Reproducibility of assessment of full-dilatation Cesarean section scar in women undergoing second-trimester screening for preterm birth

Abstract

Objective: To assess the reproducibility of a standardized method of measuring the Cesarean section (CS) scar, CS scar niche and their position relative to the internal os of the uterine cervix by transvaginal ultrasound in pregnant women with a previous full-dilatation CS. Methods: This was a prospective, single-center reproducibility study on women with a singleton pregnancy and a previous full-dilatation CS who underwent transvaginal ultrasound assessment of cervical length and CS scar characteristics at 14–24 weeks' gestation. The CS scar was identified as a hypoechogenic linear discontinuity of the myometrium at the anterior wall of the lower uterine segment or cervix. The CS scar niche was identified as an indentation at the site of the scar with a depth of at least 2 mm. The CS scar position was evaluated by measuring the distance to the internal cervical os. CS scar niche parameters, including its length, depth, width, and residual and adjacent myometrial thickness, were assessed in the sagittal and transverse planes. Qualitative reproducibility was assessed by agreement regarding visibility of the CS scar and niche. Quantitative reproducibility of CS scar measurements was assessed using three sets of images: (1) real-time two-dimensional (2D) images (real-time acquisition and caliper placement on 2D images by two operators), (2) offline 2D still images (offline caliper placement by two operators on stored 2D images acquired by one operator) and (3) three-dimensional (3D) volume images (volume manipulation and caliper placement on 2D images extracted by two operators). Agreement on CS scar visibility and the presence of a niche was analyzed using kappa coefficients. Intraobserver and interobserver reproducibility of quantitative measurements was assessed using Bland–Altman plots. Results: To achieve the desired statistical power, 72 women were recruited. The CS scar was visualized in > 80% of images. Interobserver agreement for scar visualization and presence of a niche in real-time 2D images was excellent (kappa coefficients of 0.84 and 0.85, respectively). Overall, reproducibility was higher for real-time 2D and offline 2D still images than for 3D volume images. The 95% limits of agreement (LOA) for intraobserver reproducibility were between ± 1.1 and ± 3.6 mm for all sets of images; the 95% LOA for interobserver reproducibility were between ± 2.0 and ± 6.3 mm. Measurement of the distance from the CS scar to the internal cervical os was the most reproducible 2D measurement (intraobserver and interobserver 95% LOA within ± 1.6 and ± 2.7 mm, respectively). Overall, niche measurements were the least reproducible measurements (intraobserver 95% LOA between ± 1.6 and ± 3.6 mm; interobserver 95% LOA between ± 3.1 and ± 6.3 mm). There was no consistent difference between measurements obtained by reacquisition of 2D images (planes obtained twice and caliper placed), caliper placement on 2D stored images or volume manipulation (planes obtained twice and caliper placed). Conclusions: The CS scar position and scar niche in pregnant women with a previous full-dilatation CS can be assessed in the second trimester of a subsequent pregnancy using either 2D or 3D volume ultrasound imaging with a high level of reproducibility. Overall, the most reproducible CS scar parameter is the distance from the CS scar to the internal cervical os. The method proposed in this study should enable clinicians to assess the CS scar reliably and may help predict pregnancy outcome. © 2022 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.