A review and meta-analysis evaluates the seroprevalence of SARS-CoV-2 antibodies in healthcare workers (HCWs) around the world. And it seems that about 9% of HCWs have had COVID-19 up to the end of August 2020.

The review included 49 studies and a total of almost 130,000 individual HCWs. Studies were included up to August 2020, so these findings represent findings from the “first wave” of COVID-19 only. The following factors were associated with seropositivity: the sensitivity of the antibody test; male gender; Black, Asian and Hispanic HCWs; work in a COVID-19 unit; patient-related work; front-line HCWs; healthcare assistants; shortage of personal protective equipment; self-reported belief of previous SARS-CoV-2 infection; previous positive polymerase chain reaction test; and household contact with suspected or confirmed cases of COVID-19. I would be interested to see some sub-analyses of the dataset to focus in on healthcare exposures, by taking out those with a known community exposure – and perhaps focussing only on front-line healthcare workers.

It seems notable that HCW in the US had the highest rate of seroprevalence (at 12.7%) compared with, say Europe, where the rate was 8.5%. The CDC has recommended an FFP3 respirator to care for patients with known or suspected COVID-19 in the US, and yet it seems from this review that US HCW were no better protected that those in the rest of the world, where guidelines are more variable (e.g. WHO and PHE recommend a medical/surgical mask unless AGPs when caring for a patient with known or suspected COVID-19, whereas ECDC recommends FFP2/3). (The confidence intervals for the estimate of seroprevalence in North America overlapped those from other parts of the world.) Clearly, the drivers for HCWs to acquire COVID-19 are much more complex than which type of respiratory protection was worn, especially in the context of a pandemic with vastly different rates of community transmission. I fully accept that this study was not planned to answer or comment on this controversial issue, but I think it’s an interesting finding. (For my Canadian readers – I do know that Canada is not part of the US – but the majority of the studies from “North America” were from the US in this study.)

Now that the pandemic has rolled on 6 months or so, it will be interesting to see what seroprevalence in HCWs looks like now. I suspect it will be considerably higher than 8.7%, so we’ll have to wait and see!