VALIDATION OF A RAPID URINE SCREENING ASSAY FOR COCAINE USE AMONG PREGNANT EMERGENCY PATIENTS

Objectives: To determine the accuracy of the OnTrak rapid urine latex agglutination immunoassay (AIA) for testing pregnant ED patients for t he cocaine metabolite benzoylecgonine (BE), and to determine the frequ ency of urine BE in pregnant ED patients and the association of test r esults with patient demographics. Methods: A test performance evaluati on was conducted using an inception cohort of pregnant patients at an urban teaching hospital ED. Patients with positive urine chorionic gon adotropin tests who had adequate remaining urine samples were studied. Patient demographics, ED diagnoses, and assay results were recorded w ithout patient identifiers: Urine was tested using the rapid AIA. The first 150 samples were confirmed using a second immunoassay and gas ch romatography with a nitrogen phosphorus detector. All positive samples also were confirmed with thin-layer chromatography, and quantificatio n by gas chromatography-mass spectrometry. Results: Twenty of 342 (5.9 %, 95% CI 3.4-8.4%) pregnant patients had urine samples positive by AI A testing for BE. Confirmation testing demonstrated no false-positive or -negative AIA result. Positive test results were not associated wit h specific patient diagnoses or demographics. Conclusions: ED screenin g for cocaine use among pregnant patients can be accurately performed using the OnTrak AIA for BE. In the ED population screened, 5.9% of th e pregnant women had urine samples positive for BE.