Predicting mid-pelvic interspinous distance in women using height and pubic arch angle.

A successful vaginal delivery requires a well-matched maternal pelvis and fetal size plus presentation. Fetal size is measured by ultrasonography before the delivery. However, the maternal pelvis is not measured before labor, although labor per se has been reported as a way to measure the pelvis. Previous studies have examined the maternal pelvis using X-ray, computed tomography, and magnetic resonance imaging (MRI) [ 1 – 3 ]; however, the use of these methodologies is limited because of radiation exposure and high costs. The maternal pelvis consists of the inlet pelvis, mid-pelvis, outlet pelvis, and pubic arch angle (PAA). The ischial spine is an important landmark in fetal engagement during labor. The station is zero when the fetal head descends to the level of this plane. The interspinous distance (ISD), i.e., the mid-pelvis, is the narrowest part of the pelvis [ 2 ], which may be commonly obstructed when labor occurs at this point. The ISD is a useful predictor of successful vaginal delivery [ 4 ]. To date, the ISD has been measured using the index and middle fingers [ 5 ], and its measurement requires substantial experience and clinical skills. Thus, cephalopelvic disproportion (CPD) at this point cannot be predicted accurately in advance. A short stature is a risk factor for CPD, for example when maternal height is <145 cm [ 6 ], as women with short stature have smaller pelvic dimensions. Women who have delivered vaginally have a larger pelvic size than women with CPD [ 7 ]. Hence, we hypothesized that height affects female pelvic size, including the ISD, and that ISD-related relational expressions may be derived utilizing anatomical elements that affect the ISD but cannot be assessed directly [ 8 ]. This study aimed to determine the anatomical relationship between patient height and ISD plus other pelvic parameters obtained using MRI.

Altogether, 710 women were included in this study. The mean ± standard deviation age of the patients was 40.5 ± 6.4 years (range, 21~50 years) and the mean height was 160.1 ± 5.5 cm. Patients’ mean inlet-AP distance, obstetric conjugate, mid-AP distance, outlet-AP distance, ISD, ischial tuberosity distance, and PAA was 129.7 ± 9.0 mm, 110.4 ± 8.1 mm, 119.7 ± 8.5 mm, 111.7 ± 8.9 mm, 108.9 ± 7.9 mm, 121.8 ± 11.7 mm, and 87.4 ± 6.6°, respectively. Regarding obstetric history, 245 women had a vaginal delivery, 146 women had a cesarean section, six women had only a stillbirth, and 313 women had no obstetric history ( Table 2 ). NSVD, Normal spontaneous vaginal delivery; c/sec, Caesarean section; AP, Anteroposterior (distance); trans, Transverse (distance). The nonsignificant Shapiro–Wilk’s test revealed a normal distribution for the following parameters: inlet-AP distance, obstetric conjugate, mid-AP distance, outlet-AP distance, ISD, ischial tuberosity distance, PAA, and height ( Fig 2 ). The scatterplots and Pearson correlation coefficient between each parameter. All parameters and height were significantly linearly related to ISD. The Pearson correlation coefficients between ISD and each parameter are listed in Table 3 . AP, Anteroposterior. The results of the multivariable linear regression models are presented in Tables 4 and 5 . In the first model ( Table 4 ), all parameters and height were used as independent variables. Stepwise linear regression was conducted, and the obstetric conjugate and mid-AP distance values were removed. The best model was ISD = 0.25268 × height − 0.07080 × inlet-AP distance + 0.12204 × outlet-AP distance + 0.29350 × ischial tuberosity distance + 0.32337 × PAA (R 2 = 0.9973 [adjusted R 2 = 0.9973], P < 0.001). The standardized coefficient was 0.37, −0.08, 0.13, 0.33, and 0.26 for height, inlet-AP distance, outlet-AP distance, ischial tuberosity distance (outlet transverse), and PAA, respectively. The most impactful factor was height. In this formula, R 2 equals 0.9973, indicating a 99.73% explanatory power. ISD, Interspinous distance; AP, Anteroposterior. ISD, Interspinous distance. In the second model ( Table 5 ), a multivariable linear regression analysis was performed using height and PAA plus the inlet-AP distance from intrapartum ultrasonography [ 8 , 10 ]. The inlet-AP distance was not significant as a regression coefficient, and thus it was removed. The best model was fitted only using height and PAA (ISD = 0.40935 × height + 0.49679 × PAA (R 2 = 0.9965 [adjusted R 2 = 0.9965], P < 0.001). The standardized coefficient was 0.60 and 0.39 for height and PAA, respectively. In this formula, R 2 equals 0.9965, indicating a 99.65% explanatory power.

