- Academic Editor
†These authors contributed equally.
Background: Placenta accreta spectrum (PAS) can easily lead to life-threatening hemorrhage. However, the association between placental thickness (PT) and massive bleeding remains unclear. Thus, this study investigated the association between PT and massive bleeding to determine which patients with suspected PAS and placenta previa were more likely to experience intraoperative hemorrhage. Methods: This retrospective cohort study was conducted between January 2018 and December 2020 at a general tertiary care hospital in Chongqing, China. Covariates included demographic, clinical, and ultrasonographic characteristics. Logistic regression analysis was used to explore the association between PT and massive bleeding. A sensitivity analysis was conducted by detecting trends in the association between PT quartile and massive bleeding risk. Results: PT was associated with a risk of massive intraoperative bleeding. The sensitivity analysis yielded a similar result using the minimally adjusted model (p for trend = 0.001), and minimal changes were observed using the crude and fully adjusted models (p for trend = 0.001 and 0.037, respectively). The risk of major bleeding was significantly higher in the fourth quartile (Q4) versus first quartile (Q1) group (odds ratio = 2.26, p = 0.034). A linear relationship was observed between PT and the risk of massive bleeding. Conclusions: PT was independently and linearly associated with the risk of massive bleeding. The risk of intraoperative hemorrhage was significantly higher in the higher PT (Q4) than lower PT (Q1) group. Clinical Trial Registration: The study was registered at Chinese Clinical Trial Registry (https://www.chictr.org.cn), registration number: ChiCTR2100044798.
Placenta accreta spectrum (PAS), an abnormal placental attachment induced by the invasion of placental villi into the myometrium, is a life-threatening severe obstetric disease [1, 2, 3]. Placenta previa covers the endocervical os. In China, the frequency of PAS associated with placenta previa increases annually, with the risk of a second cesarean delivery reaching as high as 3.4% in some cases [4].
The risk of intraoperative bleeding in PAS with placenta previa is substantially higher than that in placenta previa alone. According to Bailit et al. [5], the median estimated intraoperative blood loss of PAS is 2000 mL, reaching 8000 mL in emergency surgeries. Several models have been developed to predict significant intraoperative bleeding and massive blood transfusions in obstetric patients [6, 7, 8, 9, 10]. However, these models are mainly based on patients with placenta previa, and the risk variables derived from them are inconsistent [11]. A few studies linked massive bleeding with PAS to the post-placental clear space, cross-border blood vessels in the region of sub-placental vascularity, interruption or disappearance of the bladder line, and the presence of the cervical blood sinus [4, 12].
Study has associated placental thickness (PT) in the lower uterine segment with
PAS [13]. Predictive models for PAS include PT as a risk factor [14]. Another
study suggested that PT exceeding 5 cm may be associated with intraoperative
bleeding
This retrospective cohort study was conducted at the First Affiliated Hospital of Chongqing Medical University (no. 2021–059). Our local ethics committee approved the study and waived the requirement for informed consent owing to its retrospective nature. Here we reviewed the clinical data (extracted from the medical records) of consecutive patients with PAS and placenta previa who underwent a cesarean section (CS) between January 2018 and December 2020. Ultimately, 317 patients were included in the analysis. All reports followed the Guidelines for Strengthening the Reporting of Observational Epidemiological Studies.
The preoperative diagnosis of PAS is mainly based on the Placenta Accreta Spectrum Ultrasound Scoring System (PASUSS) [4, 14] and the International Federation of Gynecology and Obstetrics (FIGO) diagnostic criteria [15]. In 2019, the diagnosis of PAS relied primarily on the PASUSS. In this study, PAS was confirmed by postoperative pathology if pathological examination was performed intraoperatively by the surgeon. Patients with PAS but no other prenatal disorders who underwent CS met the inclusion criteria. Patients with multiple pregnancies or hematologic disorders or for whom inadequate data were available were excluded. Patients with placental pathologies were also excluded.
The prevention of hemorrhage and measures to stop bleeding during CS were
performed according to standard procedures. The surgeon selected whether to use
abdominal aortic balloon placement before cesarean delivery depending on the
patient’s previous medical history and ultrasound findings. We developed a policy
of prophylactic interruption of abdominal aortic balloon catheter placement
followed by placental separation and uterine preservation during cesarean
delivery whenever possible. The decision to remove the adherent placenta and
apply a local suture in the placental bed or leave the placenta in situ
and perform a hysterectomy was made after delivery of the newborn depending on
the operative findings. The triple-P procedure was used in some patients to
preserve the uterus whenever possible. If the placenta could not be removed or
the balloon occlusion failed to control the hemorrhage, a subtotal hysterectomy
was performed. Excising part of the placenta, manually detaching the placenta,
ligating the bleeding artery, placing a
The primary outcome indicator was the amount of intraoperative bleeding.
