Clinical utility of urine pregnancy assays to determine medical abortion outcome is limited

Abstract

Background Determining medical abortion outcome commonly includes a costly evaluation such as ultrasonography or serial serum hCG testing. Urine pregnancy testing may represent a less costly alternative. Methods This prospective diagnostic test evaluation study was part of a multisite randomized trial of 1080 women undergoing medical abortion up to 63 days' gestation who returned 1 and 2 weeks after receiving mifepristone. Low-sensitivity (LS) and high-sensitivity (HS) urine pregnancy tests were performed at each visit, and the results were compared to ultrasonography. Sensitivity, specificity, predictive values and likelihood ratios of each urine test were determined. Results In the first week following abortion, 14.8% of the LS tests and 7.9% of the HS tests correctly predicted outcome. None of the LS tests and only 0.2% of the HS tests were falsely negative; however, 85.2% of the LS tests and 91.8% of the HS tests were falsely positive. In the second week following abortion, 39.1% of the LS tests and 33.8% of the HS tests correctly predicted the medical abortion outcome. Only 0.2% of the LS tests and 0.3% of the HS were falsely negative; however, 60.8% of the LS tests and 65.8% of the HS tests were falsely positive. Conclusions Both LS and HS urine pregnancy assays reliably assess clinical outcomes of medical abortions in cases of negative test results. However, the clinical utility of urine assay testing is limited because of the high rate of false-positive results.