Inferring the effectiveness of government interventions against COVID-19

Brauner, Jan Markus; Mindermann, Sören; Sharma, Mrinank; Johnston, David; Salvatier, John; Gavenčiak, Tomáš; Stephenson, Anna B.; Leech, Gavin; Altman, George T.; Mikulik, Vladimir; Norman, Alexander John; Monrad, Joshua Teperowski; Besiroglu, Tamay; Ge, Hong; Hartwick, Meghan A.; Teh, Yee Whye; Chindelevitch, Leonid; Gal, Yarin; Kulveit, Jan

doi:10.1126/science.abd9338

Cited by 915 publications

(1,044 citation statements)

References 67 publications

Supporting

Mentioning

978

Contrasting

Unclassified

Order By: Relevance

“…Reductions of R per intervention Other papers, e.g. [12,87], report, for various interventions, corresponding percent reductions of R. Our model clarifies that it is structurally more robust to report the additive reduction of the reinfections by a certain intervention. For example, according to our model closing schools removes the school reinfections from the dynamics, and in consequence reduces R by about 0.15.…”

Section: Methods and Resultssupporting

confidence: 54%

Predicting the effects of COVID-19 related interventions in urban settings by combining activity-based modelling, agent-based simulation, and mobile phone data

Müller

Balmer²,

Charlton

et al. 2021

Preprint

View full text Add to dashboard Cite

Epidemiological simulations as a method are used to better understand and predict the spreading of infectious diseases, for example of COVID-19.This paper presents an approach that combines a well-established approach from transportation modelling that uses person-centric data-driven human mobility modelling with a mechanistic infection model and a person-centric disease progression model. The model includes the consequences of different room sizes, air exchange rates, disease import, changed activity participation rates over time (coming from mobility data), masks, indoors vs. outdoors leisure activities, and of contact tracing. The model is validated against the infection dynamics in Berlin (Germany).The model can be used to understand the contributions of different activity types to the infection dynamics over time. The model predicts the effects of contact reductions, school closures/vacations, masks, or the effect of moving leisure activities from outdoors to indoors in fall, and is thus able to quantitatively predict the consequences of interventions. It is shown that these effects are best given as additive changes of the reinfection rate R. The model also explains why contact reductions have decreasing marginal returns, i.e. the first 50% of contact reductions have considerably more effect than the second 50%.Our work shows that is is possible to build detailed epidemiological simulations from microscopic mobility models relatively quickly. They can be used to investigate mechanical aspects of the dynamics, such as the transmission from political decisions via human behavior to infections, consequences of different lockdown measures, or consequences of wearing masks in certain situations. The results can be used to inform political decisions.Author summaryEvidently, there is an interest in models that are able to predict the effect of interventions in the face of pandemic diseases. The so-called compartmental models have difficulties to include effects that stem from spatial, demographic or temporal inhomongeneities. Person-centric models, often using social contact matrices, are difficult and time-consuming to build up. In the present paper, we describe how we built a largely data-driven person-centric infection model within less than a month when COVID-19 took hold in Germany. The model is based on our extensive experience with mobility modelling, and a synthetic data pipeline that starts with mobile phone data, while taking the infection dynamics and the disease progression from the literature. The approach makes the model portable to all places that have similar so-called activity-based models of travel in place, which are many places world-wide, and the number is continuously increasing. The model has been used since its inception to regularly advise the German government on expected consequences of interventions.

show abstract

Section: Methods and Resultssupporting

confidence: 54%

Predicting the effects of COVID-19 related interventions in urban settings by combining activity-based modelling, agent-based simulation, and mobile phone data

Müller

Balmer²,

Charlton

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The role of schools in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection and transmission is highly debated and ranges from "there is no evidence of danger" (1) to "overlooked or at risk" (2,3). This pronounced contradiction has its roots in a puzzle that has many missing pieces.…”

Section: Introductionmentioning

confidence: 99%

“…School outbreaks have been described (2,9), but it is not clear how frequently they occur and how much they contribute to the spread of SARS-CoV-2 on a population level. Retrospective modeling studies based on many assumptions are highly controversial (1,3,10,11). Some groups have shown that closing schools and universities without making a distinction between these different settings, was associated with a 38% reduction of the reproduction factor R during the first wave of the SARS-CoV-2 pandemic (3).…”

