Diagnostic Accuracy of MRI to Predict Peripartum Hysterectomy and Neonatal Mortality in Total Placenta Previa: A Retrospective Cohort Study.
Diagnostic Accuracy of MRI to Predict Peripartum Hysterectomy and Neonatal Mortality in Total Placenta Previa: A Retrospective Cohort Study.
To assess the reliability of placental magnetic resonance imaging measurements in predicting peripartum hysterectomy and neonatal outcomes in patients with total placenta previa.
Study design and patients This study analyzed the outcomes of a cohort of 372 pregnant women diagnosed with PP over five years, from November 2017 to June 2023. The patient population for the study included women between the ages of 17 and 42 who were in their third trimester of pregnancy (27th to 37th weeks of gestation) and had received a t-PP diagnosis via color Doppler ultrasonography (cd-USG) and/or p-MRI. The resulting cohort compared a final sample population of 277 singleton pregnant women with t-PP, which comprised 150 pregnant women who underwent antenatal p-MRI examinations in the third trimester and 127 pregnant women who did not undergo a p-MRI.
Standards of reference The study covered all singleton pregnancies in which t-PP extended to both the anterior and posterior uterine walls, resulting in complete coverage of the internal cervical os by the placenta. Transvaginal and transabdominal USG were planned between 32 and 34 weeks for all patients who were followed up in our clinic due to t-PP and were scheduled to be delivered. The gestational ages of the participants were determined using the last menstrual date and crown-rump length (CRL) measurements from a first-trimester ultrasound. Two perinatologists with at least ten years of experience examined each patient's cd-USG for signs of PAS. Ultrasonographic signs identified included placental lacunae, placenta previa involving the cervix, loss of the clear zone, and bladder wall interruption.
Two radiologists with at least ten years of experience, independently of each other, evaluated p-MRI scans of pregnant women diagnosed with t-PP without prior knowledge of the original reports.
Definitive diagnosis and pathological reporting Pathological reports and approvals were granted according to a rigorous protocol involving at least two pathologists, each with a minimum of seven years of professional expertise in their field.
p-MRI image interpretations The assessment focused on analyzing the different positions of the atypical placenta in the lower uterine segment, including the anterior, posterior, anterolateral, and posterolateral locations. Precise measurements were taken to calculate the placental volume in the S1 and S2 sectors, cervical canal length, and cervical canal dilatation. Based on sagittal p-MRI scans, the placental invasion was classified into two main areas. Implementing a plane perpendicular to the upper bladder's axis facilitated the boundary of these sectors. The term S1 refers to the upper uterine segment that forms the part of the uterus that contacts the bladder's posterior wall. Likewise, the lower uterine segment, identified as S2, was formed in the lower posterior wall of the bladder. Radiologists conducted a sector-wise (S1 and S2) examination of the invasion topography, which allowed them to identify specific invasion locations, including parametrial and cervical involvement and bladder interruption.
Multidisciplinary team The operations of patients who gave birth due to PAS were performed by the same multidisciplinary team, which included perinatologists, gynecological and obstetric surgeons, gynecological oncologists, urologists, neonatologists, cardiovascular surgeons, anesthesiology and reanimation specialists, interventional radiologists, and allied health personnel with ten years of experience.
Scheduled surgery and preoperative preparations All participants in our study underwent cesarean births during the third trimester of pregnancy. Except for urgent situations, all scheduled preterm cesarean deliveries and p-TAH procedures were conducted during the gestational period of 34+0 to 36+6 weeks. Among the patients admitted to our clinic due to t-PP, surgical intervention was performed before the planned delivery date in cases of uterine contraction, the onset of vaginal bleeding, or unexpected medical indications for the mother and fetus.
Statistical Method and Power Analysis G*Power V. 3.1.9.6 estimated sample size. All p-MRI values of patients with and without PAS were compared in the reference study, and the difference was substantial. Using 95% confidence (1-α), 95% test power (1-β), d=0.8 effect size, and the two-way hypothesis, the study contained 84 cases, with a minimum of 42 in each group. The study included 155 cesarean sections and 122 TAH cases; post hoc analysis revealed a power of 99.9%. 16 Data were analyzed using IBM SPSS V23. ROC analysis determined the surgical procedure variable cut-off values. Univariate and multivariate binary logistic regression analyses examined independent surgical risk factors. Backward-Wald method was used for multivariate binary logistic regression. The analysis findings present mean ± standard deviation, median (minimum-maximum), and frequency (percentage) for quantitative and categorical variables. Intraclass correlation coefficients and Kappa (k) tests were conducted to assess intra- and inter-observer agreement reliability. The significance level was set at p<0.05.
Inclusion Criteria:
Exclusion Criteria: