National and subnational short-term forecasting of COVID-19 in Germany and Poland during early 2021.
Journal
Communications medicine
ISSN: 2730-664X
Titre abrégé: Commun Med (Lond)
Pays: England
ID NLM: 9918250414506676
Informations de publication
Date de publication:
31 Oct 2022
31 Oct 2022
Historique:
received:
19
11
2021
accepted:
22
09
2022
entrez:
9
11
2022
pubmed:
10
11
2022
medline:
10
11
2022
Statut:
epublish
Résumé
During the COVID-19 pandemic there has been a strong interest in forecasts of the short-term development of epidemiological indicators to inform decision makers. In this study we evaluate probabilistic real-time predictions of confirmed cases and deaths from COVID-19 in Germany and Poland for the period from January through April 2021. We evaluate probabilistic real-time predictions of confirmed cases and deaths from COVID-19 in Germany and Poland. These were issued by 15 different forecasting models, run by independent research teams. Moreover, we study the performance of combined ensemble forecasts. Evaluation of probabilistic forecasts is based on proper scoring rules, along with interval coverage proportions to assess calibration. The presented work is part of a pre-registered evaluation study. We find that many, though not all, models outperform a simple baseline model up to four weeks ahead for the considered targets. Ensemble methods show very good relative performance. The addressed time period is characterized by rather stable non-pharmaceutical interventions in both countries, making short-term predictions more straightforward than in previous periods. However, major trend changes in reported cases, like the rebound in cases due to the rise of the B.1.1.7 (Alpha) variant in March 2021, prove challenging to predict. Multi-model approaches can help to improve the performance of epidemiological forecasts. However, while death numbers can be predicted with some success based on current case and hospitalization data, predictability of case numbers remains low beyond quite short time horizons. Additional data sources including sequencing and mobility data, which were not extensively used in the present study, may help to improve performance. We compare forecasts of weekly case and death numbers for COVID-19 in Germany and Poland based on 15 different modelling approaches. These cover the period from January to April 2021 and address numbers of cases and deaths one and two weeks into the future, along with the respective uncertainties. We find that combining different forecasts into one forecast can enable better predictions. However, case numbers over longer periods were challenging to predict. Additional data sources, such as information about different versions of the SARS-CoV-2 virus present in the population, might improve forecasts in the future.
Sections du résumé
BACKGROUND
BACKGROUND
During the COVID-19 pandemic there has been a strong interest in forecasts of the short-term development of epidemiological indicators to inform decision makers. In this study we evaluate probabilistic real-time predictions of confirmed cases and deaths from COVID-19 in Germany and Poland for the period from January through April 2021.
METHODS
METHODS
We evaluate probabilistic real-time predictions of confirmed cases and deaths from COVID-19 in Germany and Poland. These were issued by 15 different forecasting models, run by independent research teams. Moreover, we study the performance of combined ensemble forecasts. Evaluation of probabilistic forecasts is based on proper scoring rules, along with interval coverage proportions to assess calibration. The presented work is part of a pre-registered evaluation study.
RESULTS
RESULTS
We find that many, though not all, models outperform a simple baseline model up to four weeks ahead for the considered targets. Ensemble methods show very good relative performance. The addressed time period is characterized by rather stable non-pharmaceutical interventions in both countries, making short-term predictions more straightforward than in previous periods. However, major trend changes in reported cases, like the rebound in cases due to the rise of the B.1.1.7 (Alpha) variant in March 2021, prove challenging to predict.
CONCLUSIONS
CONCLUSIONS
Multi-model approaches can help to improve the performance of epidemiological forecasts. However, while death numbers can be predicted with some success based on current case and hospitalization data, predictability of case numbers remains low beyond quite short time horizons. Additional data sources including sequencing and mobility data, which were not extensively used in the present study, may help to improve performance.
We compare forecasts of weekly case and death numbers for COVID-19 in Germany and Poland based on 15 different modelling approaches. These cover the period from January to April 2021 and address numbers of cases and deaths one and two weeks into the future, along with the respective uncertainties. We find that combining different forecasts into one forecast can enable better predictions. However, case numbers over longer periods were challenging to predict. Additional data sources, such as information about different versions of the SARS-CoV-2 virus present in the population, might improve forecasts in the future.
Autres résumés
Type: plain-language-summary
(eng)
We compare forecasts of weekly case and death numbers for COVID-19 in Germany and Poland based on 15 different modelling approaches. These cover the period from January to April 2021 and address numbers of cases and deaths one and two weeks into the future, along with the respective uncertainties. We find that combining different forecasts into one forecast can enable better predictions. However, case numbers over longer periods were challenging to predict. Additional data sources, such as information about different versions of the SARS-CoV-2 virus present in the population, might improve forecasts in the future.
Identifiants
pubmed: 36352249
doi: 10.1038/s43856-022-00191-8
pii: 10.1038/s43856-022-00191-8
pmc: PMC9622804
doi:
Types de publication
Journal Article
Langues
eng
Pagination
136Subventions
Organisme : Helmholtz Association
ID : SIMCARD
Informations de copyright
© 2022. The Author(s).
Références
Proc Natl Acad Sci U S A. 2019 Feb 19;116(8):3146-3154
pubmed: 30647115
Nat Hum Behav. 2021 Apr;5(4):529-538
pubmed: 33686204
Science. 2021 Apr 9;372(6538):
pubmed: 33658326
Viruses. 2022 Jul 02;14(7):
pubmed: 35891447
Nat Med. 2021 Jan;27(1):94-105
pubmed: 33097835
Eur J Oper Res. 2023 Jan 1;304(1):25-41
pubmed: 34219901
PLoS Comput Biol. 2022 Sep 19;18(9):e1010405
pubmed: 36121848
PLoS One. 2020 Sep 4;15(9):e0238559
pubmed: 32886696
PLoS Comput Biol. 2021 Feb 12;17(2):e1008618
pubmed: 33577550
Sci Rep. 2019 Jan 24;9(1):683
pubmed: 30679458
Nat Commun. 2020 Sep 30;11(1):4962
pubmed: 32999286
MMWR Morb Mortal Wkly Rep. 2021 May 14;70(19):719-724
pubmed: 33988185
Proc Natl Acad Sci U S A. 2020 Nov 17;117(46):28549-28551
pubmed: 33144506
NPJ Digit Med. 2021 Oct 8;4(1):146
pubmed: 34625656
Proc Natl Acad Sci U S A. 2020 Oct 20;117(42):26190-26196
pubmed: 33004629
Lancet Infect Dis. 2022 Aug 31;:
pubmed: 36057267
Nat Commun. 2021 Aug 27;12(1):5173
pubmed: 34453047
Proc Natl Acad Sci U S A. 2019 Nov 26;116(48):24268-24274
pubmed: 31712420
Int J Forecast. 2022 Jul 1;:
pubmed: 35791416
BMC Infect Dis. 2018 May 30;18(1):245
pubmed: 29843621
Arch Public Health. 2020 Jul 14;78:63
pubmed: 32685147
Proc Natl Acad Sci U S A. 2022 Apr 12;119(15):e2113561119
pubmed: 35394862
Proc Natl Acad Sci U S A. 2021 Dec 21;118(51):
pubmed: 34903655
Lancet Infect Dis. 2020 May;20(5):533-534
pubmed: 32087114
PLoS One. 2020 Mar 31;15(3):e0231236
pubmed: 32231392
Proc Natl Acad Sci U S A. 2022 Aug 9;119(32):e2112656119
pubmed: 35921436