Hello everyone, I hope you can help me with the statistical interpretation of Bland-Altman plots. We validated a diagnostic test for the IgM-NS1 dengue-specific ELISA assay (In-house) against three commercial ELISA kits for IgM dengue-specific (EUROIMMUNE, VIRION, PAN BIO). We conducted a comparative analysis based on the ROC curve, but the reviewers have asked us to present the comparative analysis using Bland-Altman plots (Figure 1: Euroimmune and In-house, Figure 2: Virion and In-house, Figure 3: Pan Bio and In-house). Unfortunately, I lack sufficient knowledge about this type of analysis, and we don't have a statistical expert on our team. From the images, we observed a skew in all three analyses. However, we would like to argue to the reviewers that the measurement values in the commercial ELISA kits differ from our in-house method because the former employs arbitrary values to determine positive or negative results, whereas the latter utilizes absolute optical density. Therefore, we suggest that this type of analysis should not be used in our comparative methods analysis. What are your thoughts on the graphics? And what suggestions do you have for our next steps?

More Sebastián Castro-Trujillo's questions See All
Similar questions and discussions