ABSTRACT
BackgroundShort-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. MethodsWe 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 from a standardised source over the next one to four 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. ResultsOver 52 weeks we collected and combined up to 28 forecast models for 32 countries. 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 84% of participating models forecasts of incident cases (with a total N=862), and 92% of participating models forecasts of deaths (N=746). Across a one to four week time horizon, ensemble performance declined with longer forecast periods when forecasting cases, but remained stable over four 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. ConclusionsOur 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 two weeks. Code and data availabilityAll data and code are publicly available on Github: covid19-forecast-hub-europe/euro-hub-ensemble.
ABSTRACT
IntroductionCurrently, information on infection and transmission risks of students and teachers in schools, the effect of infection control measures for schools as well as the contribution of schools to the overall population transmission of SARS-CoV-2 in Germany are limited to regional data sets restricted to short phases of the pandemic. MethodsWe used German federal state (NUTS-2) and county (NUTS-3) data from national and regional public health and education agencies to assess infection risk and secondary attack rates (SARs) from March 2020 to October 2021 in Germany. We used multiple regression analysis and infection dynamic modelling, accounting for urbanity, socioeconomic factors, local population infection dynamics and age-specific underdetection to investigate the effects of infection control measures. ResultsWe included (1) nation-wide NUTS-2 level data from calendar weeks (W) 46-50/2020 and W08-40/2021 with 304676 infections in students and 32992 in teachers; (2) NUTS-3 level data from W09-25/2021 with 85788 student and 9427 teacher infections and (3) detailed data from 5 regions covering W09/2020 to W27/2021 with 12814 infections, 43238 contacts and 4165 secondary cases for students (for teachers 14801, 5893 and 472 respectively). In counties with mandatory surgical mask wearing during class in all schools infection risk of students and teachers was reduced by 56/100.000 persons per 14 days and by 30% and 24% relative to the population respectively. Overall contribution to population infections of contacts in school settings was 2-13%. It was lowest during school closures and vacation and highest during normal presence class intervals. Infection risk for students increased with age and was similar to or lower than the population risk during second and third waves in Germany and higher in summer 2021. Infection risk of teachers was higher than the population during the second wave and similar or lower thereafter with stricter measures in place. SARs for students and staff were below 5% in schools throughout the study period. SARs in households more than doubled from 14% W21-39/2020 to 29-33% in W08-23/2021. Most contacts were reported for schools, yet most secondary cases originated in households. In schools, staff predominantly infected staff and students predominantly infected students. ConclusionOpen schools under hygiene measures and testing strategies contribute up to 13% of nation-wide infections in Germany and as little as 2% during vacations/school closures. Tighter infection control measures stabilised school SARs whilst household SARs more than doubled as more transmissible variants became prevalent in Germany. Mandatory mask wearing during class in all school types effectively reduces secondary transmission in schools, as do reduced attendance class models.
ABSTRACT
IntroductionCurrent estimates of pandemic spread using infectious disease models in Germany for SARS-CoV-2 often do not use age-specific infection parameters and are not always based on known contact matrices of the population. They also do not usually include setting-based information of reported cases and do not account for age-specific underdetection of reported cases. Here, we report likely pandemic spread using an age-structured model to understand the age- and setting-specific contribution of contacts to transmission during all phases of the COVID-19 pandemic in Germany. MethodsWe developed a deterministic SEIRS model using a pre-pandemic contact matrix. The model is optimized to fit reported age-specific SARS-CoV-2 incidences from the Robert Koch Institute, includes information on setting-specific reported cases in schools and integrates age and pandemic period-specific parameters for underdetection of reported cases deduced from a large population-based seroprevalence study. ResultsWe showed that taking underreporting into account, younger adults and teenagers are the main contributors to infections during the first three pandemic waves in Germany. Overall, the contribution of contacts in schools to the total cases in the population was below 10% during the third wave. DiscussionAccounting for the pandemic phase and age-specific underreporting seems important to correctly identify those parts of the population where quarantine, testing, vaccination, and contact-reduction measures are likely to be most effective and efficient. In the future, we will aim to compare current model estimates with currently emerging during-pandemic age-specific contact survey data.
