Development and statistical assessment of a paper-based immunoassay for detection of tumor markers.

Paper-based assays are an attractive low-cost option for clinical chemistry testing, due to characteristics such as short time of analysis, low consumption of samples and reagents, and high portability of assays. However, little attention has been given to the evaluation of the performance of these simple tests, which should include the use of a statistical approach to define the choice of best cut-off value for the test. The choice of the cut-off value impacts on the sensitivity and specificity of the bioassay. Here, we developed a paper-based immunoassay for the detection of the carcinoembryonic antigen (CEA) and performed a statistical assessment to establish the assay's cut-off value using the Youden's J index (68.28 A.U.), what allowed for a gain in sensibility (0.86) and specificity (1.0). We also discuss about the importance of defining a gray zone as a safety margin for test (±12% over the cut-off value), eliminating all false positives and false negatives outcomes and avoiding misleading results. The test accuracy was calculated as the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, presenting a value of 0.97, what classifies this test as highly accurate. We propose here a low-cost method capable of detecting carcinoembryonic antigen (CEA) in human serum samples, highlighting the importance of statistical tools to evaluate a new low-cost diagnostic method.