An analytical framework for Test, Trace and Isolate in the UK: optimising and targeting deployment alongside other measures.

Lead Research Organisation: London Sch of Hygiene and Trop Medicine
Department Name: Public Health and Policy

Abstract

We will develop a robust analytical framework to guide decisions about the deployment of Test, Trace and Isolate (TTI) to control the SARS CoV2 epidemic over the next 12 months. As the UK eases lockdown and workplaces, leisure venues and schools re-open, refining and targeting TTI is necessary to prevent the return of rapid exponential epidemic growth. This will require quick feedback, assessment and integration with real-time epidemiological, behavioural and operational data. We will develop mathematical models of SARS CoV2 transmission and TTI processes, to assess how TTI could be most effectively refined and targeted in association with other control measures, and to develop metrics to assess TTI performance. We will consider how both the design of the TTI system and environmental factors affect people's behaviour. We will respond to emerging questions but currently prioritise: 1) optimal combinations of TTI, screening and physical distancing measures, both to reduce transmission and to protect vulnerable groups; 2) appraisal of the role of behavioural responses to take up and adhere to TTI policies, including trade-offs and the consequences for how policies are supported; 3) signals to initiate stepped up intervention intensity; 4) assessment of our understanding of transmission patterns.

Technical Summary

We will develop a robust analytical framework to guide decisions about the deployment of Test, Trace and Isolate (TTI) to control the UK SARS CoV2 epidemic over the next 12 months. This framework will be underpinned by branching process models of SARS CoV2 transmission and contact tracing, which are flexible enough to capture factors like households and clusters while being mathematically transparent enough to allow for direct calibration to data and quantification of uncertainty in predictions. We will provide evidence to inform key decisions about optimal combinations of tracing, screening and distancing measures. The inclusion of detailed network structure, settings and heterogeneity in contact patterns enables us to address issues such as: the role of uptake and adherence to TTI; the impact of TTI on transmission; signals to initiate stepped up intervention intensity; and assessment of the consistency of observed contact tracing data with epidemiological hypotheses. The research will take a socio-technical approach, recognizing that to ensure robust decision-support outputs from mathematical models must be situated within an understanding of human behaviour, and that both the technology and the environment in which it is deployed will influence the results obtained. The framework will be informed throughout by a contextual understanding obtained from key stakeholders, including members of the public who test positive, members of the public who are listed as contacts and asked to self-isolate (who do not necessarily test
positive) and people working within the TTI and outbreak teams.
 
Description Presentation of paper on daily testing among contacts of SARS-CoV-2 cases to the Scientific Advisory Group on Emergencies Meeting 68, Nov 16, 2020
Geographic Reach National 
Policy Influence Type Participation in a advisory committee
Impact Our Test, Trace and Isolating modelling project responded to a commission to consider a specific testing and contact tracing strategy in which contacts of a SARS-CoV-2 case test daily with rapid tests (lateral flow assay tests) in lieu of quarantining (with case isolation as per standard policy). Our modelling analysis, which contributed findings that considered the separate contact patterns and processes of within and outside housheold contacts, found that such a strategy could be as or more effective than the existign policy, prompting its consideration for further pilot testing. Our preliminary analysis has been published as part of SAGE papers: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/950771/s0897-testing-of-traced-contacts.pdf
URL https://www.gov.uk/government/publications/sage-68-minutes-coronavirus-covid-19-response-16-november...
 
Description Presentation of paper on quarantine and testing strategies for travellers returning from abroad for control of the SARS-CoV-2 epidemic in the UK to the Scientific Advisory Group on Emergencies Meeting 71, Dec 3, 2020 (2020)
Geographic Reach National 
Policy Influence Type Participation in a advisory committee
Impact We presented a modelling paper to SAGE meeting no 71 on quarantine and testing strategies for international arrivals returning to household in the United Kingdom to reduce the number of SARS-CoV-2 cases that might arise from infected travellers. Our model considered the fact that in the UK at the time, returning travellers could be returning to households where household members were not under quarantine. We found simulated epidemics initiated by returning travellers were potentially larger when daily lateral flow testing policies were used in lieu of either a 14 day or 5 day + PCR test-to-release quaratine strategy. We cannot attribute specific policy decisions to this modelling work, but subsequent decisions were in alignment with our findings. Benefits to public health, societal well-being and the economy are realised via improved control of the SARS-CoV-2 epidemic.
URL https://www.gov.uk/government/publications/sage-71-minutes-coronavirus-covid-19-response-3-december-...
 
Description Regular and ongoing presentation of reports to SPI-M on epidemic control of SARS-CoV-2/COVID-19 (Scientific Advisory Group for Emergencies sub-committee on modelling), October 2020 - present
Geographic Reach National 
Policy Influence Type Participation in a advisory committee
Impact As as COVID-19 Rapid Response project, we have prioritised responding to commissions and sending research findings via scientific government advisory channels. Since October 2020, our project has regularly provided papers on modelling and routine data analyses of Test, Trace and Isolate (TTI) effectiveness, design choices and possible strategies (20 reports sent to 11 meetings). We have presented papers relating to: - Effects on TTI of testing demand exceeding capacity - Trends in testing and positivity by age, area and over time - A framework for optimal allocation of limited testing capacity - Approaches to considering TTI effectiveness strategies given prevalence - Potential effectiveness of TTI in controlling the epidemic after exiting lockdown measures - Trade-offs between duration of quarantine period and uptake of testing and adherence to quarantine. - Symptom phenotypes across several large community-based datasets and implications for TTI effectiveness of changes to the symptoms criteria for testing - Effectiveness of daily testing strategies for contacts of a case in lieu of quarantine - Effectiveness of quarantine and testing strategies for international travellers returning from abroad. Modelling evidence presented to SPI-M influences government policy via SAGE, and via direct interactions and influence on other bodies including NHS Test and Trace. The topics reflect both commissions from organisations such as NHS Test and Trace, and self-initiated topics. This advice contributes to improved epidemic control of SARS CoV-2 and COVID-19 in the United Kingdom, reducing morbidity and mortality associated with the disease and other health, social and economic harms that have resulted from the epidemic.
URL https://www.gov.uk/government/publications/scientific-advisory-group-for-emergencies-sage-coronaviru...
 
