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

Sherratt, K., Gruson, H., Grah, R., Tuček, Vít, Šmíd, Martin, Zajíček, Milan.

In eLife, volume 12, issue April 2023 , 2023, ISSN 2050-084X.

BACKGROUND: Short-term forecasts of infectious disease contribute to situational awareness and capacity planning. Based on best practice in other fields and recent insights in infectious disease epidemiology, one can maximise forecasts’ predictive performance by combining independent models into an ensemble. Here we report the performance of ensemble predictions of COVID-19 cases and deaths across Europe from March 2021 to March 2022. METHODS: We created the European COVID-19 Forecast Hub, an online open-access platform where modellers upload weekly forecasts for 32 countries with results publicly visualised and evaluated. We created a weekly ensemble forecast from the equally-weighted average across individual models' predictive quantiles. We measured forecast accuracy using a baseline and relative Weighted Interval Score (rWIS). We retrospectively explored ensemble methods, including weighting by past performance. RESULTS: We collected weekly forecasts from 48 models, of which we evaluated 29 models alongside the ensemble model. The ensemble had a consistently strong performance across countries over time, performing better on rWIS than 91% of forecasts for deaths (N=763 predictions from 20 models), and 83% forecasts for cases (N=886 predictions from 23 models). Performance remained stable over a 4-week horizon for death forecasts but declined with longer horizons for cases. Among ensemble methods, the most influential choice came from using a median average instead of the mean, regardless of weighting component models. CONCLUSIONS: Our results support combining independent models into an ensemble forecast to improve epidemiological predictions, and suggest that median averages yield better performance than methods based on means. We highlight that forecast consumers should place more weight on incident death forecasts than case forecasts at horizons greater than two weeks. FUNDING: European Commission, Ministerio de Ciencia, Innovación y Universidades, FEDER; Ag?ncia de Qualitat i Avaluació Sanit?ries de Catalunya; Netzwerk Universitätsmedizin; Health Protection Research Unit; Wellcome Trust; European Centre for Disease Prevention and Control; Ministry of Science and Higher Education of Poland; Federal Ministry of Education and Research; Los Alamos National Laboratory; German Free State of Saxony; NCBiR; FISR 2020 Covid-19 I Fase; Spanish Ministry of Health / REACT-UE (FEDER); National Institutes of General Medical Sciences; Ministerio de Sanidad/ISCIII; PERISCOPE European H2020; PERISCOPE European H2021; InPresa; National Institutes of Health, NSF, US Centers for Disease Control and Prevention, Google, University of Virginia, Defense Threat Reduction Agency.

Influence of COVID-19-Related Restrictions on the Prevalence of Overweight and Obese Czech Children.

Vážná, A., Vignerová, J., Brabec, Marek, Novák, J., Procházka, B., Gabera, A., Sedlák, P..

In International Journal of Environmental Research and Public Health, volume 19, issue č. 19 , 2022, ISSN 1661-7827.

Apart from influencing the health of the worldwide population, the COVID-19 pandemic changed the day-to-day life of all, including children. A sedentary lifestyle along with the transformation of eating and sleep habits took place in the child population. These changes created a highly obesogenic environment. Our aim was to evaluate the current weight in the child population and identify the real effects of the pandemic. Height and weight data were collected by pediatricians from the pre-COVID-19 and post-COVID-19 periods from 3517 children (1759 boys and 1758 girls) aged 4.71 to 17.33 years. We found a significant rise in the z-score BMI between pediatric visits in the years 2019 and 2021 in both sexes aged 7, 9, 11, and 13 years. Especially alarming were the percentages of (severely) obese boys at the ages of 9 and 11 years, which exceed even the percentages of overweight boys. With the use of statistical modeling, we registered the most dramatic increment at around 12 years of age in both sexes. Based on our research in the Czech Republic, we can confirm the predictions that were given at the beginning of the pandemic that COVID-19-related restrictions worsened the already present problem of obesity and excess weight in children.

Delays, Masks, the Elderly, and Schools: First Covid-19 Wave in the Czech Republic.

