Leeds-based Emis said the observational analysis of 2.6 million healthcare records – the largest population study of Covid-19 in children to date – suggests a link between ethnicity and Covid-19, with important implications for global public health strategies to combat the virus, such as access to tests.

Data for the analysis was obtained from the QResearch database of anonymised health records from 1,300 GP practices across England and linked to Covid-19 test results and hospital admissions data. The research was funded by the Medical Research Council.

In the study, published in JAMA Paediatrics, the research team from the Universities of Oxford, Leicester, Nottingham, Cambridge and Southampton analysed a nationally-representative sample of children’s electronic healthcare records from the QResearch database to understand whether the established link between ethnicity and Covid-19 in adults was similar in children.

Compared with white children, the odds of a positive test were higher in children from Asian (1.8 times more likely), Black (1.12 times more likely) and mixed/other ethnicity (1.14 times more likely) backgrounds.

Asian children were 1.62 times more likely to be admitted to hospital with confirmed Covid-19 than white children, while hospital stays in Black, mixed race and children from other ethnicities were around 36 hours longer than in white children. There was one death from Covid-19 in the study cohort.

Dr Shaun O’Hanlon, chief medical officer at Emis, said: “We’re delighted to support the research programme to enable fast and meaningful analysis on Covid-19.

“I’d like to thank all of the GP practices using Emis systems that have opted into the QResearch programme to enable ground-breaking research like this.

“Building up evidence of trends and patterns through research means collectively as a healthcare industry we are better placed to make decisions that positively impact patient outcomes on a large scale.”

The testing across different races and ethnicities in this study supports similar findings from the US, providing a clearer picture of inequity in healthcare access across the two nations.

The researchers accounted for key demographic factors such as age, sex, geography, deprivation, household size and underlying health conditions. They caution that because the study was observational, certain biases could not be ruled out, which should be considered before drawing any further conclusions.

Oxford University researchers also used QResearch to create a model to predict which factors would place patients at higher risk if they caught coronavirus.

QCovidResearch works out a patient’s risk of hospitalisation and death. This information became invaluable when prioritising patient groups for vaccinations.

As a result of this model, in February 2021, 1.5 million people were added to the Shielded Patient List, ensuring the most vulnerable were prioritised for the vaccine. The research that created the QCovid algorithm was published in the BMJ.