29 June 2021

Latest study by Medical School Clinical Professor of Cardiology Dinos Missouris et al. published in QJM: An International Journal of Medicine earlier this month, concludes that a machine learning model using routine laboratory parameters can detect atypical and asymptomatic presentations of Covid-19 and might be an adjunct to existing screening measures.

Professor Missouris commented that ‘machine learning is a promising method by which to identify cases of Covid-19 in a hospital population.  The aim of our study was to train and validate a machine learning model capable of differentiating Covid-19 positive from negative patients using routine blood tests and to assess the model’s accuracy against atypical and asymptomatic presentations.  The results produced high percentages of accuracy, sensitivity and specificity.  The rapid and early detection of Covid-19 facilitates timely intervention and prevention of transmission’.

Professor Missouris concluded that ‘further appraisal should be carried out before such a model can be approved for clinical use, and if sanctioned, it should be used as an adjunctive tool to clinical assessment’.

The model is available for use and evaluation on request.

Access the study here