Comparison of hysterosalpingography, hysterosonography and hysteroscopy in evaluation of the uterine cavity in patients with recurrent pregnancy losses

To evaluate the accuracy and acceptability of hysterosonography (HS) in the diagnosis of uterine anomalies in patients with recurrent pregnancy loss. Sixty non-pregnant patients with a history of at least three previous recurrent miscarriages were submitted to the HS, HSG and hysteroscopy (HTC) exams. The findings fall into three categories: synechiae, polypoid lesions and abnormal uterine cavity shape. The HTC exam was used as the gold standard. The findings of the HS and the HSG were compared to the HTC's and the agreement was evaluated using the Kappa coefficient. Significance was established at < 0.05. Sensitivity, specificity, positive and negative predictive values were determined for each method. Uterine anomalies were present in 38.3% (23) of the patients: 16 (26.7%) with synechiae, 3 (5.0%) with polypoid lesions and 8 (13.3%) with abnormal uterine cavity shape. Accuracy of the HS and the HSG was 90.9 and 85.2%. General sensitivity of the HS was superior to the HSG (90.5 vs. 75.0%). It also had a higher degree of agreement with the HTC (Kappa = 0.81 vs. 0.68). The HS caused significantly less pain than the other two methods. The HS had higher diagnostic accuracy in the detection of uterine cavity anomalies and it was better tolerated by the patients when compared to hysterosalpingography and hysteroscopy.