Enzyme immunoassay validation for qualitative detection of cocaine in sweat.

A solid-phase enzyme immunoassay (EIA) involving microtiter plates was modified for analysis of cocaine in sweat. Sweat was collected with the PharmChek sweat patch and drugs were eluted from the collection pad of the patch. The sweat contained primarily parent cocaine. The assay was determined to have cross-reactivity for cocaine of 102% relative to 100% for the benzoylecgonine (BE) calibrators and for cocaethylene of 148%. The optimum cutoff concentration for this modified assay, determined by receiver-operating characteristic curve analysis, was 10 micrograms/L cocaine or BE equivalents. At this concentration the assay had 94.5% sensitivity and 99.1% specificity vs gas chromatography-mass spectrometry (GC-MS) as an acceptable indicator of the true clinical state. The positive predictive value at a prevalence of 50% was 99%. Threshold analysis for positives suggested that the 95% confidence interval for a positive result by the EIA was between 12.5 and 15 micrograms/L and that quality-control samples at 5 and 15 micrograms/L could be run with each batch to certify the precision around the cutoff. All positive samples must be confirmed by GC-MS. The sensitivity and specificity of the overall analysis system (immunoassay screen and GC-MS confirmation) was 86% and 97%, with known cocaine dosing of volunteers as the acceptable indicator of the true clinical state.