Up to 40% of first wave COVID-19 Patients May Have Acquired Infection in Hospital, says SAGE Paper

Yesterday the Scientific Advisory Group for Emergencies released a document which provides an estimate of the number of COVID-19 hospital patients who were infected through nosocomial transmission between February and July of last year. Turns out the number may be as high as four in 10. The Telegraph has the story.

It was previously thought around one-fifth of infections had been caught in hospitals, but researchers at Public Health England and the London School of Hygiene and Tropical Medicine (LSHTM) also looked at the number of people testing positive within two weeks of discharge.

Under the most conservative estimate, which only included people infected after at least 15 days in hospital, just 8.8% of infections were found to be “nosocomial”, or acquired in hospital, equating to 7,906 people.

However, under the least conservative estimates – which included those testing positive within three days of entering hospital or 14 days after discharge – the number rises to 40.5% of hospital infections, a total of 36,152 people.

Separate modelling by the same group using different data from the COVID-19 Clinical Information Network concludes that there may have been 31,070 hospital infections in England over the first wave.

Not only that. This is from the document itself:

A number of the nosocomially infected patients will be discharged and lead to onward community transmission. Modelling this onward transmission crudely… up to the 4th generation of transmission, gives 50,550 onward infections. A proportion of these infections will progress to disease and subsequently be hospitalised. Note that onward transmission in the community from infectious staff is not included.

Adding only those hospital-linked community cases who end up being hospitalised into the estimation of the contribution of nosocomial infections to the first wave gives an estimate of 31.4% of infections in hospital that are nosocomial or due to transmission from a nosocomial case

Going on to discuss the impact in England overall the paper says:

Whilst the proportion of hospitalised COVID-19 cases that are linked to hospital transmission is considerable, at the population level, the proportion is relatively small. A simple calculation assuming 5% of infections are hospitalised and of these hospital cases, if 25% are due to nosocomial infection, the complete prevention of nosocomial transmission would have led to approximately 1% impact on the number of infections in the English epidemic overall. However, since hospitalised patients tend to be old and/or frail, the impact in terms of morbidity and mortality would nonetheless be expected to be substantial.

An exploration of the counterfactual (i.e. numbers of hospital admissions of COVID-19 cases in the absence of nosocomial transmission) shows little impact of nosocomial transmission initially, but in the last quarter of the first wave the impact of nosocomial transmission may have been to prolong the epidemic (potentially by several weeks)

Worth reading in full.

The paper is heavily caveated, of course, stating at the top that it makes some large assumptions about cases during a period of limited testing. But these are numbers which, according to the the Telegraph, the Government was reluctant to disclose:

SAGE minutes show that experts have been aware of the paper since October, but the Government has repeatedly refused to release it until now and rejected a Freedom of Information request from the Telegraph to see the modelling.

The paper supports the case that lockdown sceptics have been making for some time, namely, that a key problem to be tackled in respect of COVID-19 is transmission in hospitals and care homes, and is not addressed by stay-at-home orders or mask mandates. The Telegraph quotes Paul Hunter, Professor in Medicine at the University of East Anglia.

“This is not that surprising. As a country, we did not manage the first wave of the epidemic at all well, especially as regards the protection of our most vulnerable individuals and our healthcare and care workers.

“The exact proportion of people with a nosocomial depends in part on how you define that. The 40% figure is probably too high, but maybe not by that much. But a high risk of hospital-associated infection is an expected outcome of a situation where we were not able to provide adequate supplies of PPE to healthcare workers.”

Worth reading in full.

https://lockdownsceptics.org/