Applying SARS-CoV-2 serological testing to understand the effectiveness of local infection control measures for COVID-19 infection in Victoria, Australia 2020

The availability of early serological assays for COVID-19 infection allows investigation of aspects of SARS-CoV-2 transmission which may not be evident on PCR testing. Here we describe two serological surveys in different settings in 2020 which assess the effectiveness of local infection prevention practices: i) a retrospective cohort of household contacts previously quarantined; and ii) health care workers in a tertiary hospital. Serological testing of household contacts did not diagnose any additional cases to that which had been notified by testing symptomatic contacts with PCR. The secondary household attack rate for these 53 contacts from 42 households was 36.2% (95% CI: 37.4, 74.5%). Sero-positivity in health care workers increased from 0.5% in May 2020 to 3.8-5.5% in August to October 2020. Multivariate analysis demonstrated that male HCW were less likely to be sero-positive (OR 0.19 95% CI 0.03-0.7, p=0.04), while there was an increased risk of sero-positivity associated with nursing staff (OR 8.1 95% CI 1.7-145.6, p=0.04), staff who experience anosmia (OR 22.7, 95% CI 5.4-112.4, p < 0.001); and those caring for COVID-19 patients (OR 4.1 95% CI 1.8-10.5, p=0.01). In this low prevalence setting with rapid access to PCR testing, serological testing provided a small additional benefit to PCR testing alone. Measures to detect COVID-19 cases in household contacts were found to be sufficient, while infection control measures during a period with a high burden of cases in a tertiary hospital were found to be inadequate to prevent COVID-19 infections in health care workers.