Temporal trends and forecasting of COVID-19 hospitalisations and deaths in Scotland using a national real-time patient-level data platform: a statistical modelling study

July 2021: A paper published in The Lancet Digital Health describes the EAVE II cohort and its use to monitor and forecast COVID-19 health outcomes across the Scottish population since March 2020.

Summary in Plain English

Why are real-time patient-level data needed?

The government, healthcare providers, scientists and others need real-time patient data to understand the course of the COVID-19 pandemic. Digital data platforms can be used to monitor and precisely forecast outcomes of the pandemic such as hospitalisation and death rates at a national level. This in turn can influence public health policy on decisions such as the need to enforce or relax lockdown measures.

What made this study distinctive?

For this study researchers have created the first bespoke patient-level health dataset for the whole Scottish population. Other studies have looked into parts of the national population. The data come from many sources: primary and secondary care as well virus testing and death records. In addition, a number of population characteristics were considered:

• sex,
• age,
• socioeconomic status,
• nursing home residence,
• the urban/rural six-fold classification (UR6),
• clinical characteristics of vulnerable groups.

This custom dataset allowed researchers to look into:

• short-term progression of COVID-19,
• the development of severe COVID-19 disease and Long COVID,
• deaths in people who have COVID-19 in Scotland.

What has been discovered?

Available evidence suggests, for instance, that COVID-19 hospitalisations and deaths in people who tested positive were linked to certain characteristics. These include being male and having chronic heart disease. COVID-19 deaths were defined as being sooner that 28 days after testing positive.

The study also found that in the second wave of the pandemic, there was an increased incidence in older patients with multiple diseases.

What is next?

As results of this unique dataset have provided accurate forecasts of hospitalisation and death rates for Scotland, the government have used the findings to adapt their public health policy.

With more data becoming available, this new statistical prediction model will be further improved to predict COVID-19 cases and health outcomes. The results of the study may also be applicable to countries other than the UK which have similar health systems, population structures and experiences with the pandemic.

August 2021 update

We have updated this model and used it to forecast the number of COVID-19 cases, hospital admissions and deaths linked to increased movement around the time of the UEFA European Football Championship.

Note

This plain English summary was created with the support and feedback of the EAVE II Patient Advisory Group (PAG). This one in particular was authored by Kamil S.

To learn more about the PAG, see: Our EAVE II Patient Advisory Group (PAG) | The University of Edinburgh