Title Household structured branching process model of SARS-CoV-2 infection and contact tracing 
Description Through the course of our Test, Trace and Isolate (TTI) modelling project we have documented and made further developments to our household-structured branching process model of SARS-CoV-2 infection and contact tracing. The model 
Type Of Material Computer model/algorithm 
Year Produced 2021 
Provided To Others? Yes  
Impact This model and adaptations to it has been used in research to provide advice to the UK government on TTI strategies and effectiveness for control of the SARS-CoV-2 epidemic, including research sent to SPI-M, SAGE and in advice given to NHS Test and Trace. 
URL https://github.com/martyn1fyles/HouseholdContactTracing
 
Title Model of SARS-CoV-2 infection and contact tracing: code for investigating trade-offs between quarantine periods, uptake of testing and adherence to isolation and quarantine 
Description This mathematical model has been adapted from that of a previous project to investigate Test, Trace and Isolate (TTI) strategies for SARS CoV-2 epidemic control in the United Kingdom. It uses a population-based branching process of infections and contact tracing to simulate epidemics and estimate the effect of TTI on the probability of a large SARS-CoV-2 outbreak. This model has been utilised throughout this project for research and advice on SARS CoV-2 epidemic control. One application of the model was to investigate trade-offs between quarantine periods, uptake of testing and adherence to isolation and quarantine, documented here. A fuller descripiton of the model is given here: https://www.medrxiv.org/content/10.1101/2020.08.20.20178558v1. 
Type Of Material Computer model/algorithm 
Year Produced 2020 
Provided To Others? Yes  
Impact The findings of research using this model have been presented to SPI-M, the modelling sub-committee of the Scientific Advisory Group on Emergencies since October 2020. The specific analsyes documented here were presented November 11, 2020. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/937444/S0877_SAGE67_201111_SPI-M-O_Consensus_Statement.pdf 
URL https://github.com/timcdlucas/ringbp/releases/tag/sevenday1.0
 
Description Evidence for symptom phenotypes of COVID-19 among community based cases across several large studies 
Organisation Evergreen Life Group
Country United Kingdom 
Sector Private 
PI Contribution Our team conceived and led the project, designed the main analysis and developed analysis code, which we shared for partners to adapt.
Collaborator Contribution We have been working with researchers from each partner organisation. We have met regularly and jointly discussed and agreed analysis and write-ups of findings. In some cases, they have conducted analsyes on their own data and in some cases we have run analyses directly.
Impact Thus far we have sent two reports to SPI-M, the modelling sub-committee of the UK's Scientific Advisory Group for Emergencies.
Start Year 2020
 
Description Evidence for symptom phenotypes of COVID-19 among community based cases across several large studies 
Organisation King's College London
Country United Kingdom 
Sector Academic/University 
PI Contribution Our team conceived and led the project, designed the main analysis and developed analysis code, which we shared for partners to adapt.
Collaborator Contribution We have been working with researchers from each partner organisation. We have met regularly and jointly discussed and agreed analysis and write-ups of findings. In some cases, they have conducted analsyes on their own data and in some cases we have run analyses directly.
Impact Thus far we have sent two reports to SPI-M, the modelling sub-committee of the UK's Scientific Advisory Group for Emergencies.
Start Year 2020
 
Description Evidence for symptom phenotypes of COVID-19 among community based cases across several large studies 
Organisation Office for National Statistics
Country United Kingdom 
Sector Private 
PI Contribution Our team conceived and led the project, designed the main analysis and developed analysis code, which we shared for partners to adapt.
Collaborator Contribution We have been working with researchers from each partner organisation. We have met regularly and jointly discussed and agreed analysis and write-ups of findings. In some cases, they have conducted analsyes on their own data and in some cases we have run analyses directly.
Impact Thus far we have sent two reports to SPI-M, the modelling sub-committee of the UK's Scientific Advisory Group for Emergencies.
Start Year 2020
 
Description Evidence for symptom phenotypes of COVID-19 among community based cases across several large studies 
Organisation University of Oxford
Department Nuffield Department of Population Health
Country United Kingdom 
Sector Academic/University 
PI Contribution Our team conceived and led the project, designed the main analysis and developed analysis code, which we shared for partners to adapt.
Collaborator Contribution We have been working with researchers from each partner organisation. We have met regularly and jointly discussed and agreed analysis and write-ups of findings. In some cases, they have conducted analsyes on their own data and in some cases we have run analyses directly.
Impact Thus far we have sent two reports to SPI-M, the modelling sub-committee of the UK's Scientific Advisory Group for Emergencies.
Start Year 2020
 
Description Advisory panel with public health stakeholders 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact In order to ensure that our modelling of the Test, Trace and Isolate policies for SARS-CoV-2 control in the UK is responding to the primary concerns of stakeholders and reflects key aspects fo the process, we have established an advisory panel of public health professionals. These discussions have fed into our project's considerations of what strategies might improve TTI and would merit further investigation.
Year(s) Of Engagement Activity 2020,2021