Performance of Immunoglobulin G Serology on Finger Prick Capillary Dried Blood Spot Samples to Detect a SARS-CoV-2 Antibody Response

• Published in: Microbiology spectrum Vol. 10; no. 2; p. e0140521
• Main Authors: Nikiforuk, Aidan M, McMillan, Brynn, Bartlett, Sofia R, Márquez, Ana Citlali, Pidduck, Tamara, Kustra, Jesse, Goldfarb, David M, Barakauskas, Vilte, Sinclair, Graham, Patrick, David M, Sadarangani, Manish, Ogilvie, Gina S, Krajden, Mel, Morshed, Muhammad, Sekirov, Inna, Jassem, Agatha N
• Format: Journal Article
• Language: English
• Published: United States American Society for Microbiology 27.04.2022
• Subjects: Adult; Antibodies, Viral; Antibody Formation; Child; Clinical Microbiology; COVID-19 - diagnosis; COVID-19 Vaccines; Cross-Sectional Studies; diagnostic accuracy; dried blood spots; Humans; Immunoglobulin G; public health; Research Article; SARS-CoV-2; Seroepidemiologic Studies; seropositivity; vaccine evaluation; COVID-19; epidemiology; SARS-CoV-2; diagnostic accuracy; data analysis; vaccine evaluation; seropositivity; dried blood spots; public health
• ISSN: 2165-0497; 2165-0497
• DOI: 10.1128/spectrum.01405-21

We investigate the diagnostic accuracy and predictive value of finger prick capillary dried blood spot (DBS) samples tested by a quantitative multiplex anti-immunoglobulin G (IgG) assay to detect severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies after infection or vaccination. This cross-sectional study involved participants (  = 6,841) from several serological surveys conducted in nonhospitalized children and adults throughout 2020 and 2021 in British Columbia (BC), Canada. Analysis used paired DBS and serum samples from a subset of participants (  = 642) prior to vaccination to establish signal thresholds and calculate diagnostic accuracy by logistic regression. Discrimination of the logistic regression model was assessed by receiver operator curve (ROC) analysis in an  = 2,000 bootstrap of the paired sample (  = 642). The model was cross-validated in a subset of vaccinated persons (  = 90). Unpaired DBS samples (  = 6,723) were used to evaluate anti-IgG signal distributions. In comparison to paired serum, DBS samples from an unvaccinated population possessed a sensitivity of 79% (95% confidence interval [95% CI]: 58 to 91%) and specificity of 97% (95% CI: 95 to 98%). ROC analysis found that DBS samples accurately classify SARS-CoV-2 seroconversion at an 88% percent rate (area under the curve [AUC] = 88% [95% CI: 80 to 95%]). In coronavirus disease 2019 (COVID-19) vaccine dose one or two recipients, the sensitivity of DBS testing increased to 97% (95% CI: 83 to 99%) and 100% (95% CI: 88 to 100%). Modeling found that DBS testing possesses a high positive predictive value (98% [95% CI: 97 to 98%]) in a population with 75% seroprevalence. We demonstrate that DBS testing should be considered to reliably detect SARS-CoV-2 seropositivity from natural infection or vaccination. Dried blood spot samples have comparable diagnostic accuracy to serum collected by venipuncture when tested by an electrochemiluminescent assay for antibodies and should be considered to reliably detect seropositivity following SARS-CoV-2 infection and/or vaccination.