Reproducibility of three-dimensional ultrasound endometrial volume measurements in patients with postmenopausal bleeding

Objective To evaluate the reproducibility of transvaginal three-dimensional (3D) endometrial volume measurement in patients with postmenopausal bleeding and to compare the reproducibility of this technique to that of two-dimensional (2D) endometrial thickness measurement. Methods In a prospective, blinded study, transvaginal ultrasound examinations were performed in 51 consecutive patients with postmenopausal bleeding. Three-dimensional volume and 2D thickness measurements were made and intraobserver and interobserver reproducibility of each technique were assessed. Results The intraobserver correlation of 3D endometrial volume measurement of the first observer was 0.97 and that of the second observer was 0.99. Thus, mean intraobserver correlation was 0.98. The mean interobserver correlation was 0.95 (0.95 vs. 0.96). There was no significant difference in reproducibility at different volume cut-offs. The mean intra- and interobserver correlation of endometrium volume measurements for five patients with endometrial carcinoma did not differ significantly front that for Patients without carcinoma (0.98, 0.98 vs. 0.98, 0.95). The intraobserver correlation of 2D endometrial thickness measurements front the first observer was 0.71 and that front second observer 0.87. Thus, mean intraobserver correlation of the endometrial thickness measurements was 0.79. The mean interobserver correlation was 0.76 (0.84 vs. 0.70). Conclusions Endometrial volume and thickness measurements by 3D and 2D ultrasound, respectively, show good reproducibility but the reproducibility of 3D ultrasound is better.