Methods
The funding for this research came from the Swanson School of Engineering Undergraduate Research Grant and NSF GRFP Grant #1747452. Funding sources had no involvement in the study design, data collection, analysis and interpretation of data, writing of the report, or decision to publish.
This retrospective study was approved by the Institutional Review Board at the University of Pittsburgh and Northshore University HealthSystem. Images from 63 female patients between the ages of 20 and 49 that had a pelvic MRI scan with or without contrast at Magee-Womens Hospital or Northshore University HealthSystem between 2005 and 2018 were included in this study. Exclusion criteria were history of pelvic surgery (not including cesarean delivery (CD)), pelvic masses, and incomplete scans (did not include the necessary bony anatomy) or incomplete birth history.
Subjects were categorized into groups based on parity and gravidity, which resulted in 23 nulliparous, 14 pregnant and vaginally nulliparous, and 26 parous women: For the purposes of this study the women that were placed in the nulliparous and parous groups were not currently pregnant and were at least one year postpartum To delineate the effects of pregnancy on the sacrum-coccyx shape, we then examined the pregnant and parous groups by their respective number of deliveries (CDs for the pregnant group, and combined CDs and vaginal deliveries for the parous group) and normalized those values with respect to the nulliparous average. These will be referred to as the pregnant and parous subgroups. The reasoning behind the relatively small sample of pregnant women is likely due to one of the exclusion criteria. Because we wanted to isolate the effects of pregnancy, we eliminated pregnant subjects that were vaginally parous. This is why all previous births of women in the pregnant group were from CD.
Using HOROS v3.3.5 (Nimble Co LLC, Annapolis, MD USA) the midsagittal plane of the sacrum and coccyx was identified. The sacrum was defined as the first 5 vertebrae inferior to the sacral promontory. Any remaining vertebrae were defined as the coccyx. This resulted in 3, 4, or 5 coccygeal vertebrae. Twelve length, angle, and curvature measurements were made using definitions from previous literature to define the sacrum, coccyx, and combined sacrum-coccyx shape13. These measures included a count (the number of coccygeal vertebrae), length (measured as both a straight and curved length), angle, and curvature index. A curvature index was defined as a ratio between a straight length (the shortest distance between the top and bottom of a structure) and curved length (the average of the anterior and posterior borders of a structure) multiplied by 100. A curvature index of 100 indicates that a structure is perfectly straight. The sacrococcygeal straight length, curved length (anterior, posterior, and average), and curvature index are shown in Figure 1a. Angles were defined as the included angle between two straight lines, thus making an angle closer to 180 degrees straighter. The sacrococcygeal angle is shown in Figure 1b. The sacral angle was excluded from this study as many scans did not include necessary sacral landmarks13.
All of these measures were defined and measured in previous literature that intended to quantify the change in shape of the sacrum and coccyx that aimed to investigate and define normal adult sacrococcygeal morphometry and were as follows: Coccygeal curved and straight lengths were measured from the middle of the upper border of Co1 to the coccygeal tip; sacral curved and straight lengths were measured from the middle of the upper border of S1 to the middle of the of the inferior border of S5; sacrococcygeal curved and straight lengths were measured from S1 to the tip of the coccyx; sacrococcygeal angle was the included angle between the middle of the superior portion of S1, the middle of the superior portion of the Co1, and the tip of the coccyx (Figure 1b); and the coccygeal angle was the included angle between the line drawn through the middle of the superior and inferior edges of Co1 and the line drawn through the middle of the superior and inferior edges of the most inferior coccygeal vertebrae.
Statistical analyses were conducted in IBM SPSS Statistics v25 (IBM Corp., Armonk, NY USA) and consisted of a One-Way Independent MANCOVA followed by univariate ANOVAs with multiple comparisons and Benjamini-Hochberg (BH) corrections post-hoc14. The covariate was the age of the patient. In a BH correction, an allowable false discovery rate (the rate at which a null hypothesis is rejected incorrectly) is chosen (10% for this study). A critical value is calculated using the rank of the p-value from the MANCOVA analysis, the allowable false discovery rate, and the number of measurements. A p-value smaller than the critical value is considered significant. Those variables were then considered in a univariate analysis. Using the rank of the p-values instead of the numerical value of them means that a BH correction is a less conservative alternative to a Bonferroni correction. Measures with significant differences between groups were followed-up with additional multiple comparisons. Homogeneity of variances were tested, and independent samples were assumed.