Subtotal versus total hysterectomy in placenta accreta spectrum management at a tertiary centre in saudi arabia: a retrospective study Authors: Nawal A. Alharbi, Ghada I. Abdulhamed, Syeda I. Mohsina, Abdulaziz A. Alsharif, Shaden A. Almojel Published: 2026 ABSTRACT BackgroundPlacenta accreta spectrum (PAS) is a major contributor to life-threatening obstetric hemorrhage and is increasingly encountered worldwide due to rising caesarean delivery rates. Caesarean hysterectomy remains the cornerstone of management in most cases; however, the optimal surgical techniques, subtotal versus total hysterectomy, continue to be debated. Current evidence is limited and heterogeneous, and practice often varies according to surgeon preference and intraoperative findings. By evaluating outcomes in a large cohort at a national tertiary referral center, this study aims to clarify the comparative safety and effectiveness of subtotal versus total hysterectomy in PAS management.MethodsWe conducted a retrospective cohort study of women with PAS who underwent caesarean hysterectomy between 2018 and 2023. The primary outcome was intraoperative quantified blood loss (QBL). Secondary outcomes included operative time and bladder injury. Multivariable linear regression on log-transformed QBL and operative time was used to estimate adjusted geometric mean ratios, while bladder injury was analyzed using multivariable logistic regression. All models adjusted for surgical urgency, antenatal PAS severity, suspected cervical involvement, number of prior caesarean deliveries, gestational age at delivery, and surgeon identifier.ResultsAmong the included patients, subtotal hysterectomy was associated with significantly lower intraoperative blood loss compared with total hysterectomy (adjusted geometric mean ratio 0.36, 95% CI 0.32–0.41; p-value < 0.001). Operative time was also significantly shorter in the subtotal hysterectomy group (adjusted geometric mean ratio 0.54, 95% CI 0.50–0.59; p-value < 0.001). In addition, subtotal hysterectomy was associated with a significantly lower odds of bladder injury (adjusted odds ratio 0.19, 95% CI 0.05–0.68; p-value = 0.013).ConclusionSubtotal hysterectomy was associated with significantly lower intraoperative blood loss, shorter operative time, and fewer bladder injuries compared to total hysterectomy for placenta accreta spectrum. These findings support subtotal hysterectomy as a feasible and potentially safer surgical option in selected cases. While these findings support subtotal hysterectomy as a viable option, the retrospective design and potential selection bias highlight the need for individualized decision-making. Further multicenter prospective research with standardized surgical criteria is needed to confirm these results, minimize confounding from PAS severity, and determine the most effective surgical approach. Background Placenta accreta spectrum (PAS) comprises abnormal placental attachments, including accreta, increta, and percreta. PAS is a significant contributor to maternal morbidity and mortality due to massive obstetric hemorrhage [1–3]. These conditions, commonly linked to uterine scarring from previous caesarean deliveries, are increasingly encountered worldwide, reaching approximately 0.5% (1/200) in the last decade [4–6]. Several retrospective cohort studies on PAS have demonstrated reduced maternal morbidity when managed by multidisciplinary teams (MDTs) with specialized surgical expertise to tailor individualized management plans based on the severity and complexity of accreta spectrum disorders [7–10]. Hysterectomy remains the mainstay of treatment, though the optimal surgical approach—subtotal versus total hysterectomy—remains debated [11, 12]. This study aims to assess the efficacy and outcomes of subtotal hysterectomy compared to total hysterectomy in PAS management at a tertiary center with multidisciplinary expertise. Materials and methods Study design and population A retrospective cohort study was conducted at King Saud Medical City, Riyadh, Saudi Arabia. We included 198 women with antenatally diagnosed placenta accreta spectrum (PAS) who underwent caesarean hysterectomy between January 2018 and December 2023. Inclusion criteria comprised patients with a history of prior caesarean section and a diagnosis of PAS. Cases of placenta previa without evidence of PAS were excluded. PAS diagnosis was established antenatally using ultrasound with colour Doppler and, in selected cases, magnetic resonance imaging, and was confirmed intraoperatively. Histopathological examination did not consistently allow definitive subclassification into accreta, increta, or percreta due to