Low-sensitivity urine pregnancy testing to assess medical abortion outcome: A systematic review

The objective was to summarize data on the accuracy and acceptability of a strategy for identifying ongoing pregnancy after medical abortion treatment using a low-sensitivity pregnancy test (LSPT).

We searched PubMed to identify studies that evaluated the use of a single posttreatment LSPT for detection of ongoing pregnancy after treatment with mifepristone and misoprostol. We extracted, assessed and summarized data from each study.

We found 10 studies that evaluated 6 LSPTs with human chorionic gonadotropin detection thresholds of 1000, 1500 or 2000 mIU/mL. The three earliest studies compared the pregnancy test strategy to standard assessment in the same women; the sensitivity of a positive or invalid LSPT result for detecting ongoing pregnancy ranged from 67% to 100%. Three randomized trials found no significant difference in detection of ongoing pregnancy between the LSPT strategy and routine in-person follow-up. Four noncomparative studies found that of the 12 women who had ongoing pregnancies diagnosed after performing an LSPT, 8 (67%) had positive or invalid LSPT results. Across the 10 studies, 30 of the 59 total ongoing pregnancies (51%) were identified based on symptoms without a positive or invalid LSPT result. Women expressed satisfaction with the LSPT strategy. Risk of bias in the seven later studies was high.

Despite their limitations, most studies showed that the LSPT strategy had moderate sensitivity for identifying ongoing pregnancy and can enable the majority of medical abortion patients to assess treatment outcome at home. However, the LSPT itself had a limited role in the detection of treatment failures in the studies.

The LSPT strategy shows promise for reducing the need for in-person follow-up after medical abortion. A range of home-based options should be validated to meet the varied needs of women and abortion providers in diverse settings.