Cocaine use during pregnancy: sensitive detection by hair assay

This paper compares the sensitivity of two tests of cocaine use, interview and urine test, with that of a radioimmunoassay of hair. Interviews and hair samples were provided by 397 obstetric patients in one New York City hospital; urine samples were obtained in 377. Of these patients, 241 were receiving prenatal care (were registered) and 156 delivered without prenatal care (were unregistered). The 241 registered patients were derived from 400, comprising all reporting use of cocaine ever ("ever-users"), all reporting use by the father but not themselves ("lifestylers"), and a sample of women who were neither ever-users nor lifestylers. The 156 unregistered patients were derived from 352 women interviewed at delivery, unselected for reported use. Thirty-two percent reported ever using cocaine, 45% of these within 6 months before interview. Urine tests were positive in 20%, hair tests in 59%. The estimated sensitivity of the hair test (92%) was 3.1 times higher than that of the urine test (95% confidence interval [CI] = 2.5, 3.8) and 4 times higher than that of reported use in the past 6 months (95% CI = 3.2, 5.0). Self-report and urine tests alone miss most of cocaine users during pregnancy. Hair tests greatly improve detection and thus can enhance evaluations of the effects of prenatal cocaine use on fetal and child development.