Intraoperative blood loss was calculated based on pre-and postoperative
hemoglobin values [17]. Women with a calculated blood loss
Measurements were performed only on transabdominal and transvaginal images obtained in the longitudinal plane [19]. The PT was measured at the thickest portion by an obstetric ultrasound-specialized radiologist with 5 years of post-fellowship experience. If the placental cord insertion was centric or eccentric by 2 cm, the PT was measured in the sagittal view in the area of placental cord insertion. When the placental cord insertion was marginal or velamentous, the maximum PT in the longitudinal plane was calculated linearly. The PT was measured the day before or the day of surgery. PT is also expressed in quartiles, where Q1 represents the first quartile, Q2 represents the second quartile, etc.
The patients’ clinical characteristics were collected by review of their medical records during hospitalization for CS, including age, weight, height, gestational weeks, gravidity, curettage, previous number of CS, intraoperative estimated blood loss, incidence of transfusion, incidence of intensive care unit (ICU) admission, incidence of hysterectomy, operation time, and hospitalization duration. Furthermore, neonatal outcomes were assessed.
All ultrasound-related indicators including PT and cervical canal length were extracted from the latest preoperative obstetric ultrasound reports.
Those for whom PT was unmeasured were excluded, while missing values for
indicators such as cervical canal length were interpolated using the mean values.
Continuous data are presented as mean
Subsequently, smooth curve fitting was performed to determine the association between the risk of massive bleeding and PT. A univariate analysis was performed to assess candidate variables as risk factors for massive bleeding; associations between potential risk factors and outcomes are shown as odds ratio (OR) and 95% confidence interval (CI). We simultaneously show the results of unadjusted and minimally adjusted analyses (age, weight, height, and gestational age) and those from fully adjusted analyses (age, weight, height, gestational age, previous CS, gravidity, cervical canal length, and intra-aortic balloon occlusion). Moreover, a sensitivity analysis was performed to detect trends in the association between PT quartile and massive bleeding risk. A subgroup analysis of the PAS categories was conducted using a stratified logistic regression model. All analyses were performed using R statistical software (R 4.1.2, http://www.R-project.org; the R Foundation) and a free statistics analysis platform.
During the study period, 337 patients with suspected PAS with placenta previa underwent cesarean deliveries at the institution. A total of 317 patients were included in the analysis (Fig. 1). Of them, 133 (41.9%) experienced massive intraoperative bleeding. Moreover, 78 (24.6%) required a transfusion of two or more blood product units (fresh-frozen plasma and/or red blood cells), 12 (3.8%) were admitted to the ICU, 11 (3.5%) required a hysterectomy, and 25 (7.9%) underwent angiographic embolization. The baseline clinical and ultrasonic characteristics of the intra-and postoperative indicators are presented in Tables 1,2. Among the PT group, a significant difference was noted in gestational age, cervical canal length, and PAS category. Bleeding, the probability of blood component transfusion, surgical duration, and hospitalization duration differed between the PT groups. A linear association was observed between the likelihood of massive bleeding and PT (Fig. 2).
Flowchart of the study population. PAS, placenta accreta spectrum.
Relationship between placental thickness and risk of massive bleeding.
