Comparison of the sensitivity and specificity of six immunoassays for the detection of amphetamines in urine.

We analyzed 225 urine samples with FPIA (Abbott Amphetamine/Methamphetamine II on ADx and AxSYM), EMIT (Emit II Plus Monoclonal Amphetamine/Metamphetamine assay and EMIT II Plus Amphetamines assay, EMIT N), and KIMS (standard protocol and MDMA protocol, KIMS and KIMS X, respectively) immunoassays and compared their sensitivity and specificity. All assays were calibrated and used semi-quantitatively. All samples that screened positive by any amphetamine screening method and 15% of the negative samples were confirmed by liquid chromatography-tandem mass spectrometry (LC-MS-MS). A sample was considered positive for amphetamines if amphetamine, methamphetamine, methylenedioxymethamphetamine, methylenedio-xyethylamphetamine, or methylenedioxyamphetamine was present at 250 ng/mL. Ninety (40%) of the samples were positive by LC-MS-MS. The areas under the receiver operating characteristic curve varied between 0.972 (KIMS X) and 1.000 (ADx). The optimal cut-off concentrations varied between 271 ng/mL (EMIT N) and 723 ng/mL (AxSYM). The sensitivity was 100% for ADx and between 93 and 95% for the other assays. The specificity varied between 88% (KIMS) and 100% (EMIT N). Use of a 500 ng/mL screening cut-off would have resulted in identical or very similar results for ADx and KIMS X and large increases in the false positives for AxSYM and EMIT and the false negatives for EMIT N and KIMS.