Reproducibility of transvaginal three-dimensional endometrial volume measurements during ovarian stimulation.

Before estimating the clinical role of a method, the reproducibility has to be evaluated precisely. This study aimed to document the reproducibility of the endometrial volume measurement by three-dimensional (3D)-ultrasound. The volume measurements were done on 57 consecutive in-vitro fertilization (IVF) patients with either the full planar (contour) or the three distance method. A paired t-test provided no statistically significant difference between the two methods. Linear regression analysis, using the full planar method as independent and the other as dependent variables, yielded the following results: intercept = 0.348, not statistically different from 0; slope = 0.962, statistically different from 1 (P < 0.01). Interobserver reliability for the three distance method was 0.6667 and for the full planar method was 0.9565. Intra-observer reliability for the three distance method was 0.8426 and for the full planar method 0.9394. The correspondence between and within observers seemed to be good. Both methods are reliable, but the full planar method seems to provide slightly better reproducibility in regard to endometrium volume measurement.