Variables | Total (n = 317) | Q1 (n = 68) | Q2 (n = 70) | Q3 (n = 90) | Q4 (n = 89) | p-value | |
Age (year), Mean |
32.9 |
33.1 |
33.4 |
32.4 |
33.0 |
0.545 | |
Height (centimeter), Mean |
157.9 |
157.6 |
158.1 |
157.6 |
158.1 |
0.864 | |
Weight (kg), Mean |
68.4 |
68.0 |
68.8 |
67.8 |
68.9 |
0.814 | |
Gravidity, Mean |
3.9 |
3.4 |
4.0 |
3.9 |
4.0 |
0.128 | |
Previous CS, n (%) | 0.418 | ||||||
0 | 74 (23.3) | 19 (27.9) | 14 (20) | 25 (27.8) | 16 (18) | ||
1 | 194 (61.2) | 40 (58.8) | 47 (67.1) | 49 (54.4) | 58 (65.2) | ||
2 | 48 (15.1) | 8 (11.8) | 9 (12.9) | 16 (17.8) | 15 (16.9) | ||
1 (0.3) | 1 (1.5) | 0 (0) | 0 (0) | 0 (0) | |||
Curettage, n (%) | 0.126 | ||||||
0 | 69 (21.8) | 21 (30.9) | 15 (21.4) | 18 (20) | 15 (16.9) | ||
1 | 84 (26.5) | 10 (14.7) | 20 (28.6) | 25 (27.8) | 29 (32.6) | ||
2 | 81 (25.6) | 18 (26.5) | 13 (18.6) | 29 (32.2) | 21 (23.6) | ||
83 (26.2) | 19 (27.9) | 22 (31.4) | 18 (20) | 24 (27) | |||
Gestational age (week), Mean |
36.3 |
36.1 |
36.4 |
36.9 |
35.9 |
||
Diabetes, n (%) | 83 (26.2) | 22 (32.4) | 20 (28.6) | 24 (26.7) | 17 (19.1) | 0.276 | |
Hypertension, n (%) | 11 (3.5) | 7 (10.3) | 1 (1.4) | 1 (1.1) | 2 (2.2) | 0.013 | |
Suspected PAS, n (%) | |||||||
No PAS | 18 (5.7) | 16 (23.5) | 2 (2.9) | 0 (0) | 0 (0) | ||
Accreta | 103 (32.5) | 24 (35.3) | 19 (27.1) | 41 (45.6) | 19 (21.3) | ||
Increta | 170 (53.6) | 27 (39.7) | 43 (61.4) | 44 (48.9) | 56 (62.9) | ||
Percreta | 26 (8.2) | 1 (1.5) | 6 (8.6) | 5 (5.6) | 14 (15.7) | ||
Cervical canal length (mm), Mean |
32.6 |
34.7 |
32.1 |
33.0 |
30.8 |
SD, standard deviation; CS, cesarean section; Q1, the first quartile; Q2, the second quartile; Q3, the third quartile; Q4, the fourth quartile; PAS, placenta accreta spectrum.
Variables | Total (n = 317) | Q1 (n = 68) | Q2 (n = 70) | Q3 (n = 90) | Q4 (n = 89) | p-value | |
Blood loss (mL), Median (IQR) | 611.0 (436.0, 1124.0) | 546.0 (412.8, 1009.0) | 566.0 (412.0, 1089.0) | 573.0 (423.2, 1008.0) | 1024.0 (554.0, 1578.0) | ||
IABO, n (%) | 168 (53.0) | 21 (30.9) | 40 (57.1) | 45 (50) | 62 (69.7) | ||
Blood transfusion, n (%) | 78 (24.6) | 11 (16.2) | 15 (21.4) | 12 (13.3) | 40 (44.9) | ||
Blood transfusion (Unit), Median (IQR) | 0.0 (0.0, 0.0) | 0.0 (0.0, 0.0) | 0.0 (0.0, 0.0) | 0.0 (0.0, 0.0) | 0.0 (0.0, 3.0) | ||
Plasma, n (%) | 55 (17.4) | 4 (5.9) | 12 (17.1) | 8 (8.9) | 31 (34.8) | ||
Cryoprecipitate, n (%) | 18 (5.7) | 1 (1.5) | 5 (7.1) | 2 (2.2) | 10 (11.2) | 0.029 | |
Autologous blood collection, n (%) | 43 (13.6) | 4 (5.9) | 9 (12.9) | 7 (7.8) | 23 (25.8) | ||
Postoperative transfusion, n (%) | 60 (18.9) | 6 (8.8) | 18 (25.7) | 10 (11.1) | 26 (29.2) | ||
Uterine gauze stuffing, n (%) | 45 (14.2) | 6 (8.8) | 5 (7.1) | 12 (13.3) | 22 (24.7) | 0.006 | |
Intrauterine balloon tamponade, n (%) | 111 (35.0) | 28 (41.2) | 16 (22.9) | 35 (38.9) | 32 (36) | 0.097 | |
Uterine artery embolization, n (%) | 25 (7.9) | 2 (2.9) | 5 (7.1) | 4 (4.4) | 14 (15.7) | 0.01 | |
Uterine bondage, n (%) | 141 (44.5) | 25 (36.8) | 33 (47.1) | 41 (45.6) | 42 (47.2) | 0.542 | |
Cervical lift suture, n (%) | 226 (71.3) | 51 (75) | 48 (68.6) | 59 (65.6) | 68 (76.4) | 0.349 | |
Triple-P procedure, n (%) | 58 (18.3) | 8 (11.8) | 14 (20) | 19 (21.1) | 17 (19.1) | 0.459 | |
Placenta left in situ, n (%) | 21 (6.6) | 2 (2.9) | 2 (2.9) | 7 (7.8) | 10 (11.2) | 0.113 | |
Hysterectomy, n (%) | 11 (3.5) | 0 (0) | 5 (7.1) | 1 (1.1) | 5 (5.6) | 0.028 | |
ICU admission, n (%) | 12 (3.8) | 0 (0) | 4 (5.7) | 2 (2.2) | 6 (6.7) | 0.088 | |
