Endometrial thickness following medical abortion is not predictive of subsequent surgical intervention

Objectives. To evaluate the ability of endometrial thickness after medical abortion to predict the need for subsequent dilatation and curettage (D&C). Methods. We pooled data from two multicenter medical abortion trials involving 2208 women who received mifepristone orally followed by misoprostol vaginally. Women returned for transvaginal ultrasonography approximately 7 days later. The endometrial thickness was measured if no gestational sac was present. Final status was confirmed by a phone interview at 5 weeks. The area under the receiver-operating characteristics (ROC) curve was calculated to assess the overall ability of endometrial thickness to predict the need for subsequent D&C. Endometrial thickness was dichotomized using threshold values at 5-mm increments from 10 to 30 mm. The sensitivity, specificity, negative predictive value and positive predictive value were calculated to evaluate the ability of each endometrial thickness threshold value to predict subsequent D&C. Multivariable regression analysis was performed to adjust endometrial thickness values for study, treatment group, and study site. Results. At 7 days after misoprostol treatment, 1870 women (84.7%) had endometrial thickness assessed. Thirty of these women (1.6%) subsequently underwent D&C. The mean endometrial thickness was 14.5 mm for women who underwent D&C and 10.9 mm for those who did not (difference 3.5 mm (95% CI, 1.8-5.3 mm)). Endometrial thickness was poorly predictive of the need for D&C, with an area under the ROC curve of 0.65. All endometrial thickness thresholds had positive predictive values of 25% or less. The results were unchanged by adjustment of endometrial thickness values by multivariable modeling. Conclusions. Although endometrial thickness following successful expulsion of the gestational sac is thicker in women who will eventually require surgical intervention after medical abortion, endometrial thickness is not a clinically useful predictor of the subsequent need for D&C.