This is a summary, written by members of the CITF Secretariat, of:

Tuite AR, Fisman D, Abe KT, Rathod B, Pasculescu A, Colwill K, Gingras AC, Yi QL, O’Brien SF, Drews SJ. Estimating SARS-CoV-2 Seroprevalence in Canadian Blood Donors, April 2020 to March 2021: Improving Accuracy with Multiple Assays. Microbiol Spectr. 2022 Feb 23. doi: 10.1128/spectrum.02563-21

The results and/or conclusions contained in the research do not necessarily reflect the views of all CITF members.

Using various serological assays in parallel on the same set of samples can produce inconsistent results. A solution is to use composite reference standards and latent class analysis to help decipher true seropositivity. Using these two analytical methods, Drs. Anne-Claude Gingras, from the University of Toronto, and Sheila O’Brien and Steven Drews from Canadian Blood Services, estimated true seropositivity among nearly 18,000 blood donors from April 2020 to March 2021 in a paper published in Microbiology Spectrum. Both methods produced similar estimates and confirmed that Canadian seroprevalence was below 10% as of March 2021.

The problem: Because different serological assays can produce different results even on the same sample, it is very difficult to determine a true measure of seropositivity (whether due to prior infection and/or vaccination). Moreover, the issue is further complicated as there is currently no gold standard among assays.

The solution: Composite reference standards and latent class analysis (LCA) are two mathematical models applied to test results to help decipher true seropositivity. For the former, they used the assumption that seropositivity on two or more assays represented a true positive case. The latter estimated true seropositivity using a model fed with historical data of factors associated with seropositivity gathered from multiple tests. Samples from blood donors were all processed on four different assays detecting IgG against components of SARS-CoV-2 and results were used in the modelling.

Key points:

  • Using the composite reference standard method, 2.1% of tested samples were deemed ‘seropositive’ over the study period. Seropositivity rose throughout the study period, from 0.5% in April 2020 to 5.7% in March 2021.
  • Using the LCA model, 2.5% of the samples were deemed ‘seropositive’ over the study period. In this analysis, seropositivity rose from 0.8% in April 2020 to 6.3% in March 2021.
  • Still using the LCA model, donors belonging to a racialized group were more likely to be seropositive compared to self-identified white donors. Additionally, the Atlantic provinces had more samples that were seronegative than any other province.

Findings were based on 17,999 unique Canadian Blood Services donor samples.