Correlation between self-reported cocaine use and urine toxicology in an inner-city prenatal population.

To determine the prevalence of recent cocaine use and the accuracy of self-reported use, the results of a urine assay for the major cocaine metabolite benzoylecgonine were compared with self-reported cocaine use in an inner-city prenatal population offered routine voluntary urine toxicology screening at the time of registration for prenatal care. During a 1-year period, 6866 women registered for prenatal care and 5200 (76%) consented to urine assays for cocaine metabolites. Of the women consenting to urine assays, 253 (5%) had positive assays for benzoylecgonine. Women with positive assays were significantly more likely than those with negative assays to be older (mean [SD] 27 [5] versus 23 [6] years), black, single, and unemployed. In addition, women with positive assays were significantly more likely to be multiparous, report > two sexual partners in the previous year, and acknowledge a history of a sexually transmitted disease (STD). Forty-seven percent of women with positive assays acknowledged cocaine use in the 6 months prior to sampling. Women with positive assays who denied cocaine use were significantly more likely than those who admitted use to be younger (mean [SD] 26 [5] versus 28 [4] years), to report > or = two sexual partners in the past year, and acknowledge a history of an STD. This analysis revealed a poor correlation between self-reported cocaine use and the results of urine assays for cocaine metabolites among women seeking prenatal care in an inner-city institution.