Three-Dimensional Volume Ultrasound Assessment of Cesarean Scar Niche and Cervix in Pregnant Women: A Reproducibility Study.

Abstract

To assess the reproducibility of standardized 3-dimensional (3D) ultrasound volume analysis of the dimensions and the position of cesarean birth (CB) scar niche relative to the cervix in pregnant women. This prospective single-center study in women with 1 previous CB ≥8 cm cervical dilatation acquired ultrasound volumes between 11 and 24 weeks' gestation in a mid-sagittal plane. Two experienced operators processed the volumes using virtual organ computer-aided analysis. A CB scar niche was defined as an indentation at the scar site of ≥2 mm in depth. Niche and cervix volumes were calculated using manual contouring. Agreement for categorical variables was expressed using intraclass correlation coefficient (ICC). The Bland-Altman method was used to assess numerical variable reproducibility. To achieve the desired statistical power, 52 participants were included. The intraobserver agreement on niche classification relative to the internal os was 100%, with an interobserver kappa coefficient of 0.98 (95% confidence interval [CI] 0.97-0.99, P < .05). The intraobserver ICC for niche volume was 0.94 (95% CI 0.90-0.96; P < .001), with a mean difference of -15.32 mm3 (±109.32). The interobserver ICC was 0.78 (95% CI 0.62-0.87; P < .001), with a mean difference of -21.57 mm3 (±202.01). The ICC for niche/cervix volume ratio were 0.94 (95% CI 0.90-0.96; P < .001) and 0.79 (95% CI 0.63-0.87; P < .001) for intra- and interobserver reproducibility, respectively. This study demonstrates that 3D CB scar sonographic features are highly reproducible in pregnant women with a history of advanced labor CB. The validated protocol can guide future research on the association with subsequent adverse pregnancy outcomes.