Berec, Luděk, Smyčka, J., Levínský, René, Hromádková, Eva, Šoltés, Michal, Šlerka, J., Tuček, V., Trnka, J., Šmíd, Martin, Zajíček, Milan, Diviák, T., Neruda, Roman, Vidnerová, Petra.

In Bulletin of Mathematical Biology, volume 84, issue č. 8 , 2022, ISSN 0092-8240.

Running across the globe for nearly 2 years, the Covid-19 pandemic keeps demonstrating its strength. Despite a lot of understanding, uncertainty regarding the efficiency of interventions still persists. We developed an age-structured epidemic model parameterized with epidemiological and sociological data for the first Covid-19 wave in the Czech Republic and found that (1) starting the spring 2020 lockdown 4 days earlier might prevent half of the confirmed cases by the end of lockdown period, (2) personal protective measures such as face masks appear more effective than just a realized reduction in social contacts, (3) the strategy of sheltering just the elderly is not at all effective, and (4) leaving schools open is a risky strategy. Despite vaccination programs, evidence-based choice and timing of non-pharmaceutical interventions remains an effective weapon against the Covid-19 pandemic.

Protection provided by vaccination, booster doses and previous infection against covid-19 infection, hospitalisation or death over time in Czechia.

Berec, Luděk, Šmíd, Martin, Přibylová, L., Májek, O., Pavlík, T., Jarkovský, J., Zajíček, Milan, Weiner, J., Barusová, Tamara, Trnka, J..

In PLoS ONE, volume 17, issue č. 7 , 2022, ISSN 1932-6203.

Studies demonstrating the waning of post-vaccination and post-infection immunity against covid-19 generally analyzed a limited range of vaccines or subsets of populations. Using Czech national health data from the beginning of the covid-19 pandemic till November 20, 2021 we estimated the risks of reinfection, breakthrough infection, hospitalization and death by a Cox regression adjusted for sex, age, vaccine type and vaccination status. Vaccine effectiveness against infection declined from 87% at 0-2 months after the second dose to 53% at 7-8 months for BNT162b2 vaccine, from 90% at 0-2 months to 65% at 7-8 months for mRNA-1273, and from 83% at 0-2 months to 55% at 5-6 months for the ChAdOx1-S. Effectiveness against hospitalization and deaths declined by about 15% and 10%, respectively, during the first 6-8 months. Boosters (third dose) returned the protection to the levels observed shortly after dose 2. In unvaccinated, previously infected individuals the protection against infection declined from 97% after 2 months to 72% at 18 months. Our results confirm the waning of vaccination-induced immunity against infection and a smaller decline in the protection against hospitalization and death. Boosting restores the original vaccine effectiveness. Post-infection immunity also decreases over time.

Using a Deep Neural Network in a Relative Risk Model to Estimate Vaccination Protection for COVID-19.

Suchopárová, Gabriela, Vidnerová, Petra, Neruda, Roman, Šmíd, Martin.

In Engineering Applications of Neural Networks, pages 310-320 , , ISSN 1865-0929.

The proportional hazard Cox model is traditionally used in survival analysis to estimate the effect of several variables on the hazard rate of an event. Recently, neural networks were proposed to improve the flexibility of the Cox model. In this work, we focus on an extension of the Cox model, namely on a non-proportional relative risk model, where the neural network approximates a non-linear time-dependent risk function. We address the issue of the lack of time-varying variables in this model, and to this end, we design a deep neural network model capable of time-varying regression. The target application of our model is the waning of post-vaccination and post-infection immunity in COVID-19. This task setting is challenging due to the presence of multiple time-varying variables and different epidemic intensities at infection times. The advantage of our model is that it enables a fine-grained analysis of risks depending on the time since vaccination and/or infection, all approximated using a single non-linear function. A case study on a data set containing all COVID-19 cases in the Czech Republic until the end of 2021 has been performed. The vaccine effectiveness for different age groups, vaccine types, and the number of doses received was estimated using our model as a function of time. The results are in accordance with previous findings while allowing greater flexibility in the analysis due to a continuous representation of the waning function.

Pandemic-Driven Innovations Contribute to the Development of Information-Based Medicine.

