England and Its Opportunity Costs for Nonadmitted Patients.

Background: Norovirus places a substantial burden on healthcare systems, arising from infected patients, disease outbreaks, beds kept unoccupied for infection control, and staff absences due to infection. In settings with high rates of bed occupancy, opportunity costs arise from patients who cannot be admitted due to beds being unavailable. With several treatments and vaccines against norovirus in development, quantifying the expected economic burden is timely. Methods: The number of inpatients with norovirus-associated gastroenteritis in England was modeled using infectious and noninfectious gastrointestinal Hospital Episode Statistics codes and laboratory reports of gastrointestinal pathogens collected at Public Health England. The excess length of stay from norovirus was estimated with a multistate model and local outbreak data. Unoccupied bed-days and staff absences were estimated from national outbreak surveillance. The burden was valued conventionally using accounting expenditures and wages, which we contrasted to the opportunity costs from forgone patients using a novel methodology. Results: Between July 2013 and June 2016, 17.7% (95% confidence interval [CI], 15.6%‒21.6%) of primary and 23.8% (95% CI, 20.6%‒29.9%) of secondary gastrointestinal diagnoses were norovirus attributable. Annually, the estimated median 290000 (interquartile range, 282000‒297000) occupied and unoccupied bed-days used for norovirus displaced 57800 patients. Conventional costs for the National Health Service reached £107.6 million; the economic burden approximated to £297.7 million and a loss of 6300 quality-adjusted life-years annually. Conclusions: In England, norovirus is now the second-largest contributor of the gastrointestinal hospital burden. With the projected impact being greater than previously estimated, improved capture of relevant opportunity costs seems imperative for diseases such as norovirus.

Estimating the opportunity costs of bed-days.

Opportunity costs of bed-days are fundamental to understanding the value of healthcare systems. They greatly influence burden of disease estimations and economic evaluations involving stays in healthcare facilities. However, different estimation techniques employ assumptions that differ crucially in whether to consider the value of the second-best alternative use forgone, of any available alternative use, or the value of the actually chosen alternative. Informed by economic theory, this paper provides a taxonomic framework of methodologies for estimating the opportunity costs of resources. This taxonomy is then applied to bed-days by classifying existing approaches accordingly. We highlight differences in valuation between approaches and the perspective adopted, and we use our framework to appraise the assumptions and biases underlying the standard approaches that have been widely adopted mostly unquestioned in the past, such as the conventional use of reference costs and administrative accounting data. Drawing on these findings, we present a novel approach for estimating the opportunity costs of bed-days in terms of health forgone for the second-best patient, but expressed monetarily. This alternative approach effectively re-connects to the concept of choice and explicitly considers net benefits. It is broadly applicable across settings and for other resources besides bed-days.

The Use of Evidence to Design an Essential Package of Health Services in Pakistan: A Review and Analysis of Prioritisation Decisions at Different Stages of the Appraisal Process.

BACKGROUND: Pakistan embarked on a process of designing an essential package of health services (EPHS) as a pathway towards universal health coverage (UHC). The EPHS design followed an evidence-informed deliberative process; evidence on 170 interventions was introduced along multiple stages of appraisal engaging different stakeholders tasked with prioritising interventions for inclusion. We report on the composition of the package at different stages, analyse trends of prioritised and deprioritised interventions and reflect on the trade-offs made. METHODS: Quantitative evidence on cost-effectiveness, budget impact, and avoidable burden of disease was presented to stakeholders in stages. We recorded which interventions were prioritised and deprioritised at each stage and carried out three analyses: (1) a review of total number of interventions prioritised at each stage, along with associated costs per capita and disability-adjusted life years (DALYs) averted, to understand changes in affordability and efficiency in the package, (2) an analysis of interventions broken down by decision criteria and intervention characteristics to analyse prioritisation trends across different stages, and (3) a description of the trajectory of interventions broken down by current coverage and cost-effectiveness. RESULTS: Value for money generally increased throughout the process, although not uniformly. Stakeholders largely prioritised interventions with low budget impact and those preventing a high burden of disease. Highly cost-effective interventions were also prioritised, but less consistently throughout the stages of the process. Interventions with high current coverage were overwhelmingly prioritised for inclusion. CONCLUSION: Evidence-informed deliberative processes can produce actionable and affordable health benefit packages. While cost-effective interventions are generally preferred, other factors play a role and limit efficiency.

