select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to a study looking at hospital-acquired COVID-19 infections

A study, published in the Journal of Hospital Infection, looked at hospital-acquired COVID-19 infections.


Prof Tim Cook, Honorary Professor in Anaesthesia at the University of Bristol, said

“The paper tells us 1 in 8 (12.5%) patients in hospital with COVID-19 acquired it in the hospital (while accepting this may be as high as 23%).

“It is important to note this is different from stating that 1 in 8 of all hospitalised patients get COVID while in hospital – we’d need to know how many patents in hospital didn’t get COVID-19 to provide a risk of getting infected while having been admitted to hospital for other reasons. With somewhere between 196 and 365 patients having acquired the illness in 8 hospitals over 60 days (total 480 days) this is a rate of slightly less than one infection every day or two.

“The interpretation that patients diagnosed in hospital do better than those who acquire the illness in the community needs to be considered very carefully. COVID-19 is a mild disease for many, and for some there are no symptoms at all. The patients with mild disease do not get admitted to hospital.

“The patients who have community-acquired COVID-19 all appear to have been admitted to hospital because of the illness – and as only around 1 in 10 of all COVID-19 patients are admitted to hospital they fall into a group most severely affected by the disease. The group who acquired COVID-19 while in hospital for some other reason could have had any range of severity of illness from asymptomatic through mild to severe and we’d only expect about 1 in 10 to have severe disease.

“So it would be altogether surprising if the mortality rates were similar between these two very different groups of patients – the hospital-acquired group should have much better outcomes. The fact their outcomes were only modestly better is therefore perhaps a concern.

“Finally, as the study used a definition of hospital-acquired infection as one that occurred 15 days after admission, it will only capture patients who were already staying a long time in hospital, and these are more likely to be older, frailer and sicker than average hospital patients. The conservative definition will mean patients who acquired the infection in a shorter period will have been missed, as will patients who became infected but then left hospital before two weeks of admission.

“Taken together the results are not entirely reassuring but are also difficult to interpret. They inform us that during the peak of the pandemic a significant proportion of cases of COVID-19 in hospitalised patients were acquired in hospital – but they do not tell us how often this happens, or how many infections actually occur. Those patients who acquire hospital-acquired COVID may be a disproportionately vulnerable group, but this cannot be stated with certainty as it may be due to study definitions. The lower mortality in the hospital-acquired group may have several explanations but an important one is that many in this group will have had mild COVID-19, whereas all those hospitalised with community-acquired COVID-19 had, by definition, severe disease.” 


Prof Duncan Young, Professor of Intensive Care Medicine, University of Oxford, said: 

Normally nosocomial infections (hospital acquired infections, HAI) are defined as new infections occurring 48 hours or more after admission to a healthcare facility. However, this definition is usually used for bacterial infections which have a short incubation time. As the SARS-CoV virus has a long and probably variable incubation time, this definition is hard to apply and I suspect this is why the more conservative 14+ day threshold was used by the study team.

“The team appears not to have followed patients after discharge to catch patients who got infected during a short hospital stay (<14 days) but developed the disease symptoms later after discharge from hospital. Thus the population they studied had developed hospital-acquired COVID disease whilst still an inpatient and so had a greater than two week hospital stay with a median of a month.  They were therefore studying nosocomial COVID infections in a very long-stay hospital population, not all hospital admissions. Patients with 14+ day stays are very likely to be old, frail, and have chronic diseases because otherwise they would have been treated and discharged much earlier. The result cited in the press release that “the group that became infected whilst in hospital were older, more frail and more likely to have pre-existing health conditions compared to the group who were admitted to hospital for COVID-19 infection” may be simply a result of defining a nosocomial infection as one occurring in long-stay inpatients.

“Even at a less detailed level you might expect patients in hospital who get COVID in hospital to have more underlying diseases than those admitted with COVID. The inpatients were already in hospital for a non-COVID disease, so naturally they had more underlying disease than previously fit individuals admitted with COVID alone.

“One should be careful about interpreting the risk of contracting COVID in hospital as 1 in 8. The 1 in 8 result is the proportion of all hospitalised COVID cases that were diagnosed after hospital stays of 14 days or greater. The true risk of catching COVID in hospital would be the number of hospital-acquired COVID cases as a proportion of all admissions (i.e. 197 in all the patients admitted to the seven hospital groups in the study over about three months), which would be very considerably smaller.

“There was no mortality difference between COVID cases from the community and nosocomial cases. However, I am not sure this is a valid comparison. Only severe community acquired COVID cases would have been hospitalised, whereas any already hospitalised patient with even mild COVID symptoms would be included in the study. Thus the community-acquired cases likely had, on average, much more severe COVID disease. We also do not know if the patients with nosocomial infections who died did so as a result of their COVID infection or of the underlying disease that was keeping them in hospital for so long, whilst it is likely nearly all of the community acquired COVID cases died of COVID.”



‘Nosocomial COVID-19 infection: examining the risk of mortality. The COPE-Nosocomial study (COVID in Older PEople)’ by Carter et al. was published in the Journal of Hospital Infection at 04:00 UK time on Tuesday 25th August.

DOI: 10.1016/j.jhin.2020.07.013


All our previous output on this subject can be seen at this weblink:


Declared interests

None received.


in this section

filter RoundUps by year

search by tag