Kalina, Jan.

In Digital Innovation for Healthcare in COVID-19: Pandemic Strategies and Solutions, pages 245-262 , .

The outbreak of the COVID-19 pandemic accelerated trends towards introducing innovative digital tools tailor-made for various applications in medical care or public health. This chapter is focused on decision support systems as important examples of artificial intelligence tools with an increasing popularity. Their potential to contribute to targeting the measures adopted against the COVID-19 pandemic is discussed. In connection with applying artificial intelligence tools in healthcare, their ability to perform epidemic modeling or to contribute to targeting public health interventions are discussed as well. The expected remarkable transforms of medical care including its informatization is described here by the concept of information-based medicine, which exceeds the limited pre-pandemic ideals of evidence-based medicine. The same is true for the concept of information-based public health.

Importance of vaccine action and availability and epidemic severity for delaying the second vaccine dose.

Berec, Luděk, Levínský, R., Weiner, J., Šmíd, Martin, Neruda, Roman, Vidnerová, Petra, Suchopárová, Gabriela.

In Scientific Reports, volume 12, issue č. 1 , 2022, ISSN 2045-2322.

Following initial optimism regarding potentially rapid vaccination, delays and shortages in vaccine supplies occurred in many countries during spring 2021. Various strategies to counter this gloomy reality and speed up vaccination have been set forth, of which the most popular has been to delay the second vaccine dose for a longer period than originally recommended by the manufacturers. Controversy has surrounded this strategy, and overly simplistic models have been developed to shed light on this issue. Here we use three different epidemic models, all accounting for then actual COVID- 19 epidemic in the Czech Republic, including the real vaccination rollout, to explore when delaying the second vaccine dose by another 3 weeks from 21 to 42 days is advantageous. Using COVID-19-related deaths as a quantity to compare various model scenarios, we find that the way of vaccine action at the beginning of the infection course (preventing infection and symptoms appearance), mild epidemic and sufficient vaccine supply rate call for the original inter-dose period of 21 days regardless of vaccine efficacy. On the contrary, for the vaccine action at the end of infection course (preventing severe symptoms and death), severe epidemic and low vaccine supply rate, the 42-day inter-dose period is preferable, at any plausible vaccine efficacy.

Model-M: An agent-based epidemic model of a middle-sized municipality.

Berec, L., Diviák, T., Kuběna, Aleš Antonín, Levínský, René, Neruda, Roman, Suchopárová, Gabriela, Šlerka, J., Šmíd, Martin, Tuček, V., Vidnerová, Petra, Zajíček, Milan, Zapletal, František.

, 2022,

This report presents a technical description of our agent-based epidemic model of a particular middle-sized municipality. We have developed a realistic model with 56 thousand inhabitants and 2.7 millions of social contacts. These form a multi-layer social network that serves as a base of our epidemic simulation. The disease is modeled by our extended SEIR model with parameters fitted to real epidemics data for Czech Republic. The model is able to simulate a whole range of non-pharmaceutical interventions on individual level, such as protective measures and physical distancing, testing, contact tracing, isolation and quarantine. The effect of government-issued measures such as contact restrictions in different environments (schools, restaurants, vendors, etc.) can also be simulated.

Epicity.

Berec, Luděk, Diviák, T., Kuběna, Aleš Antonín, Levínský, René, Neruda, Roman, Suchopárová, Gabriela, Šlerka, J., Šmíd, Martin, Trnka, Jan, Tuček, Vít, Vidnerová, Petra, Vrbenský, Karel, Zajíček, Milan, Zapletal, František.

, 2021,

Model M is an agent-based epidemic model for COVID-19 computational experiments on realistic multi-graph social networks. It allows to simulate projections of main epidemic indicators with respect to various interventions. These include lockdowns, closures of different contact layers (leisure, schools, etc.), social distancing, testing and quarantine, contact tracing, and vaccination.

Simulation of non-pharmaceutical interventions in an agent based epidemic model.

Vidnerová, Petra, Neruda, Roman, Suchopárová, Gabriela, Berec, L., Diviák, T., Kuběna, Aleš Antonín, Levínský, René, Šlerka, J., Šmíd, Martin, Trnka, J., Tuček, V., Vrbenský, Karel, Zajíček, Milan.

In Proceedings of the 21st Conference Information Technologies – Applications and Theory (ITAT 2021), pages 263-268 , , ISSN 1613-0073.

The standard SEIR equation-based models represent the state-of-the-art approach in epidemiological modelling. Their drawbacks include unrealistic infection-related contact estimates and difficulties in modelling nonpharmaceutical interventions, such as contact reductions or partial closures. In this paper, we present our agent-based model that addresses the above-mentioned issues. It works with a population of individuals (agents) and their contacts are modelled as a multi-graph social network according to real data based on a Czech county. Custom algorithmic procedures simulating testing, quarantine and partial closures of various contact types are implemented. The model can serve as a tool for relative comparison of the efficacy of various policies. It was also used for a study comparing various interventions in Czech primary and secondary schools, using a graph based on real data from a selected Czech school.

Could the new coronavirus have infected humans prior November 2019?.

Tichopád, A., Pecen, Ladislav, Sedlák, V..

In PLoS ONE, volume 16, issue č. 8 , 2021, ISSN 1932-6203.

The pandemic caused by the SARS-CoV-2 virus is believed to originate in China from where it spread to other parts of the world. The first cluster of diseased individuals was reported in China as early as in December 2019. It has also been well established that the virus stroke Italy later in January or in February 2020, hence distinctly after the outbreak in China. The work by Apolone et al. published in the Italian Medical Journal in November 2020 and retracted upon expression of concern on 22 March 2021, however propose that the virus could have stroke people already in September 2019, possibly following even earlier outbreak in China. By fitting an early part of the epidemic curve with the exponential and extrapolating it backwards, we could estimate the day-zero of the epidemic and calculated its confidence intervals in Italy and China. We also calculated how probable it is that Italy encountered the virus prior 1 January 2020. We determined an early portion of the epidemic curve representing unhindered exponential growth which fit the exponential model with high determination >0.97 in both countries. We conservatively suggest that the day-zero in China and Italy was 8 December 2019 (95% CI: 3 Dec., 20 Dec.) and 22 January 2020 (95% CI: 16 Jan., 29 Jan.), respectively. Given the uncertainty of the very early data in China and adjusting hence our model to fit the exponentially behaved data only, we can even admit that the pandemic originated through November 2019 (95% CI: 31 Oct., 22 Dec.). With high confidence (p <0.01) China encountered the virus prior Italy. We generally view any pre-pandemic presence of the virus in humans before November 2019 as very unlikely. The later established dynamics of the epidemics data suggests that the country of the origin was China.

COVID-19-free pracoviště: Vyšetření protilátek proti koronaviru jako základ testovací strategie ve firmách.

Krátká, Z., Sedláčková, L., Luxová, Š., Hrubá, D., Katina, Stanislav.

In Časopis lékařů českých, volume 160, issue č. 4, pages 126-132 , 2021, ISSN 0008-7335.