Characterising information gains and losses when collecting multiple epidemic model outputs.

BACKGROUND: Collaborative comparisons and combinations of epidemic models are used as policy-relevant evidence during epidemic outbreaks. In the process of collecting multiple model projections, such collaborations may gain or lose relevant information. Typically, modellers contribute a probabilistic summary at each time-step. We compared this to directly collecting simulated trajectories. We aimed to explore information on key epidemic quantities; ensemble uncertainty; and performance against data, investigating potential to continuously gain information from a single cross-sectional collection of model results. METHODS: We compared projections from the European COVID-19 Scenario Modelling Hub. Five teams modelled incidence in Belgium, the Netherlands, and Spain. We compared July 2022 projections by incidence, peaks, and cumulative totals. We created a probabilistic ensemble drawn from all trajectories, and compared to ensembles from a median across each model's quantiles, or a linear opinion pool. We measured the predictive accuracy of individual trajectories against observations, using this in a weighted ensemble. We repeated this sequentially against increasing weeks of observed data. We evaluated these ensembles to reflect performance with varying observed data. RESULTS: By collecting modelled trajectories, we showed policy-relevant epidemic characteristics. Trajectories contained a right-skewed distribution well represented by an ensemble of trajectories or a linear opinion pool, but not models' quantile intervals. Ensembles weighted by performance typically retained the range of plausible incidence over time, and in some cases narrowed this by excluding some epidemic shapes. CONCLUSIONS: We observed several information gains from collecting modelled trajectories rather than quantile distributions, including potential for continuously updated information from a single model collection. The value of information gains and losses may vary with each collaborative effort's aims, depending on the needs of projection users. Understanding the differing information potential of methods to collect model projections can support the accuracy, sustainability, and communication of collaborative infectious disease modelling efforts.

Cost of the COVID-19 pandemic versus the cost-effectiveness of mitigation strategies in EU/UK/OECD: a systematic review.

OBJECTIVES: The economic burden of COVID-19 pandemic is substantial, with both direct and indirect costs playing a significant role. DESIGN: A systematic literature review was conducted to estimate the cost of the COVID-19 pandemic and the cost-effectiveness of pharmaceutical or non-pharmaceutical interventions. All cost data were adjusted to the 2021 Euro, and interventions compared with null. DATA SOURCES: Ovid MEDLINE and EMBASE were searched from January 2020 through 22 April 2021. ELIGIBILITY CRITERIA: Studies regarding COVID-19 outbreak or public health preparedness measures or interventions with outcome measures related to the direct and indirect costs for disease and preparedness and/or response in countries of the European Union (EU), the European Economic Area (EEA), the UK and the Organisation for Economic Co-operation and Development (OECD) of all relevant epidemiological designs which estimate cost within the selected time frame were considered eligible. DATA EXTRACTION AND SYNTHESIS: Studies were searched, screened and coded independently by two reviewers with high measure of inter-rater agreement. Data were extracted to a predefined data extraction sheet. The risk of bias was assessed using the Consensus on Health Economic Criteria checklist. RESULTS: We included data from 41 economic studies. Ten studies evaluated the cost of the COVID-19 pandemic, while 31 assessed the cost-benefit of public health surveillance, preparedness and response measures. Overall, the economic burden of the COVID-19 pandemic was found to be substantial. Community screening, bed provision policies, investing in personal-protective-equipment and vaccination strategies were cost-effective. Physical distancing measures were associated with health benefits; however, their cost-effectiveness was dependent on the duration, compliance and the phase of the epidemic in which it was implemented. CONCLUSIONS: COVID-19 pandemic is associated with substantial short-term and long-term economic costs to healthcare systems, payers and societies, while interventions including testing and screening policies, vaccination and physical distancing policies were identified as those presenting cost-effective options to deal with the pandemic, dependent on population vaccination and the Re at the stage of the pandemic.