tissue disruption and autolysis reported by the pathology department, which limited assessment of the myometrial–placental interface. Consequently, histopathological subclassification was not included in the analytical models. Surgical technique Subtotal hysterectomy was performed with the placenta left in situ (Fig. 1A, B). Ureters were identified, dissection, and mobilization of the adherent bladder downwards, and ligation of the uterine arteries. The uterus was elevated, and the lower uterine segment beneath the placenta was manually compressed and stretched upward in both the anterior and posterior planes to facilitate upward displacement of the placenta. Simultaneously, bilateral clamps were placed at the lower uterine segment below the placenta and above the cervix. The uterus was then excised, followed by transfixation of the pedicles and meticulous hemostasis and closure of the uterine stump. Following uterine removal, no active bleeding was noted from the cervical canal; only mucous discharge was observed, and the cervix appeared free of any placental tissue involvement (Fig. 1C, D). Fig. 1A–D. Intraoperative Findings During Elective Subtotal Hysterectomy for Placenta Percreta at 34 Weeks’ Gestation in a Patient With Four Previous Caesarean Deliveries. A The exteriorized gravid uterus shows prominent vascularity, with placental tissue visible just beneath the uterine serosa. B Transection of the uterine body with placenta left in situ above the cervix following devascularization. C Inspection of the cervix after the procedure shows only mucous secretions without active bleeding. D Intraoperative view following subtotal hysterectomy, demonstrating the uterine stump with vascular clamps applied at a level just above the uterosacral ligaments. The preserved tissue appears thick and myometrial in nature, with no vaginal cuff opening, confirming that a subtotal hysterectomy was performed A–D. Intraoperative Findings During Elective Subtotal Hysterectomy for Placenta Percreta at 34 Weeks’ Gestation in a Patient With Four Previous Caesarean Deliveries. A The exteriorized gravid uterus shows prominent vascularity, with placental tissue visible just beneath the uterine serosa. B Transection of the uterine body with placenta left in situ above the cervix following devascularization. C Inspection of the cervix after the procedure shows only mucous secretions without active bleeding. D Intraoperative view following subtotal hysterectomy, demonstrating the uterine stump with vascular clamps applied at a level just above the uterosacral ligaments. The preserved tissue appears thick and myometrial in nature, with no vaginal cuff opening, confirming that a subtotal hysterectomy was performed The steps for a total hysterectomy are essentially similar, except that they include additional dissection to open the vaginal vault, remove the cervix, and close the vaginal cuff. All patients were routinely admitted to the intensive care unit for at least 24 h postoperatively as part of our standardized PAS protocol. Criteria for surgical technique selection The selection of surgical approach (subtotal versus total hysterectomy) was undertaken intraoperatively at the discretion of the senior surgeon, informed by placental topography, the extent and depth of placental involvement, the anticipated complexity of bladder dissection, intraoperative hemodynamic stability, and overall operative judgment. These explicit criteria clarify the clinical rationale and acknowledge the potential for selection bias. While preoperative and postoperative management of PAS followed an established institutional protocol, there was no standardized institutional consensus regarding the choice of surgical technique during the study period. Data collection and statistical analysis Data collected included maternal age, gestational age at delivery, number of prior caesarean sections, antenatal PAS severity, intraoperative findings, and surgical complications. The primary outcome was quantified intraoperative blood loss (QBL, litres). Secondary outcomes included operative time (hours), intraoperative and postoperative complications (including bladder or ureteric injury and bowel injury), need for blood transfusion, reoperation, and maternal mortality. Log-linear regression models were used for QBL and operative time, with heteroscedasticity-robust standard errors; results are reported as adjusted geometric mean ratios (aGMRs) with 95% confidence intervals. Multivariable logistic regression was used to analyse bladder injury, with results reported as adjusted odds ratios (aORs) and 95% confidence intervals. Prespecified covariates included surgical