Does the standard hospital warning score work for COVID-19?
The NEWS2 score was a weak predictor of severe COVID-19, but adding age and a few routine blood tests improved it across hospitals in the UK, Norway and China.
The question
The National Early Warning Score (NEWS2) is recommended in the UK to judge how sick hospital patients are, including those with COVID-19. But little was known about how well it spots the patients who go on to need intensive care or die.
What we did
We trained models on 1,276 patients admitted to King’s College Hospital, then tested them on 6,237 patients from five other UK NHS trusts, a hospital in Norway and two hospitals in Wuhan. The outcome was transfer to intensive care or death within 14 days of admission.
What we found
NEWS2 plus age had poor-to-moderate discrimination (AUC 0.700 in the training cohort). Adding routine measurements, including supplemental oxygen and blood markers such as urea, C-reactive protein, kidney function and neutrophil count, gave moderate improvement that held up across hospitals, though the model tended to underestimate risk in most sites.
Why it matters
A few common measurements can sharpen an existing score without new tests. But the miscalibration shows that early warning scores need careful local checking for COVID-19.