This prospective cohort study was conducted at Zeynep Kamil Education and Training Hospital, Istanbul, Turkey, and included 89 premenopausal women. All participants were examined first by TVS, further investigated with SIS and HS, and finally dilatation and curettage was performed when needed. The results obtained from these three methods were compared with the pathologic diagnoses. The positive and negative likelihood ratios (LR+ and LR鈭? of TVS, SIS and HS were calculated by comparison with the final pathological diagnosis. In addition, area under the curve (AUC) values were also calculated.
Polypoid lesion was the most common abnormal pathology. LR+ and LR鈭?of TVS, SIS, and HS were 3.13 and 0.15, 9.83 and 0.07, 13.7 and 0.02 respectively in detection of any abnormal pathology, and the AUCs of TVS, SIS, and HS were 0.804, 0.920, and 0.954 respectively. When the three procedures were compared with each other separately, HS had the best diagnostic accuracy, and the diagnostic accuracy of HS and SIS was superior to TVS (p1 = 0.000, p2 = 0.000). For the detection of polypoid lesions, HS was the most accurate diagnostic procedure (AUC = 0.947), followed by SIS (AUC = 0.894) and TVS (AUC = 0.778).
HS provides the most accurate diagnosis and allows treatment in the same session in premenopausal women with abnormal uterine bleeding.