urgency (elective vs. emergency), antenatal PAS severity, number of previous caesarean deliveries, gestational age at delivery, and operating surgeon. Variables with less than 5% missing data were analysed using complete-case analysis. Sensitivity analyses excluded cases with extreme haemorrhage (QBL > 6 L). Analyses were performed using SPSS version 26.0 and R version 4.3.0. Statistical significance was set at p-value < 0.05 (two-sided). Surgical team and data sources All procedures were led by a consultant obstetrician, with intraoperative support from a second senior obstetrician or a gynaecologic oncologic surgeon according to the on-call rota. The centre serves as a national tertiary PAS referral unit, and senior surgical support was routinely available. Clinical data were extracted from electronic medical records and the institutional PAS database. Sample size and descriptive statistics As this was a retrospective study, no a priori sample size calculation was performed. A post-hoc detectable-effect assessment demonstrated greater than 95% power to detect the observed difference in quantified blood loss at a two-sided α level of 0.05. Continuous variables were assessed for normality using the Shapiro–Wilk test and visual inspection of histograms and Q–Q plots. Categorical variables were compared using the chi-square or Fisher’s exact test, as appropriate, and continuous variables using the independent t-test or Mann–Whitney U test. Results are presented as means (standard deviation) or medians (interquartile range). Results A total of 198 patients with placenta accreta spectrum (PAS) underwent caesarean hysterectomy during the study period. Of these, 121 (61%) underwent subtotal hysterectomy, and 77 (39%) underwent total hysterectomy. Elective procedures accounted for 113 cases (57%), while 85 (43%) were performed emergently, most commonly for acute intrapartum or antenatal haemorrhage (Table 1). Table 1Distribution of hysterectomy type and surgical urgency (n = 198)Type of hysterectomyElective, n (%)Emergency, n (%)Total, n (%)Total hysterectomy51 (66.2%)26 (33.8%)77 (39%)Subtotal hysterectomy62 (51.2%)59 (48.8%)121 (61%)Total cases113 (57%)85 (43%)198 (100%)Most hysterectomies performed were subtotal (61%), with a near-equal distribution between elective (51.2%) and emergency (48.8%) cases. Total hysterectomies accounted for 39% of procedures, predominantly elective (66%) Table 1Distribution of hysterectomy type and surgical urgency (n = 198)Type of hysterectomyElective, n (%)Emergency, n (%)Total, n (%)Total hysterectomy51 (66.2%)26 (33.8%)77 (39%)Subtotal hysterectomy62 (51.2%)59 (48.8%)121 (61%)Total cases113 (57%)85 (43%)198 (100%)Most hysterectomies performed were subtotal (61%), with a near-equal distribution between elective (51.2%) and emergency (48.8%) cases. Total hysterectomies accounted for 39% of procedures, predominantly elective (66%) Distribution of hysterectomy type and surgical urgency (n = 198) Most hysterectomies performed were subtotal (61%), with a near-equal distribution between elective (51.2%) and emergency (48.8%) cases. Total hysterectomies accounted for 39% of procedures, predominantly elective (66%) Baseline characteristics The mean maternal age was 36.5 years, and the mean gestational age at delivery was 35.1 weeks. The number of prior caesarean deliveries ranged from 1 to 9, with 61.6% of women having had four to five previous caesarean sections. PAS subtype classification was primarily based on antenatal imaging and intraoperative findings, with 115 (58%) cases classified as accreta, 49 (25%) as increta, and 34 (17%) as percreta (Table 2). Table 2Baseline characteristics, PAS subtypes, and number of prior caesarean deliveries (n = 198)VariableValue / Number of patients (%)Maternal characteristics Maternal age, mean (years)36.5 Gestational age at delivery, mean (weeks)35.1Placenta accreta spectrum subtype Accreta115 (58%) Increta49 (25%) Percreta34 (17%)Number of prior caesarean deliveries 18 (4.0%) 2–344 (22.2%) 4–5122 (61.6%) > 524 (12.2%) Range1–9Most patients had a history of multiple caesarean deliveries, with 61.6% having 4–5 prior caesarean sections. Placenta accreta was the most common subtype (58%), followed by increta (25%) and percreta (17%). The mean maternal age was 36.5 years, and the mean gestational age at delivery was 35.1 weeks Table 2Baseline characteristics, PAS subtypes, and number of prior caesarean deliveries (n = 198)VariableValue / Number of patients (%)Maternal characteristics Maternal age, mean (years)36.5 Gestational age at delivery, mean (weeks)35.1Placenta accreta spectrum subtype Accreta115 (58%) Increta49 (25%) Percreta34 (17%)Number of prior caesarean deliveries 18 (4.0%) 2–344 (22.2%) 4–5122 (61.6%) > 524 (12.2%) Range1–9Most patients had a history of multiple caesarean deliveries, with 61.6% having 4–5 prior caesarean sections. Placenta accreta was the most common subtype (58%), followed by increta (25%) and percreta (17%). The mean maternal age was 36.5 years, and the mean gestational age at delivery was 35.1 weeks Baseline characteristics, PAS subtypes, and number of prior caesarean deliveries (n = 198) Most patients had a history of multiple caesarean deliveries, with 61.6% having 4–5 prior caesarean sections. Placenta accreta was the most common subtype (58%), followed by increta (25%) and percreta (17%). The mean maternal age was 36.5 years, and the mean gestational age at delivery was 35.1 weeks Primary outcome Subtotal hysterectomy was associated with significantly lower intraoperative quantified blood loss compared with total hysterectomy (adjusted geometric mean ratio [aGMR] 0.36; 95% CI 0.32–0.41; p-value < 0.001). Secondary outcomes Subtotal hysterectomy was associated with a significantly shorter operative time (aGMR 0.54; 95% CI 0.50–0.59; p-value < 0.001) and lower odds of bladder injury (adjusted odds ratio [aOR] 0.19; 95% CI 0.05–0.68; p-value = 0.013). Bladder injury occurred more frequently in the total hysterectomy group (15 vs. 3 cases). Ureteric injury occurred in two patients in each group. There were no bowel injuries, reoperations for haemorrhage, or maternal deaths (Tables 3 and 4). All patients required blood transfusion support and received postoperative intensive care monitoring for a minimum of 24 h in accordance with institutional protocol; therefore, between-group comparisons for these outcomes were not applicable. Table 3Maternal morbidity and mortality according to type of hysterectomy (n = 198). A. Maternal morbidityOutcomeSubtotal hysterectomy (n = 121)Total hysterectomy (n = 77)Mean / Effect difference (95% CI)p-value Blood loss (L), mean (95% CI)2.25 (2.10–2.40)6.25 (5.90–6.60)−4.00 (−4.50 to −3.50)Operative time (hours), mean (95% CI)2.6 (2.5–2.7)4.8 (4.6–5.0)2.2 (−2.4 to −2.0)Bladder injury, n (%)3 (2.5%)15 (19.5%)Ureteric injury, n (%)2 (1.7%)2 (2.6%)—Bowel injury, n (%)00—Return to theatre for hemorrhage, n (%)00— Table 3Maternal morbidity and mortality according to type of hysterectomy (n = 198). A. Maternal morbidityOutcomeSubtotal hysterectomy (n = 121)Total hysterectomy (n = 77)Mean / Effect difference (95% CI)p-value Blood loss (L), mean (95% CI)2.25 (2.10–2.40)6.25 (5.90–6.60)−4.00 (−4.50 to −3.50)Operative time (hours), mean (95% CI)2.6 (2.5–2.7)4.8 (4.6–5.0)2.2 (−2.4 to −2.0)Bladder injury, n (%)3 (2.5%)15 (19.5%)Ureteric injury, n (%)2 (1.7%)2 (2.6%)—Bowel injury, n (%)00—Return to theatre for hemorrhage, n (%)00— Maternal morbidity and mortality according to type of hysterectomy (n = 198). A. Maternal morbidity Table 4Maternal morbidity and mortality according to type of hysterectomy (n = 198). B. Maternal mortalityOutcomeSubtotal hysterectomy (n = 121)Total hysterectomy (n = 77)Maternal death, n (%)0095% confidence intervals (CI) are reported for continuous outcomes. No cases of bowel injury, return to theatre for haemorrhage, or maternal death were observed in either groupAll applicable comparisons include p-values Table 4Maternal morbidity and mortality according to type of hysterectomy (n = 198). B. Maternal mortalityOutcomeSubtotal hysterectomy (n = 121)Total hysterectomy (n = 77)Maternal death, n (%)0095% confidence intervals (CI) are reported for continuous outcomes. No cases of bowel injury, return to theatre for haemorrhage, or maternal death were observed in either groupAll applicable comparisons include p-values Maternal morbidity and mortality according to type of hysterectomy (n = 198). B. Maternal mortality 95% confidence intervals (CI) are reported for continuous outcomes. No cases of bowel injury, return to theatre for haemorrhage, or maternal death were observed in either group All applicable comparisons include p-values Sensitivity and subgroup analyses The associations between surgical approach and the primary and secondary outcomes remained statistically significant after exclusion of cases with extreme haemorrhage (quantified blood loss > 6 L). No significant effect modification was observed by surgical urgency (elective vs. emergency) or PAS severity. Adjusted analysis of surgical outcomes After adjustment for surgical urgency, antenatal PAS severity, suspected cervical involvement, number of prior caesarean deliveries, gestational age at delivery, and surgeon identifier, subtotal hysterectomy remained independently associated with improved operative