COVID-19

The aim of this project was to provide timely information on the local progression of COVID-19 to NHS Grampian to enable the optimal provision of additional resources to deal with changing clinical demand.

We have developed a variety of monitoring and modelling tools. A novel social distancing score aims to identify areas that might be at greater risk of transmitting COVID-19, due to likely constraints in the ability to adhere to social distancing guidelines. This is being used to inform COVID-19 testing provision. A tool for visualising clusters of positive cases and their contacts is being used by the NHS Grampian Health Intelligence team to track local outbreaks. We developed an epidemiological model that has been customised to fit the Grampian outbreak profile. This is used and updated daily by the Health Intelligence team to predict the number of hospitalisations, and estimate the effect that different scenarios (e.g. changes in levels of restrictions) will have on local healthcare demand. We are using the same data to understand the dynamics of hospital pathways with a multi-state statistical modelling framework.

Using the tools we have developed, we hope to continue providing essential information to NHS Grampian for the duration of the pandemic.

The tools we have developed are already being used by NHS Grampian to help inform the local COVID-19 response. They are freely available when possible to encourage use by other NHS boards.

Work Package 1 (Real time monitoring and visualisation)

Aimed to develop data visualisation tools to provide rapid feedback to the NHS to support COVID-19 planning. We have produced:

• A tool for visualising clusters of positive cases and their contacts, freely available on GitHub provides the ability to track local outbreaks. A version of this has been modified and is used internally within NHS Grampian to monitor local clusters on a weekly basis. The code has been shared directly with other NHS boards.
• A novel social distancing score to identify areas that might be at greater risk of transmitting COVID-19. The score combines three measures: percent of people with essential jobs; percent of people living in overcrowded homes; local population density. It is also freely available on GitHub and used by NHS Grampian to inform COVID-19 testing provision.

Work Package 2 (Customised Grampian model)

Focused on modelling tools customised to fit the Grampian outbreak profile. We have developed an epidemiological model using the same compartments as Public Health Scotland, but estimating the local parameters directly from Grampian data. This work was presented at the NHS Grampian R&D conference, October 2021: Blana D, Chadwick EK, Butler J, Nath M, Osler G, Wilde K, Webb S, Black C, Okpo E, and Evans J., Integrating real time surveillance and dynamic modelling of COVID-19 in Grampian to inform an adaptive healthcare response.

We discussed the project in an episode of the Cafe Connect podcast “Following the science: how data modelling is used to inform the COVID-19 response at a local level”, with project PI Dr Dimitra Blana, co-applicants Prof Corri Black and Dr Graham Osler, and Dr Nick Fluck, medical director for NHS Grampian.

Vulnerable patients with multiple health problems need regular clinic appointments and hospital admissions. However, to minimise their chances of being infected with COVID-19, the government has advised them to self-isolate and avoid contact with hospitals and health care settings. This strategy could make their underlying conditions worse, unless the NHS can find a different way of looking after them. We will use clinical data from vulnerable patients whose current treatment is being disrupted to find out what their needs are and to plan alternative ways of continuing their care.