Section: Introductionmentioning

confidence: 99%

Surveillance of Acute SARS-CoV-2 Infections in School Children and Point-Prevalence During a Time of High Community Transmission in Switzerland

et al. 2021

View full text Add to dashboard Cite

Background: Switzerland had one of the highest incidence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections in Europe during the second wave. Schools were open as in most of Europe with specific preventive measures in place. However, the frequency and transmission of acute unrecognized, asymptomatic or oligosymptomatic infections in schools during this time of high community transmission is unknown. Thereof, our aim was to pilot a surveillance system that detects acute SARS-CoV-2 infections in schools and possible transmission within classes.Methods: Fourteen out of the randomly selected sample of the Ciao Corona cohort study participated between December 1 and 11, a time when incidence rate for SARS-CoV-2 infections was high for the canton of Zurich. We determined point-prevalence of acute SARS-CoV-2 infections of school children attending primary and secondary school. A buccal swab for polymerase chain reaction (PCR) and a rapid diagnostic test (RDT) to detect SARS-CoV-2 were taken twice 1 week apart (T1 and T2) in a cohort of children from randomly selected classes. A questionnaire assessed demographics and symptoms compatible with a SARS-CoV-2 infection during the past 5 days.Results: Out of 1,299 invited children, 641 (49%) 6- to 16-year-old children and 66 teachers from 14 schools and 67 classes participated in at least one of two testings. None of the teachers but one child had a positive PCR at T1, corresponding to a point-prevalence in children of 0.2% (95% CI 0.0–1.1%), and no positive PCR was detected at T2. The child with positive PCR at T1 was negative on the RDT at T1 and both tests were negative at T2. There were 7 (0.6%) false positive RDTs in children and 2 (1.7%) false positive RDTs in teachers at T1 or T2 among 5 schools (overall prevalence 0.7%). All 9 initially positive RDTs were negative in a new buccal sample taken 2 h to 2 days later, also confirmed by PCR. Thirty-five percent of children and 8% of teachers reported mild symptoms during the 5 days prior to testing.Conclusion: In a setting of high incidence of SARS-CoV-2 infections, unrecognized virus spread within schools was very low. Schools appear to be safe with the protective measures in place (e.g., clearly symptomatic children have to stay at home, prompt contact tracing with individual and class-level quarantine, and structured infection prevention measures in school). Specificity of the RDT was within the lower boundary of performance and needs further evaluation for its use in schools. Given the low point prevalence even in a setting of very high incidence, a targeted test, track, isolate and quarantine (TTIQ) strategy for symptomatic children and school personnel adapted to school settings is likely more suitable approach than surveillance on entire classes and schools.Clinical Trial Registration:https://clinicaltrials.gov/ct2/show/NCT04448717, ClinicalTrials.gov NCT04448717.

show abstract

“…To date, most data-driven policy analyses have been retrospective and focused on overall effectiveness across a number of counties, states, or countries [7][8][9][10]. These analyses can be broadly useful for policy makers deciding which controls to apply when attempting to contain an outbreak already in progress.…”

Section: Introductionmentioning

confidence: 99%

County-Specific, Real-Time Projection of the Effect of Business Closures on the COVID-19 Pandemic

Yurk

Abu‐Mostafa

2021

Preprint

View full text Add to dashboard Cite

Public health policies such as business closures have been one of our most effective tools in slowing the spread of COVID-19, but they also impose costs. This has created demand from policy makers for models which can predict when and where such policies will be most effective to head off a surge and where they could safely be loosened. No current model combines data-driven, real-time policy effect predictions with county-level granularity. We present a neural net-based model for predicting the effect of business closures or re-openings on the COVID-19 time-varying reproduction number Rt in real time for every county in California. When trained on data from May through September the model accurately captured relative county dynamics during the October/November California COVID-19 surge (r2=0.76), indicating robust out-of-sample performance. To showcase the model's potential utility we present a case study of various counties in mid-October. Even when counties imposed similar restrictions at the time, our model successfully distinguished counties in need of drastic and immediate action to head off a surge from counties in less dire need of intervention. While this study focuses on business closures in California, the presented model architecture could be applied to other policies around world.

show abstract

Inferring the effectiveness of government interventions against COVID-19

Cited by 915 publications

References 67 publications

Predicting the effects of COVID-19 related interventions in urban settings by combining activity-based modelling, agent-based simulation, and mobile phone data

Predicting the effects of COVID-19 related interventions in urban settings by combining activity-based modelling, agent-based simulation, and mobile phone data

Surveillance of Acute SARS-CoV-2 Infections in School Children and Point-Prevalence During a Time of High Community Transmission in Switzerland

County-Specific, Real-Time Projection of the Effect of Business Closures on the COVID-19 Pandemic

Contact Info

Product

Resources

About