Study population characteristics (n=119).

Background
Dengue virus (DENV) seroprevalence studies often rely on Enzyme-Linked Immunosorbent Assay (ELISA) testing of serum samples, but ELISA testing of dried blood spot (DBS) samples offer several advantages for field-based research in resource-limited settings. However, DBS’ limited sample volume can be challenging for test sensitivity, requiring validation studies with standard methods (e.g., analysis of serum through ELISAs or Plaque Reduction Neutralization Tests (PRNTs)). In preparation for a large cluster randomized controlled trial, we conducted a pilot study in 2019 to validate the use of DBS compared to serum samples for DENV IgG testing. We aimed to identify the optimal DBS dilution for IgG detection and to estimate the correlation, magnitude of agreement, and sensitivity and specificity of IgG detection in DBS versus serum samples.
Methodology/Principal Findings
We conducted this pilot validation study among 119 healthy participants in Fortaleza, Brazil to evaluate and optimize the detection of DENV IgG from DBS compared to serum. Each participant provided paired DBS and venous blood samples, which were evaluated for DENV IgG using the Panbio Dengue IgG indirect ELISA. DBS elution diluted 1:4 was optimal compared with serum results, with high correlation (r= 0.98) and near-perfect agreement (kappa = 0.95). At this dilution, DBS had a sensitivity of 100%, a specificity of 92.3%, a 97.9% positive predictive value, and a 100% negative predictive value compared with serum.
Conclusions/Significance
These results validate using DBS instead of serum for detection of prior dengue infection among similar populations in endemic regions, without sacrificing test sensitivity and specificity. The validity of using DBS for ELISA to detect prior dengue infection could have important implications for field-based research. A limitation to this study was that the potential for misclassification due to cross-reactivity (e.g., with Zika virus, Yellow Fever vaccine) was not assessed.