In conclusion, this study predicts ISD values using pelvic bone and height data and demonstrates that the ISD can be predicted using the patient’s height and measured PAA. A comparison study of MR and ultrasound images will be required to predict ISD cutoff values in the delivery room. This should allow clinicians to anticipate CPD.

This study retrospectively analyzed the data of women who underwent pelvic MRI due to various reasons such as myomas, adenomyosis, and gynecological cancer from January 2014 to June 2020 at our hospital. Of the 750 women with MRI data, 710 women of reproductive age (20–50 years) were recruited. The exclusion criteria were congenital pelvic deformity and history of pelvic trauma. Data on height and obstetric history were collected from medical records, and MRI were obtained from a picture archiving and communication system. Magnetic resonance pelvimetry was performed with the patient in the supine position, using three MRI systems (SIGNA 3.0T HDXT, GE Healthcare, Chicago, IL, USA, 2006; SIGNA ARCHITEC, GE Healthcare, 2018; and SIGNA HDX 1.5T, GE Healthcare, 2006). We measured the following seven parameters on MR images [ 4 ] ( Table 1 and Fig 1 ): (1) true conjugate (inlet-anteroposterior [AP] distance), from the sacral promontory to the top of the symphysis pubis; (2) obstetric conjugate, from the sacral promontory to the inner margin of the symphysis pubis; (3) mid-AP distance, from the point between the S4 and S5 to the lower margin of the pubis symphysis; (4) outlet-AP distance, from the tip of the sacrum (not coccyx) to the bottom of the inner cortex of the symphysis pubis; (5) ISD (narrowest) between the ischial spines; (6) ischial tuberosity distance, the intertuberous (widest) distance between the ischial tuberosities; and (7) PAA, the angle between both ischial tuberosity and inter margin of the pubic symphysis (inferior). (A) Sagittal view: (a) true conjugate (inlet-anteroposterior [AP] distance), (b) obstetric conjugate, (c) mid-AP distance, and (d) outlet-AP distance. (B) Interspinous distance in an axial plane at the level of the ischial spines. (C) Ischial tuberosity distance. (D) Pubic arch angle. AP, Anteroposterior (distance). R version 3.6.3 (R core team, 2020) was used for the statistical analysis, and statistical significance was set at P ≤ 0.05. The Shapiro–Wilk test and Quantile–Quantile plot were used to test for normality. Pearson correlation analyses were performed to determine the correlation between ISD and all other parameters. Two multiple linear regression analyses were performed. First, all parameters were used for measuring ISD. Second, the height and inlet-AP distance plus PAA, which were measurable by intrapartum ultrasound, for measuring ISD were analyzed. The minimal sample size was 153, as determined by a preliminary study using G*power version 3.1.9.4 [ 9 ], with an effective medium size of 0.15, a P -value of 0.05, a power of 0.95, and 7 predictors. The study was approved by the Institutional Review Board and Ethics Committee of CHA Bundang Medical Center, CHA University, Seongnam City, South Korea (No. 2020-06-003, June 2, 2020). The requirement for informed consent was waived owing to the retrospective nature of the study.