Hospital-onset COVID-19 (HOCI): a systematic review

As we get into gear to prepare for the next epidemic wave of COVID-19 affecting healthcare providers (hoping that it will not come), it’s a good time to review where we have got to with the surveillance of healthcare-associated COVID-19. Colleagues at Imperial have just published a systematic review of the latest literature on this important issue. Whilst uncertainties remain about surveillance definitions and exactly what “healthcare-associated COVID-19 infection” is, the message is clear that healthcare-providers must have in place clear and rapid systems for identifying healthcare-associated COVID-19 to prevent the spread of the SARS-CoV-2 virus in healthcare facilities.

First, a word on terminology. These are what the following terms mean to me…

Hospital-onset COVID-19 infection (HOCI) is COVID-19 that is identified from a patient in hospital.
Healthcare-associated COVID-19 and hospital-associated COVID-19 is COVID-19 that could have been acquired during healthcare / in hospital.
Healthcare-acquired COVID-19 and hospital-acquired COVID-19 is COVID-19 was probably acquired during healthcare / in hospital. Hospital-acquired COVID-19 is synonymous with nosocomial COVID-19.

During the “first epidemic wave” of COVID-19 in the UK in March/April 2020, we were horribly under-prepared for what was to come. This resulted in in-hospital transmission of COVID-19. We wrote an article on this last year, outlining some of the challenges of developing a meaningful HOCI surveillance system. At this time we were only testing for symptomatic COVID-19, and so the prevalence of SARS-CoV-2 infection in our hospitals was probably greater than we knew, as it was in the community.

Over the summer of 2020, we had the opportunity to develop and implement more robust systems for the identification and prevention of healthcare-associated COVID-19. And so when the “second epidemic wave” of COVID-19 struck in November/December 2020, we were better prepared to respond. Nobody was able to prevent the in-hospital transmission of SARS-CoV-2, but at least we have rapid surveillance systems so that we could respond and prevent further transmission.

This systematic review identified nine articles that met the inclusion criteria. The review highlighted several important challenges re HOCI surveillance:

How to manage the long possible incubation period. As we know, it can take up to 14 days from exposure to SARS-CoV-2 to the point whether symptomatic or asymptomatic infection can be identified. This creates a real challenge in terms of surveillance definitions and tying down HOCI. Figure 1 (below) shows a schematic of how various different surveillance systems have managed this challenge. As you can see, the terminology is subtly but significantly different in all of the surveillance systems, and this is a challenge, because we can be talking about the same thing but with different terminology and risk getting into a bit of a mess! In practice, I have come to a much more pragmatic view of life to consider COVID-19 in the first 7 days of admission unlikely to be hospital-acquired and >7 days likely to be hospital-acquired.
What do we do with staff? 4/9 surveillance systems made some attempt to include staff in their scheme. In the context of this pandemic, staff (and any other human being in the vicinity – especially visitors) will play a key role in hospital transmission.
Real-timeliness. A HOCI surveillance system that tells you in retrospect how much hospital transmission of SARS-CoV-2 happened has some value, but isn’t helpful in practical terms. A HOCI surveillance system needs to be as close to real-time as possible to prevent healthcare-associated COVID-19.
Implementation. The review included an analysis of how well the system had been implemented and interfaced outside of a single provider context.
Incidence. Six studies reported on the incidence of healthcare-associated COVID-19, which was in the 0-15% range.