NIHR ARC Wessex is supporting this research 

Summary

The COVID-19 pandemic created an unprecedented demand for acute care services. Hospitals faced the challenge of needing to rapidly restructure care pathways and resourcing priorities to maximise survival rates for critically ill patients. University Hospital Southampton (UHS) needed evidence-based tools to support clinical and resource planning decisions in response to the pandemic. COVID-19-ED was able to provide rapid response insight to analytics questions provided by UHS.

Who worked on the project?

Professor Michael Boniface – University of Southampton

Dr Hang Phan – University of Southampton

Dr Francis Chmiel – University of Southampton

Dr Daniel Burns – University of Southampton

Professor Ben MacArthur – University of Southampton

Professor Dave Woods – University of Southampton

Dr Derek Sandeman – University Hospital Southampton NHS Foundation Trust

Dr Thomas Daniels – University Hospital Southampton NHS Foundation Trust

Dr Michael Kiuber – University Hospital Southampton NHS Foundation Trust

Neil Tape – University Hospital Southampton NHS Foundation Trust

Martin Azor – University Hospital Southampton NHS Foundation Trust

Dr Matthew Stammers – University of Southampton NHS Foundation Trust

What did we find out?

We found that computational models of COVID-19 epidemiology could be used to help forecast COVID-19 hospital demand and support the University Hospital Southampton (UHSFT) in their response to the pandemic.

We learnt how to deliver a series of models and insight throughout the phases of the pandemic first wave (alert, pre-peak, post-peak, recovery).

We learnt to adapt approaches through experimentation and as new information became available to the research team:

• Initially a 10-day rapid response was required balancing timeliness and precision of answers. Little was known about COVID-19 and a simple approach was developed to forecast the time and size of the first peak.

• We then extended the model to consider actual hospital data along with the prevalence of infected people in communities and occupancy types such as general or intensive care beds.

• Finally, the impact of social distancing was incorporated to modify community transmission rates allowing UHS to run what if scenarios considering future assumptions about social distancing policies from the UK government.

What difference did it make?

COVID-19-ED provided a source of rapid insight response for UHSFT operations teams at a time when they had little information to plan resources.

Why was it so important?

Community disease prevalence modelling is an important tool to support capacity planning and resource planning for integrated care systems.

Vital to provide rapid insights with limited information and changing disease and policy situations.

What happened next?

The knowledge gained about COVID-19 epidemic modelling led to

• UK wider leadership of data analytics for Dr Dan Burns in the COVID-19 Regulator Testing Programme

  • Kidd, S.P., Burns, D., Armson, B., Beggs, A.D., Howson, E.L., Williams, A., Snell, G., Wise, E.L., Goring, A., Vincent-Mistiaen, Z. and Grippon, S., 2022. Reverse-transcription loop-mediated isothermal amplification has high accuracy for detecting severe acute respiratory syndrome coronavirus 2 in saliva and nasopharyngeal/oropharyngeal swabs from asymptomatic and symptomatic individuals. The Journal of Molecular Diagnostics, 24(4), pp.320-336.

Ptasinska, A., Whalley, C., Bosworth, A., Poxon, C., Bryer, C., Machin, N., Grippon, S., Wise, E.L., Armson, B., Howson, E.L. and Goring, A., 2021. Diagnostic accuracy of loop-mediated isothermal amplification coupled to nanopore sequencing (LamPORE) for the detection of SARS-CoV-2 infection at scale in symptomatic and asymptomatic populations. Clinical microbiology and infection, 27(9), pp.1348-e1.

Rapid analytics support to NHSE Chief Scientific Officer’s (NHSE CSO) team including:

• Statistical validation of the PCR technology for Project Jupiter in Leamington Spa, which now covers 300,000 PCR tests a day of the UK’s Pillar 2 testing infrastructure

• Development of a calculator tool for TVG which will be used to validate several COVID-19 testing technologies

• Research design for laboratory validation studies