outcomes (Table 5s). Specifically, subtotal hysterectomy was associated with lower quantified blood loss (aGMR 0.36; 95% CI 0.32–0.41; p-value < 0.001), shorter operative time (aGMR 0.54; 95% CI 0.50–0.59; p-value < 0.001), and lower odds of bladder injury (aOR 0.19; 95% CI 0.05–0.68; p-value = 0.013). Table 5Adjusted surgical outcome analyses comparing subtotal versus total hysterectomy in PASOutcomeStatistical ModelAdjusted Effect (Subtotal vs. Total)95% CIp-valueQuantified blood loss (L)Log-linear (aGMR)0.360.32–0.41< 0.001Operative time (hours)Log-linear (aGMR)0.540.50–0.59< 0.001Bladder injuryLogistic regression (aOR)0.190.05–0.68= 0.013Models adjusted for surgical urgency, antenatal PAS severity, suspected cervical involvement, prior caesarean deliveries, gestational age, and surgeon identifier. Robust standard errors applied. Sensitivity analyses (exclusion of Quantified blood loss (QBL) >6 L) produced consistent results. aGMR is reported for log-transformed outcomes (blood loss and operative time)Abbreviations: aGMR Adjusted geometric mean ratio, aOR Adjusted odds ratio, CI Confidence interval, PAS Placenta accreta spectrum Table 5Adjusted surgical outcome analyses comparing subtotal versus total hysterectomy in PASOutcomeStatistical ModelAdjusted Effect (Subtotal vs. Total)95% CIp-valueQuantified blood loss (L)Log-linear (aGMR)0.360.32–0.41< 0.001Operative time (hours)Log-linear (aGMR)0.540.50–0.59< 0.001Bladder injuryLogistic regression (aOR)0.190.05–0.68= 0.013Models adjusted for surgical urgency, antenatal PAS severity, suspected cervical involvement, prior caesarean deliveries, gestational age, and surgeon identifier. Robust standard errors applied. Sensitivity analyses (exclusion of Quantified blood loss (QBL) >6 L) produced consistent results. aGMR is reported for log-transformed outcomes (blood loss and operative time)Abbreviations: aGMR Adjusted geometric mean ratio, aOR Adjusted odds ratio, CI Confidence interval, PAS Placenta accreta spectrum Adjusted surgical outcome analyses comparing subtotal versus total hysterectomy in PAS Models adjusted for surgical urgency, antenatal PAS severity, suspected cervical involvement, prior caesarean deliveries, gestational age, and surgeon identifier. Robust standard errors applied. Sensitivity analyses (exclusion of Quantified blood loss (QBL) >6 L) produced consistent results. aGMR is reported for log-transformed outcomes (blood loss and operative time) Abbreviations: aGMR Adjusted geometric mean ratio, aOR Adjusted odds ratio, CI Confidence interval, PAS Placenta accreta spectrum Discussion In 2018, the International Federation of Gynecology and Obstetrics (FIGO) recommended total hysterectomy as the preferred surgical approach for managing placenta previa associated with placenta accreta spectrum (PAS), particularly in cases of placenta increta and percreta. This recommendation was based on concerns regarding persistent hemorrhage and the potential for malignant transformation in the retained cervical stump following subtotal hysterectomy, which would necessitate long-term cervical surveillance [11]. Similarly, the American College of Obstetricians and Gynecologists (ACOG) also endorsed total hysterectomy as the standard treatment for placenta percreta in its 2018 guidance [1]. Our study demonstrates that subtotal hysterectomy effectively controlled hemorrhage in 121 patients with all PAS subtypes, including increta and percreta, in both elective and emergent settings. Maternal morbidity was lower in the subtotal group compared to the total hysterectomy group. Intraoperative time and blood loss were further increased in the total hysterectomy group due to a higher incidence of bladder injuries, which prolonged the operation and contributed to greater coagulopathy and hemorrhage (Figs. 2 and 3). Fig. 2A. Intraoperative blood loss [adjusted geometric mean ratio 0.36 (95% CI 0.32–0.41), p-value < 0.001]. B. Operative time [adjusted geometric mean ratio 0.54 (95% CI 0.50–0.59), p-value < 0.001] A. Intraoperative blood loss [adjusted geometric mean ratio 0.36 (95% CI 0.32–0.41), p-value < 0.001]. B. Operative time [adjusted geometric mean ratio 0.54 (95% CI 0.50–0.59), p-value < 0.001] Fig. 3Number of bladder injuries among patients undergoing subtotal versus total hysterectomy for placenta accreta spectrum. Adjusted odds ratio (aOR) for bladder injury with subtotal vs. total hysterectomy: 0.19 (95% CI 0.05–0.68), p-value = 0.013, derived from multivariable logistic regression Number of bladder injuries among patients undergoing subtotal versus total hysterectomy for placenta accreta spectrum. Adjusted odds ratio (aOR) for bladder injury with