Thousands of patients are likely to have been infected with coronavirus in UK hospitals, a study suggests.
The King's College London study of 10 UK hospital sites plus one in Italy found at least one in eight patients who had received hospital treatment for coronavirus had caught it on-site.
Researchers said that was a relatively low rate and showed there was effective infection control in place.
They analysed data on 1,500 cases to 28 April, covering the peak in the UK.
Lead author Dr Ben Carter said: "The majority of these patients had already been in hospital for a long time.
"They were older, frailer and had pre-existing health conditions."
But they also had better outcomes than those admitted with the virus, probably because of faster diagnosis and, therefore, treatment.
Only those who tested positive 15 days or more after admission were counted as hospital-acquired infections, however.
And if patients who tested positive after five to 14 days are included, the proportion increases to 23%.
But, because of the long incubation period of the virus, it is impossible to be sure how many of these patients would have been infected in hospital.
Prof Duncan Young, an expert in intensive-care medicine at Oxford University, said the study would also not have captured those infected during a short hospital stay, as patients were not followed up after discharge.
And of course the study only looked at people who were being treated in hospital for coronavirus, so did not include the vast majority of people who caught the disease and either showed no illness or just recovered at home with no need for treatment.
Those caveats do suggest the the risk of catching the virus in hospital remains still small.
Follow Nick on Twitter
Have you or a family member caught coronavirus after visiting hospital, or elsewhere? Share your experiences by emailing email@example.com.
Please include a contact number if you are willing to speak to a BBC journalist. You can also get in touch in the following ways: