BackgroundWe aimed to define endometrial pattern and endometrial thickness in predicting the outcome of early medical abortion. Study DesignWhile blinded to outcomes of abortion, we retrospectively reviewed the ultrasound scan performed 14–21 days after medical abortion. We assessed the endometrial pattern and endometrial thickness. A total of 943 women at or before 56 days of gestation who underwent medical abortions were included. Abortion was induced with mifepristone (600 mg) orally followed 48 h later with oral misoprostol (600 mcg). A successful medical abortion was defined as complete abortion without surgical intervention. Three sonographic patterns (homogenous, heterogeneous and multilayered) were devised to correlate with the outcome. ResultsOf the 940 women, 92 (9.8%) had failed medical abortions. Eighty-seven (94.6%) patients with failed treatment outcomes had a heterogeneous pattern, while no patients with failed treatments had a multilayered pattern. Based on multivariable logistic regression, women who had an endometrial thickness in the range of 10–15 or >15 mm were more likely to have failed outcomes than those with a thickness <10 mm, with ORs of 3.69 (p=.001) and 8.82 (p<.001). Compared to those with a homogenous pattern, women with a heterogeneous endometrial pattern were more likely to have failed outcomes (OR 4.5, p=.003). In addition, an endometrial thickness >10 mm in combination with a heterogeneous pattern had the highest balanced accuracy in the prediction of failed outcome (81.9%; 95% CI, 77.6–86.3). ConclusionWomen with a multilayered pattern could be reassured that they have successful medical abortion, while those with a heterogeneous pattern and/or endometrium >10 mm may need follow-up. Sonographic endometrial pattern and endometrial thickness may serve as objective criteria in the management of early medical abortions.