subtotal vs. total hysterectomy: 0.19 (95% CI 0.05–0.68), p-value = 0.013, derived from multivariable logistic regression Previous studies have reported reoperation rates of approximately 5% and maternal mortality of about 2% in the UK [13]. Our cohort experienced no reoperation or maternal deaths (Tables 3 and 4). This favorable outcome can be attributed to the management by a specialized multidisciplinary team (MDT) with extensive PAS expertise, underscoring the importance of care in centers of excellence. Given that the surgical approach was determined intraoperatively rather than through random assignment, the results presented here reflect real-world clinical decision-making in PAS management. Accordingly, the interpretation of these adjusted outcomes must consider that subtotal hysterectomy was, at times, selected when total hysterectomy was judged technically unsafe or associated with a high anticipated risk of bladder injury or hemodynamic instability. Such intraoperative selection may contribute, in part, to the differences observed between groups. Prospective multicenter studies with standardized operative criteria are warranted to minimize selection bias and validate these findings. We have not encountered any cases with definite placental involvement of cervical tissue. We recognize that this observation is based on our own operative experience, and we present it to share our findings with obstetric colleagues. These findings may support the idea that placenta accreta spectrum (PAS) may not primarily result from true trophoblastic invasion into the myometrium or cervix. Rather, it appears to be related to abnormal adhesions between the thinned lower uterine segment and the posterior bladder wall, likely secondary to repeated caesarean deliveries. This process is further exacerbated by increased vascularity in both the uterus and adjacent bladder tissues. Jauniaux et al. reported that in cases of accreta placentation, there is no histological evidence to support abnormal invasiveness of the extravillous trophoblast or penetration of chorionic villi through the uterine serosa into the pelvic cavity. Instead, the risk of placenta accreta spectrum (PAS) appears to be determined by structural factors, including the size of the caesarean scar defect, the volume of placental tissue implanted within the scar, and the residual myometrial thickness at the site of implantation. These variables influence the proximity of the placental basal plate to the uterine serosa, thereby modulating the likelihood of abnormal placental adherence [14]. Future research should focus on prospective, multicenter studies with standardized surgical selection protocols and long-term follow-up to confirm these findings and better define the optimal surgical approach for different PAS subtypes and clinical scenarios. Study limitations This study has several limitations. First, its retrospective, single-centre design introduces potential selection bias, as the choice between subtotal and total hysterectomy was determined intraoperatively by surgeon judgment rather than randomization. Second, no a priori sample size calculation was performed, reflecting the retrospective nature of the study. Third, although pre- and postoperative PAS management followed an institutional protocol, the absence of a standardized institutional consensus regarding surgical approach selection may limit generalizability to other settings. Fourth, quantified blood loss was estimated using a combination of suction volumes and weighed sponges, which may be subject to measurement variability. Fifth, some outcomes, including intensive care unit admission and blood transfusion, could not be modelled due to minimal or no variability across the cohort. Sixth, misclassification of PAS severity is possible, as subclassification relied primarily on antenatal imaging and intraoperative findings; histopathological confirmation was not consistently definitive. Conclusion In this tertiary referral center, subtotal hysterectomy was associated with lower intraoperative blood loss, shorter operative time, and fewer urologic injuries compared with total hysterectomy, without compromising maternal outcomes. These findings suggest that subtotal hysterectomy is a safe and effective option for placenta accreta spectrum when selected intraoperatively by experienced teams, particularly in cases where extensive bladder dissection is anticipated. While our results support the use of subtotal hysterectomy in both elective and emergency settings, prospective multicenter studies are needed to confirm these observations and inform standardized surgical guidance in PAS management.