Speciality Trainee in Clinical Oncology Dr Karen Mactier gives an update on some recent evidence.
Evidence of past COVID-19 infection in healthcare workers (HCWs) is becoming a focus of medical and political interest. Here is a summary of the first, medium-size study in the NHS workforce with comments on similar Spanish and German studies.
Shields A et al., SARS-CoV-2-seroconversion in health care workers medRxiv 2020, posted 19/5/20
Paper status: pre-print manuscript, not peer-reviewed.
Background: This paper describes a UK cross-sectional study of asymptomatic HCWs’ current and past exposure to SARS-CoV-2, measured using nasopharyngeal swab real time rt-RNA PCR and ELISA to anti-SARS-COV-2 spike proteins respectively. IgG, IgM and IgA antibody status were all assessed. Five hundred and fifty-four HCWs were recruited across four hospital campuses from south and central Birmingham over a 24 hour period on 25/4/20. Staff members were selected from a range of patient-facing roles after responding to an invitation email. The date chosen was approximately 5 weeks following the UK-wide lockdown and 3 weeks after peak death rate was reached locally.
Results: The point prevalence of PCR positivity was 13/554 (2.39%). Serum was available for 516 participants, 126 (24.4%) had evidence of seroconversion. There was no statistically significant difference in age, ethnicity (white vs. BAME), gender or postcode deprivation score profiles between the groups. One hundred and thirty-six (26.3%) participants reported a prior illness consistent with COVID-19 in the past four months. Participants who reported suggestive symptoms were statistically more likely to be seropostitive (37.5% vs. 17.1%, Χ2 = 21.1034). Seroconversion was higher in participants who reported their symptoms occurring in the week prior to peak death rate (77.8%). Furthermore, symptomatic participants generated quantitatively higher antibody test responses than asymptomatic individuals. The authors reported seroconversion rates by speciality area, with those working in housekeeping (34.5%), acute medicine (33.3%) and general internal medicine (30.3%) most likely to be seropositive; were lower in intensive care (14.8%) and emergency medicine (13.3%).
Discussion: This is the first published UK study of SARS-CoV-2 seroconversion in healthcare workers. There appears to be agreement between past reported symptoms in both the likelihood and magnitude of serology response and this relationship was stronger when symptoms occurred at the peak of the outbreak when the hospital was busier. This is consistent with findings from another European study; Garcia-Basterio published serology data from 578 staff at a Barcelona hospital between 28/3/20 and 9/4/20. Fifty-four (9.3%, 95% CI 7.2-12.20%) of the Spanish group antibodies to SARS-CoV-2 spike proteins and again the seroprevalence was much higher in those with who reported suggestive symptoms (OR 8.84, 95% CI 4.41-1773). However, caution must be applied with self-reported symptoms, The 217 German HCWs in Behrens et al.’s study, all working in ‘high-risk’ clinical areas including ED,ID, respiratory, paeds and ITU reported an average self-perceived likelihood of previous COVID-19 infection of 20.9%, however seroconversion rates were only 1-2%.
The difference in seroconversion rates in the Birmingham study between level 1 areas such as general medicine and ITU are striking, perhaps raising fears regarding availability and compliance with PPE protocols in different clinical areas. However the authors do not comment on whether this is a statistically significant difference, nor do they provide the number of staff participants in each group, so no definitive conclusions can be drawn.
Generalisability to Edinburgh Cancer Centre: Birmingham is clearly a large metropolitan community and has been badly affected by the COVID-19 pandemic in comparison to smaller population centres, such as Edinburgh and the Lothians. The authors did not specify seroconversion rates for staff working in ‘green’ clinical areas, such as oncology, where exposure to COVID-19 is likely to be lower.
Karen Mactier is a Clinical Oncology Registrar currently working in Edinburgh Cancer Centre