Operation time (min), Median (IQR) | 74.0 (53.0, 98.0) | 66.0 (50.5, 86.0) | 68.5 (55.0, 90.0) | 62.5 (48.5, 91.0) | 91.0 (69.0, 117.0) | ||
Hospitalization time (day), Median | 6.0 (5.0, 7.0) | 6.0 (4.0, 7.0) | 6.0 (5.0, 7.0) | 5.0 (4.0, 7.0) | 7.0 (5.0, 9.0) | ||
Neonatal sex, n (%) | 0.259 | ||||||
male | 169 (53.3) | 39 (57.4) | 33 (47.1) | 54 (60) | 43 (48.3) | ||
female | 148 (46.7) | 29 (42.6) | 37 (52.9) | 36 (40) | 46 (51.7) | ||
Neonatal weight (g), Mean |
2880.9 |
2874.9 |
2864.5 |
3039.9 |
2737.6 |
IABO, Intra-Abdominal Balloon Occlusion; ICU, intensive care unit; IQR, Interquartile range; SD, standard deviation; Q1, the first quartile; Q2, the second quartile; Q3, the third quartile; Q4, the fourth quartile.
The risk of massive intraoperative bleeding increased proportionally with PT (Supplementary Table 1). Univariate analysis indicated that cervical canal length, previous CS, and gestational age were associated with massive intraoperative bleeding.
In the crude model, PT was associated with high bleeding risk. Moreover, similar results were obtained using the minimally and fully adjusted models. In the sensitivity analysis, PT was a categorical variable (quartile), and the same trend was observed. In the fully adjusted model, a higher PT (Q4) was associated with a 126% increased risk of major bleeding compared to a lower PT (Q1) (OR: 2.26; p = 0.034) (Table 3).
Index | Non-adjusted Model | Model I | Model II | ||||
OR (95% CI) | p-value | OR (95% CI) | p-value | OR (95% CI) | p-value | ||
PT | 1.06 (1.03–1.09) | 1.06 (1.03–1.09) | 1.04 (1.01–1.07) | 0.008 | |||
PT quartile | |||||||
Q1 | 1 (Reference) | 1 (Reference) | 1 (Reference) | ||||
Q2 | 1.6 (0.79–3.25) | 0.193 | 1.7 (0.83–3.49) | 0.149 | 1.08 (0.49–2.37) | 0.843 | |
Q3 | 1.2 (0.61–2.37) | 0.6 | 1.39 (0.69–2.81) | 0.36 | 0.87 (0.4–1.88) | 0.721 | |
Q4 | 3.88 (1.98–7.62) | 3.86 (1.95–7.65) | 2.26 (1.07–4.8) | 0.034 | |||
p for trend | 0.037 |
As all Variance Inflation Factor (VIF) values were
Model I: Adjust for age, weight, height, and gestational age;
Model II: Adjust for age, weight, height, and gestational age, previous CS,
gravidity, cervical canal length, IABO.
Stratified analyses of the association between PT and massive intraoperative bleeding are presented in Table 4. The relationship between PT and massive intraoperative bleeding risk did not differ significantly among the subgroups based on disease category. The effect of PT on massive intraoperative bleeding was stable in all subgroups. The multiplicative interactions among the various groups regarding the risk of massive incident intraoperative bleeding were not significant.
Subgroup | Variable | No. total | No. event (%) | OR (95% CI) | p-value | p for interaction |
No PAS | PT | 18.0 | 0 (0) | 1 (0–Inf) | 1.0 | 0.524 |
Accreta | PT | 103.0 | 5 (4.9) | 0.92 (0.79–1.08) | 0.312 | |
Increta | PT | 170.0 | 103 (60.6) | 1.03 (1–1.07) | 0.091 | |
Percreta | PT | 26.0 | 25 (96.2) | 1.06 (0.83–1.34) | 0.655 |
PT, placental thickness; OR, odds ratio; CI, confidence interval.
This retrospective cohort study included 317 patients with suspected PAS and placenta previa and examined the independent association between PT and intraoperative hemorrhage. Our study found that an increasing PT was associated with an increased risk of massive bleeding in patients with PAS placenta previa who underwent CS. We found that PT was significantly increased in patients with massive bleeding. To the best of our knowledge, this is the first study to report the association between PT and massive bleeding in patients with PAS and placenta previa.
