Placenta accreta spectrum (PAS) is a serious obstetric complication that can cause massive postpartum bleeding. During the past several decades, the prevalence of PAS has been increasing and varies from country to country. Over the past four decades, an approximately 5- to 10-fold increase in the incidence of PAS has occurred. (19). The International Federation of Gynecology and Obstetrics (FIGO) classified PAS disorders, and it has been used in clinical practice worldwide. (18). Ultrasound is the most convenient way of screening PAS patients during routine clinical visits. There are many PAS scoring systems based on ultrasound findings from different centers. (15, 20, 21, and 22) This study aimed to validate a prenatal ultrasound (US) staging system for placenta accreta spectrum (PAS) disorders in women with placenta previa proposed by (13) And to evaluate its associations with surgical outcome, placental invasion, and the clinical staging system for PAS disorders proposed by the International Federation of Gynecology and Obstetrics (FIGO). In our study, according to the PAS system, the distribution of patients in stages of the scoring system revealed that most of them had PAS grades 2 and 3. Cali et al. reported that most women were classified as having a PAS score of 0, followed by those with a PAS score of 3 (Cali et al., 2019). This difference could presumably be due to the high rate of CS in Egypt (23) In this study, twenty patients (35%) underwent hysterectomy. Other studies reported a lower hysterectomy rate. (24) (21). This may be attributed to the use of a preoperative aortic balloon, uterine artery embolization, and balloon tamponade, leading to a reduction in bleeding from the placental bed. (24) (21). CS is considered the main risk factor for placenta accreta. (25). Although the number of previous CSs was not included in our study, we observed a statistically significant positive correlation between the PAS score and the number of previous CSs. This result is consistent with previously reported results. (26)- (27). However, several studies have used the number of previous CSs in their scoring system. They reported that the combination of placenta previa, the number of prior cesarean deliveries, and ultrasound suspicion of invasion was more predictive than ultrasound variables alone. (15)(20) (21) (27) Based on ultrasound features, several studies have reported other scoring systems to predict the degree of placental invasion. Tovbin et al. relied on the number and size of placental lacunae, the presence of bladder wall interruption, and the obliteration of the demarcation between the uterus and the placenta in their scoring system (15). This scoring system differs from that reported by Cali et al. in the number and size of placental lacunes. However, they recorded signs like those reported by Cali et al., except for ureterovesical and parametrial hypervascularity. In the retrospective study of Rac et al., their scoring system was based on the combination of the smallest sagittal myometrial thickness and grading of lacunae by measuring the number and size of lacunes for the score of placental lacunae, as proposed by Feinberg and Williams. (28) bridging vessels, the number of previous Cesarean deliveries, and the placental location (20). This scoring system differs from that of Cali et al. in terms of all ultrasonographic parameters. Zheng and colleagues reported another scoring system through a multicenter retrospective study (20 tertiary centers with 2219 patients), which included a combination of risk factors such as the number of CSs and ultrasonographic features. Some of these features, such as placental lacunae and loss of the “clear zone,” were included in the Cali et al. scoring system. However, other features, such as myometrial thinning, exophytic placental bulge, sub-placental hypervascularity, and extension into the bladder and cervix, were not included in the Cali et al. scoring system. Interestingly, they also validated their scoring system at one institution (Peking University Hospital) to avoid bias in the results from different sonographers and operators at different centers. (27). Considering the number of prediction models that have been developed, the percentage of external validation studies is small. (29). A study performed by Alsadah et al validated sonographic-based scoring systems for the prediction of morbidly adherent placenta (MAP) in high-risk populations(15) (20), and (16). They reported that Tovbin et al. had a superior ability to predict the MAP than the other two scoring systems did. (30). Further studies validated the scoring systems. (22 − 13) and (14). They reported that there was no single superior system. (31). Tovbin et al. reported better diagnostic performance for PAS, as they reported specificity and positive and negative predictive values of 98.7%, 84.2%, and 97.1%, respectively, and lower sensitivity values (69.6%), with an area under the curve of 0.94. (15). Although the Rac et al. study yielded higher specificity (100%), positive predictive value (100%), and negative predictive value (71%), with a higher area under the curve of 0.87, the sensitivity was very low compared with our results. (20) Furthermore, Zhang et al reported higher sensitivity (92.6%) and specificity (85.0%) than our findings when they evaluated a scoring system with maternal risk factors, and the AUC was slightly greater than that when ultrasound features alone. Zhang et al., 2021 were used. Moreover, Luo et al. reported higher positive and negative predictive values and false positive rates of the scoring system (96.68%, 95.44%, and 3.32%, respectively) (21). The Del Negro et al. study reported good performance, with a sensitivity of 100%, specificity of 89%, and accuracy of 92%, with an area under the curve (AUC) of 0.94 (32). Based on the superior performance and inclusion of most ultrasonographic descriptors, this scoring system appears to be the most appropriate for the prediction of placental invasion in patients with PAS. However, it has not been validated, and external validation is needed to confirm these results. Moreover, these studies were heterogeneous, showing marked differences in their type (some were retrospective, others were prospective) and the number of included patients in each study. External validation is important because prediction models, risk scores, and decision tools are becoming more integral parts of surgical practice. Additionally, it is necessary to assess whether a prediction model is accurate. (33). Furthermore, it is necessary to determine a prediction model’s reproducibility and generalizability to new and different patients. (29) The Cali scoring system was easily applied to patients. The application of the Cali scoring system in our study could predict the presence of placenta percreta. However, the level of performance was lower than that reported by Cali et al., with an area under the curve of 0.625, sensitivity of 75%, specificity of 50%, positive predictive value of 60%, negative predictive value of 66.7%, and overall accuracy of 62.5%. In contrast, Cali and colleagues validated their scoring system in comparison with other scoring systems and reported that women with PAS2 or PAS3 were affected exclusively by placenta percreta, with higher records for the area under the curve (0.85), sensitivity (91%), and specificity (78%) (13) (31). This difference in performance could be attributed to the number of enrolled patients (57 patients in our study versus 259 patients in the Cali et al study). However, our study was designed to be prospective. In our study, the PAS ultrasound scoring system and FIGO clinical grading system were applied prospectively. However, in the Cali et al. study, the correlation between the ultrasound staging system and the clinical grading system proposed by FIGO was affected by the retrospective nature of the analysis because, at the time the study was conducted, the FIGO grading system was not yet available. (13). It is assumed that the prospective design is better than the retrospective design, as sonographers can obtain well-scored sonograms through a targeted placental search, which compensates for the drawbacks of a retrospective study. (34). The depth of placental invasion is one of the major determinants of surgical outcome in women with a PAS disorder, with those affected by placenta percreta being at greater risk of intra-surgical complications such as massive hemorrhage, the need for blood transfusion, and damage to adjacent organs(35). Despite the statistically nonsignificant relationship between the PAS stage and intraoperative complications, complications mostly occurred in PAS 3 patients (50%), whereas one patient had complications in PAS 2 patients (9.1%). This result is in agreement with the study performed by Cali et al., as all cases of complications occurred in PAS 3 (27.8%) and PAS 2 (25%) patients. (13). In our study, there was a statistically significant relationship between PAS stage and operative time, blood transfusion, fresh frozen plasma transfusion, length of hospital stays, and ICU admission. This result is consistent with that previously reported by Cali et al.(13). The prospective data collection provides strength for our study. Furthermore, we evaluated the correlation of the proposed staging system for PAS disorders not only with surgical outcomes but also with the depth of placental invasion and the FIGO grading system. Most of our cases were evaluated by the same expert sonographer. Patients were operated on by expert multidisciplinary teams at the PAS. Our study had several limitations. First, it was a single-center study. Second, patients affected by PAS were not managed by the same multidisciplinary team. Third, a small number of patients were included in this study.Strengths and limitations