Česko patří mezi země nejvíce zasažené koronavirovou pandemií – přibližně 16 % obyvatel mělo pozitivní test PCR, 2–3x více lidí prodělalo infekci bez podstoupení tohoto vyšetření. Pro zaměstnavatele je velmi užitečné vědět, kolik zaměstnanců již infekci prodělalo a pro kolik osob je koronavirus nadále rizikový. Za tímto účelem je vhodné vyšetřit IgG protilátky. V současné době je však strategie testování jiná – povinně se provádí testování antigenními testy s cílem hledat infekční osoby bez ohledu na imunitu lidí. Cílem této pilotní studie bylo stanovit počet imunních osob po prodělané infekci na třech klinikách GENNET, s. r. o. Současně se antigenními testy zjišťovala infekce u neočkovaných osob, které neprodělaly COVID-19 nebo jej prodělaly před více než 3 měsíci. Soubor zahrnoval 297 jedinců, z nichž 182 (61,3 %) nebylo očkováno a 115 (38,7 %) bylo po vakcinaci. Z neočkovaných mělo 71 (39 %) osob v anamnéze pozitivní test PCR, dalších 18 (9,9 %) mělo pozitivní IgG protilátky, aniž by věděly o prodělané infekci, a 38 (20,9 %) mělo negativní IgG protilátky. Zatím nevyšetřených bylo 55 (30,2 %) osob. Sečteme-li očkované s osobami s protilátkami, pak imunních bylo 74,3 % zaměstnanců kliniky GENNET Archa, 68 % zaměstnanců kliniky GENNET Kostelní a 58,1 % kliniky GENNET Liberec. Antigenním testem bylo ve 4 kolech vyšetřeno v průměru 153 osob (přičemž 60 z nich mělo protilátky). Infekce byla zjištěna u 2 osob. Obě patřily do skupiny bez vyšetřených protilátek. Žádná osoba s protilátkami neměla pozitivní antigenní test. Lidé, kteří mají protilátky po očkování nebo po infekci, jsou vůči opakované infekci odolné a je u nich nízké riziko, že budou nadále virus šířit. Vyšetřením protilátek zaměstnavatelé získají lepší přehled o situaci na pracovištích. Na základě naší studie doporučujeme zrušit plošné antigenní testování u osob s protilátkami.

Kernel density estimation for circular data about COVID-19 in the Czech Republic.

Katina, Stanislav, Zámečník, S., Hórová, I..

In ISCB 2021: 42nd Annual Conference of the International Society for Biostatistics: Final Programme & Book of Abstracts, pages 244-244 , .

The term circular statistics describes a set of techniques used to model distributions of random variables that are cyclic in nature and these approaches can be easily adapted to temporal data recorded, e.g., daily, weekly or monthly. One of the nonparametric possibilities how to analyze these data is through kernel estimations of circular densities where the problem of how much to smooth, i.e., how to choose the bandwidth, is crucial. In this presentation we describe the existing methods: cross-validation method, smoothed cross-validation, adaptive method and propose their modifications. We apply these methods on real data from the Institute of health information and statistics of the Czech Republic about total (cumulative) number of persons with a proven COVID-19 infection according to regional hygienic stations, number of cured persons, number of deaths and tests performed for whole country and regions coded based on nomenclature of territorial units for Statistics (NUTS). The results are visualized as circular histograms (rose diagrams) and calculated standardized characteristics are superimposed with choropleth map, where NUTS are shaded in diverging color scheme. All statistical analyses are performed in the R software.

Agentní modely.

Neruda, Roman.

In Poučení z pandemie COVID-19. Program konference , .

Simulace epidemiologických opatření v modelu M.

Vidnerová, Petra.

In Poučení z pandemie COVID-19. Program konference , .

Jsou děti významnými šiřiteli koronavirové infekce, nebo se jedná o mýtus?.

Katina, Stanislav, Krátká, Z..

ProLekare.cz, 8. 4. 2021,

Koronavirová epidemie omezila přístup ke vzdělání jednomu a čtvrt milionu dětí v Česku již ve dvou školních rocích. Smyslem zavedení distanční výuky bylo zamezit šíření infekce populací, ale při zavádění těchto opatření se nebere zřetel na publikované vědecké studie sledující infekčnost dětí. Cílem tohoto sdělení je upozornit jednak na první českou epidemiologickou studii provedenou u dětí a mladé populace, a zároveň na zahraniční práce, jež se této tématice věnují.

Proč používat antigenní testy u lékaře, ale ne při plošném testování?.

Katina, Stanislav, Krátká, Z..

ProLekare.cz, 24. 3. 2021,

Proč antigenní test umožňuje poměrně úspěšně potvrdit infekci koronavirem u pacienta testovaného dobře zaškolenými zdravotníky a proč naopak odborníci nedoporučují využití antigenních testů v plošném měřítku, například ve školách. Z čeho mají obavy? Na celou problematiku se podíváme z pohledu matematických zákonitostí i klinické praxe a reality.

Los o ventilátor je férové řešení. Ministerstvo nechalo lékaře napospas v odporné situaci, říká bioetik.

Vrtišková Nejezchlebová, L., Černý, David.

