[Show abstract][Hide abstract] ABSTRACT: The proper characterization of a commercial qualitative method for determining aflatoxin B1 in some nuts is described. A qualitative method that provides binary responses of the yes/no type means that the performance parameters have been properly adapted and defined. Performance characteristics such as the cut-off limit, the detection limit, sensitivity, specificity, false-positive and negative rates, and the unreliability or uncertainty region are defined and then estimated by means of the performance characteristic curves. The commercial test kit showed the cut-off limit at 1.6 ng/g, with a sensitivity rate of 95% and a false-negative rate of zero. A modification can be performed to shift the cut-off to 2.0 ng/g, keeping the same values for the sensitivity and false-negative rate.
No preview · Article · Mar 2004 · Journal of AOAC International