Cocaine analytes in human hair: evaluation of concentration ratios in different cocaine sources, drug-user populations and surface-contaminated specimens.

Hair specimens were analyzed for cocaine (COC), benzoylecgonine (BE), cocaethylene (CE) and norcocaine (NCOC) by liquid chromatography-tandem mass spectrometry. Drug-free hair was contaminated in vitro with COC from different sources with varied COC analyte concentrations. Results were compared to COC analyte concentrations in drug users' hair following self-reported COC use (Street) and in hair from participants in controlled COC administration studies (Clinical) on a closed clinical research unit. Mean ± standard error analyte concentrations in Street drug users' hair were COC 27,889 ± 7,846 (n = 38); BE 8,132 ± 2,523 (n = 38); CE 901 ± 320 (n = 20); NCOC 345 ± 72 pg/mg (n = 32). Mean percentages to COC concentration were BE 29%, CE 3% and NCOC 1%. Concentrations in hair were lower for Clinical participants. COC contamination with higher CE, BE or NCOC content produced significantly higher concentrations (P = 0.0001) of all analytes. CE/COC and NCOC/COC ratios did not improve differentiation of COC use from COC contamination. COC concentrations in illicit and pharmaceutical COC affect concentrations in contaminated hair. Criteria for distinguishing COC use from contamination under realistic concentrations were not significantly improved by adding CE and NCOC criteria to COC cutoff concentration and BE/COC ratio criteria. Current criteria for COC hair testing in many forensic drug-testing laboratories may not effectively discriminate between COC use and environmental COC exposure.