Protocol for an OpenSAFELY cohort study collecting patient-reported outcome measures using the TPP Airmid smartphone application and linked big data to quantify the health and economic costs of long COVID (OpenPROMPT).

INTRODUCTION: The impact of long COVID on health-related quality of-life (HRQoL) and productivity is not currently known. It is important to understand who is worst affected by long COVID and the cost to the National Health Service (NHS) and society, so that strategies like booster vaccines can be prioritised to the right people. OpenPROMPT aims to understand the impact of long COVID on HRQoL in adults attending English primary care. METHODS AND ANALYSIS: We will ask people to participate in this cohort study through a smartphone app (Airmid), and completing a series of questionnaires held within the app. Questionnaires will ask about HRQoL, productivity and symptoms of long COVID. Participants will be asked to fill in the questionnaires once a month, for 90 days. Questionnaire responses will be linked, where possible, to participants' existing health records from primary care, secondary care, and COVID testing and vaccination data. Analysis will take place using the OpenSAFELY data platform and will estimate the impact of long COVID on HRQoL, productivity and cost to the NHS. ETHICS AND DISSEMINATION: The Proportionate Review Sub-Committee of the South Central-Berkshire B Research Ethics Committee has reviewed and approved the study and have agreed that we can ask people to take part (22/SC/0198). Our results will provide information to support long-term care, and make recommendations for prevention of long COVID in the future. TRIAL REGISTRATION NUMBER: NCT05552612.

Predictive performance of multi-model ensemble forecasts of COVID-19 across European nations.