In Deník N, issue 3. března 2021 , 2021, ISSN 2571-1717.

„Jste plavčík a na obou koncích bazénu se zároveň topí dva lidé. Pak nemáte povinnost zachránit oba. Musíte udělat volbu. A to je přesně ta situace, kdy má lékař jeden ventilátor a dva pacienty,“ říká expert na medicínskou etiku David Černý v rozhovoru o triáži pacientů, ke které už v českých nemocnicích dochází. A o tom, že ministerstvo nechalo lékaře napospas situaci, kterou nezavinili.

Přeplňování nemocnic jsme zabránili, do Vánoc bychom ale neměli rozvolňovat, míní Neruda z Centra pro modelování.

Neruda, Roman.

Česká televize, 2020,

Opatření proti šíření nemoci covid-19 by neměla být před Vánoci rozvolňována, domnívá se informatik Roman Neruda z Centra pro modelování biologických a společenských procesů. Epidemie podle něj zpomaluje, ale ne dostatečně rychle. Pokud zůstaneme u současných opatření, nehrozí ale podle něj již větší přetížení nemocnic.

Čísla jsou lepší než realita, PES by se měl přepočítat, myslí si informatik.

Zangová C., , Neruda, Roman.

CNN Prima News, 2020,

Andrej Babiš by chtěl před Vánoci zrealizovat dobrovolné plošné testování antigenními testy. To by mohlo ze společnosti vyčlenit bezpříznakové jedince, ale právě tyto testy zkreslují realitu, která nyní vypadá pozitivnější, než doopravdy je. V pořadu 360° Pavlíny Wolfové to řekl informatik Roman Neruda.

Model M – Mikropohled města (přednáška).

Neruda, Roman.

In Konference COVID v modelech , .

Z hlediska matematiků jsme absolutně neměli uvolnit opatření, tvrdí informatik Neruda.

Míková E., , Neruda, Roman.

CNN Prima News, 2020,

Počet nakažených koronavirem stále stoupá a mnoho z nás má své teorie, jak se bude situace vyvíjet dále. Co si o vývoji ale myslí informatici? Na to v Novém dni odpovídal Roman Neruda z Centra pro modelování biologických a společenských procesů

Ještě by se vyplatilo šlápnout na brzdu, říká informatik ze skupiny „modelářů“. Pokles všichni sledují nadšeně, my opatrně.

Vrtišková Nejezchlebová, L., Neruda, Roman.

In Deník N, issue Online 2. listopadu 2020 , 2020, ISSN 2571-1717.

Oficiální počty nakažených koronavirem v posledním týdnu mírně poklesly, což po měsících rychlého růstu všichni sledujeme s jistou úlevou. Nejen podle informatika Romana Nerudy, který se podílí na tvorbě epidemiologických modelů, bychom však s optimismem měli zatím šetřit. „Jsme přesvědčení, že i kdyby se to teď opravdu začalo snižovat, pořád by se ještě vyplatilo šlápnout víc na brzdu, abychom denní přírůstky dostali v kratším čase k tisícovce, kdy začne efektivně fungovat trasování. Ted jsme rozhodně vzdáleni od jakéhokoli rozvolňování.“

Model M – Několik poznámek k optimálnímu algoritmu pro Chytrou karanténu (přednáška).

Neruda, Roman.

In Sympozium: Změny chování české populace v době COVID-19 a jejich reflexe v epidemiologických modelech , .

Model M - Cesta k věrnému modelu města (přednáška).

Neruda, Roman.

In NZIS Open 2020 , .

Měření dopadů pandemie covid-19 dle celkového počtu úmrtí za týden bez ohledu na příčinu úmrtí a porovnání s počty zemřelých v předchozích 8 letech (přednáška).

Pecen, Ladislav.

In 41. imunoanalytické dny 11. ročník workshopu personalizované medicíny. Sborník abstraktů , .

Několik zajímavosti kolem epidemiologie covid-19 aneb jak měřit dopady pandemie a proč předpovědi někdy sedí a někdy naprosto selhávají (přednáška).

Pecen, Ladislav.

In Den otevřených dveří Ústavu informatiky AV ČR , .