Curve fitting revealed that intraoperative bleeding increased as PT increased. The relationship between PT and intraoperative hemorrhage was linear. In the subgroup analyses, the correlation between PT and the risk of major bleeding was not statistically different. This further suggests that the risk of major bleeding is related to the PAS type, with higher PT values predicting a higher PAS grade. When it can be accurately determined preoperatively, PT can assist in determining the risk of intraoperative hemorrhage.
Massive bleeding is the major cause of preventable peripartum mortality. As much
as 70–93% of these deaths are deemed preventable [20]. A rapid diagnosis and
early multidisciplinary management can improve the maternal prognosis. The early
identification of the risk of hemorrhage facilitates appropriate preoperative
countermeasures such as cross-matching, establishing extensive venous access, and
autologous blood recovery [21, 22]. Most cases of massive bleeding during CS are
caused by placental abnormalities. These risk factors include maternal history
and placental characteristics observed on ultrasonography [23]. The clinical
features included maternal age, gestational age
However, different PAS scores are associated with different bleeding risks. Wang et al. [4] showed that the risk of intraoperative hemorrhage in PAS is associated with the post-placental clear space, cross-border blood vessels in the region of sub-placental vascularity, interruption or disappearance of the bladder line, and presence of the cervical blood sinus. These ultrasound features are specific to some patients with PAS and not universal. An experienced sonographer may obtain these indicators; however, subjective factors cannot be excluded [13]. When diagnosed with suspected PAS, a more objective indicator for predicting the risk of major bleeding is needed.
Many studies have used PT to predict fetal and maternal health [24, 25]. The condition of the placenta can be a predictor of many events that occur in the mother and fetus [26]. Another study demonstrated a correlation between the PT and neonatal weight [27]. Li et al. [28] showed a correlation between PT and PAS. A threshold of 4.5 cm can help screen for patients with placenta previa and risk factors for PAS. There was a trend toward an increased PT in patients with different PAS types. PT was significantly greater in the penetrating PAS group than in the other two types. This may be related to an increased number of lacunae in the placenta. When more severe placental implantation occurs, lacunae of varying sizes develop within the placenta, thereby increasing the PT. The proportion of intra-placental lacunae increases with the degree of placental implantation. An increase in PT suggests a significantly higher likelihood of implantation or penetration. The underlying cause of an abnormally thick PT in women with PAS is unclear. However, this may be related to poor vascularization of the lower uterus and cervix as well as placental implantation in the cesarean scar, which limits placental migration [29]. One study showed that antenatally measured PT has a linear correlation with blood loss in patients with PAS [30]. Our study confirmed this finding. An increased PT suggests increased placental implantation and invasion into the myometrium and corresponding vessels. This may affect uterine contractions and complicate surgical hemostasis when the placenta peels off.
This study has some limitations. Moreover, it is limited by factors that limit all retrospective designs such as confounders and selection bias. The study was retrospective; therefore, PT was not measured prospectively. All ultrasound-related metrics were obtained by reviewing the patients ultrasound reports; therefore, some of the ultrasound metrics were not presented in the ultrasound reports, such as intraplacental hypoechogenicity. Therefore, relevant indicators were not included in the regression model, which may have been a confounding factor affecting the final results.
In conclusion, an increased PT in the lower uterine segment is associated with intraoperative hemorrhage in patients with PAS and placental previa. Preoperatively, an increased PT can provide an indicator for the assessment of intraoperative bleeding risk. The risk of intraoperative hemorrhage was significantly higher in the higher PT (Q4) than lower PT (Q1) group. A thickened PT seems to be a warning sign of massive bleeding during surgery for PAS with placenta previa.
The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.
FH participated in the design of the study, data collection and analysis, and manuscript drafting. JW and KW designed the study and drafted and revised the manuscript. TY, WW, YX, QLX, QJX, and JZ acquired and interpreted the data and followed the patients. QJX and JZ performed the data analysis and interpretation. All authors have read and approved the final version of the manuscript. All authors contributed to the editorial revision of the manuscript and agreed to take responsibility for all aspects of this work.
Ethical approval was obtained from the Ethics Committee of First Affiliated Hospital of Chongqing Medical University (approval number: 2021–059). All procedures were performed in accordance with the principles of the Declaration of Helsinki. The Ethics Committee of the First Affiliated Hospital of Chongqing Medical University waived the requirement for informed consent. The medical records used in this study were obtained from previous diagnoses and treatments. Exemption from informed consent will not adversely affect the rights or health of the subjects. The participants’ privacy and personal identities were protected.
Not applicable.
This research received no external funding.
The authors declare no conflict of interest.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.