Background: Short-term forecasts of infectious disease burden can contribute to situational awareness and aid capacity planning. Based on best practice in other fields and recent insights in infectious disease epidemiology, one can maximise the predictive performance of such forecasts if multiple models are combined into an ensemble. Here, we report on the performance of ensembles in predicting COVID-19 cases and deaths across Europe between 08 March 2021 and 07 March 2022. Methods: We used open-source tools to develop a public European COVID-19 Forecast Hub. We invited groups globally to contribute weekly forecasts for COVID-19 cases and deaths reported by a standardised source for 32 countries over the next 1-4 weeks. Teams submitted forecasts from March 2021 using standardised quantiles of the predictive distribution. Each week we created an ensemble forecast, where each predictive quantile was calculated as the equally-weighted average (initially the mean and then from 26th July the median) of all individual models' predictive quantiles. We measured the performance of each model using the relative Weighted Interval Score (WIS), comparing models' forecast accuracy relative to all other models. We retrospectively explored alternative methods for ensemble forecasts, including weighted averages based on models' past predictive performance. Results: Over 52 weeks, we collected forecasts from 48 unique models. We evaluated 29 models' forecast scores in comparison to the ensemble model. We found a weekly ensemble had a consistently strong performance across countries over time. Across all horizons and locations, the ensemble performed better on relative WIS than 83% of participating models' forecasts of incident cases (with a total N=886 predictions from 23 unique models), and 91% of participating models' forecasts of deaths (N=763 predictions from 20 models). Across a 1-4 week time horizon, ensemble performance declined with longer forecast periods when forecasting cases, but remained stable over 4 weeks for incident death forecasts. In every forecast across 32 countries, the ensemble outperformed most contributing models when forecasting either cases or deaths, frequently outperforming all of its individual component models. Among several choices of ensemble methods we found that the most influential and best choice was to use a median average of models instead of using the mean, regardless of methods of weighting component forecast models. Conclusions: Our results support the use of combining forecasts from individual models into an ensemble in order to improve predictive performance across epidemiological targets and populations during infectious disease epidemics. Our findings further suggest that median ensemble methods yield better predictive performance more than ones based on means. Our findings also highlight that forecast consumers should place more weight on incident death forecasts than incident case forecasts at forecast horizons greater than 2 weeks. Funding: AA, BH, BL, LWa, MMa, PP, SV funded by National Institutes of Health (NIH) Grant 1R01GM109718, NSF BIG DATA Grant IIS-1633028, NSF Grant No.: OAC-1916805, NSF Expeditions in Computing Grant CCF-1918656, CCF-1917819, NSF RAPID CNS-2028004, NSF RAPID OAC-2027541, US Centers for Disease Control and Prevention 75D30119C05935, a grant from Google, University of Virginia Strategic Investment Fund award number SIF160, Defense Threat Reduction Agency (DTRA) under Contract No. HDTRA1-19-D-0007, and respectively Virginia Dept of Health Grant VDH-21-501-0141, VDH-21-501-0143, VDH-21-501-0147, VDH-21-501-0145, VDH-21-501-0146, VDH-21-501-0142, VDH-21-501-0148. AF, AMa, GL funded by SMIGE - Modelli statistici inferenziali per governare l'epidemia, FISR 2020-Covid-19 I Fase, FISR2020IP-00156, Codice Progetto: PRJ-0695. AM, BK, FD, FR, JK, JN, JZ, KN, MG, MR, MS, RB funded by Ministry of Science and Higher Education of Poland with grant 28/WFSN/2021 to the University of Warsaw. BRe, CPe, JLAz funded by Ministerio de Sanidad/ISCIII. BT, PG funded by PERISCOPE European H2020 project, contract number 101016233. CP, DL, EA, MC, SA funded by European Commission - Directorate-General for Communications Networks, Content and Technology through the contract LC-01485746, and Ministerio de Ciencia, Innovacion y Universidades and FEDER, with the project PGC2018-095456-B-I00. DE., MGu funded by Spanish Ministry of Health / REACT-UE (FEDER). DO, GF, IMi, LC funded by Laboratory Directed Research and Development program of Los Alamos National Laboratory (LANL) under project number 20200700ER. DS, ELR, GG, NGR, NW, YW funded by National Institutes of General Medical Sciences (R35GM119582; the content is solely the responsibility of the authors and does not necessarily represent the official views of NIGMS or the National Institutes of Health). FB, FP funded by InPresa, Lombardy Region, Italy. HG, KS funded by European Centre for Disease Prevention and Control. IV funded by Agencia de Qualitat i Avaluacio Sanitaries de Catalunya (AQuAS) through contract 2021-021OE. JDe, SMo, VP funded by Netzwerk Universitatsmedizin (NUM) project egePan (01KX2021). JPB, SH, TH funded by Federal Ministry of Education and Research (BMBF; grant 05M18SIA). KH, MSc, YKh funded by Project SaxoCOV, funded by the German Free State of Saxony. Presentation of data, model results and simulations also funded by the NFDI4Health Task Force COVID-19 (https://www.nfdi4health.de/task-force-covid-19-2) within the framework of a DFG-project (LO-342/17-1). LP, VE funded by Mathematical and Statistical modelling project (MUNI/A/1615/2020), Online platform for real-time monitoring, analysis and management of epidemic situations (MUNI/11/02202001/2020); VE also supported by RECETOX research infrastructure (Ministry of Education, Youth and Sports of the Czech Republic: LM2018121), the CETOCOEN EXCELLENCE (CZ.02.1.01/0.0/0.0/17-043/0009632), RECETOX RI project (CZ.02.1.01/0.0/0.0/16-013/0001761). NIB funded by Health Protection Research Unit (grant code NIHR200908). SAb, SF funded by Wellcome Trust (210758/Z/18/Z).