Fig. 7: Simulation results showing confidence intervals for serosurvey prevalence calculations. | Nature Communications

Fig. 7: Simulation results showing confidence intervals for serosurvey prevalence calculations.

From: Standardization of ELISA protocols for serosurveys of the SARS-CoV-2 pandemic using clinical and at-home blood sampling

Fig. 7

Each graph displays 95% confidence intervals (CI) for the estimate of prevalence from 1000 replications of each condition (including estimating the sensitivity and specificity for each replicate). The CIs are sorted by the lower bound, with lower bounds less than 0 replaced by zero. For all graphs, the true sensitivity is 0.90 and the simulations use 100 samples to estimate the sensitivity. For the graphs in row a the true specificity is 0.99 and in row b the true specificity is 1.00. The graphs show the simulations results with estimates of specificity using sample sizes of 100 (black), 300 (red), 1000 (green). Points are plotted black, red, then green, so some of the black and green points may be covered (e.g., the lower bounds for all three colors are all zeros in the bottom left panel). The columns give results for true prevalence values of 0.001, 0.01, or 0.1. The true prevalence for each simulation is shown by a vertical gray line.

Back to article page