NUMERICAL ANALYSIS OF AN IMMUNOCHROMATOGRAPHIC TEST STRIP READER IN ABUSED DRUGS SCREENING

Abstract

A point-of-care immunoassay strip reader, Uniscan™, was applied to detect methamphetamine, opiate, and marijuana in human urine by providing numerical apparent drug concentrations. Calibration curves were determined by a nonlinear regression. The cutoff was verified using spiked controls. Clinical samples were analyzed and compared with enzyme multiplied immunoassay technique (EMIT). The discrepant results were confirmed by gas chromatography–mass spectrometry (GC-MS). The impacts of interference and cross-reactivity were determined for numerous compounds. The coefficients of the calibration curves had a high correlation coefficient. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and total recovery all had high values for spiked controls. For the 19 discrepant results of clinical samples, GC-MS confirmed that Uniscan and EMIT were correct for 11 and eight samples, respectively. For both methamphetamine and opiate, Uniscan had a lower false positive rate, a higher true negative rate, and a higher total recovery rate than EMIT. For marijuana, Uniscan had a higher true positive rate and a lower false negative rate than EMIT. The Uniscan performed excellently when compared to EMIT. It is advantageous for Uniscan to interpret the test result based on digital read-out, rather than subjective visual judgment.