ABSTRACT
Having diagnostic methods that meet certain attributes is vital to guide health decisions, the current context warrants it. In order to validate the capacity of two rapid tests to detect antibodies due to SARS-CoV-2 infection in Isla de la Juventud, Cuba, from April to May 2020, a descriptive cross-sectional evaluation study was carried out. Rapid tests: Wondfo (SARS-CoV-2 antibody test) and Lungene covid-19 IgG/IgM, were compared to the real-time polymerase chain reaction test. Two homogeneous samples of 250 each were constructed, determining validation indicators. Sensitivity values of 6.6% and 8.3% respectively were obtained for each test, while the specificity was higher for Wondfo (95%). The positive predictive values were very low, the negative ones were adequate, higher in Lungene with 94.8%. Likelihood ratio values were classified as useless. In different scenarios in terms of symptomatic cases, sensitivity of 50% was reached in an interval of 1 to 7 days for Wondfo. The area under the ROC curve for Wondfo was 0.50 (95% CI = 0.46-0.55) and 0.46 for Lungene (95%CI = 0.38-0.55). The kappa index for Wondfo was 0.025 and 0.010 for Lungene. The rapid tests explored showed very low sensitivity, positive predictive value, and inadequate likelihood ratio. The global validity of the tests did not demonstrate a good diagnostic performance, marked by the value of the area under the ROC curve. The degree of agreement was poor. © 2021, Finlay Ediciones. All rights reserved.