Drug screen technique determines prevalence of cocaine use among pregnant women

We hypothesized that 1) using the traditional screening methods of history and enzyme-multiplied immunoassay technique (EMIT) urine assay, the incidence of prenatal cocaine use in obstetric patients enrolled at the first prenatal visit would be lower than published reports, and 2) using gas chromatography/mass spectrometry (GC/MS) as a more sensitive detection technique would increase identification of cocaine users. Two hundred women were consecutively enrolled at their first prenatal visit. A drug history questionnaire and urine sample were obtained and tested by both EMIT and GC/MS for cocaine metabolites. By standard methods of maternal history and urine EMIT, four of 200 women (2.0% used cocaine during pregnancy. By maternal history and GC/MS of urine, 11 of 200 women (5.5% had evidence of cocaine use, a 175% increase in detection. By the McNemar's test with correction for continuity, GC/MS was more likely to detect cocaine exposure than EMIT (P = 0.014). Two women were detected by history alone with negative urine toxocology by either technique. Therefore, the drug screen technique determined the prevalence of cocaine use. With the standard insensitive EMIT assay coupled with history, there is high potential for misclassification of cocaine using mothers and infants as drug-free controls, thus underestimating the overall incidence of cocaine exposure and underidentifying newborns at risk for adverse effects of cocaine. GC/MS improved case identification by 175%; yet it, too, is limited in the detection of the sporadic user of cocaine.