We retrospectively investigated data of all consecutive patients diagnosed with PAS overlying the cesarean scar and treated in accordance with the eight-step protocol between December 2015 and October 2019 in our hospital. Information regarding history of CS, gestational week at surgery, surgery duration, EBL, bladder injury, and post-procedural recovery was retrieved from the hospital database. The inclusion criteria were as follows: one or more previous CS, placenta previa overlying the previous cesarean scar, and PAS confirmed by intraoperative examination. Those with spontaneous placental abruption were excluded.
Perioperative sonography was performed to determine the location and invasion of the placenta and the length and shape of the cervix. Cervical involvement was determined when the placenta covered the internal cervical orifice and blood flow could be detected inside the shortened cervical canal (length, <2 cm).15 Cases with two or more of the following signs were considered to be severe: patients who had two or more previous CS, cervical involvement, complete placenta previa, placental lacunae and turbulence, loss of myometrial interface with a width >3 cm, bladder wall interruption, or uterovesical hypervascularity, and a preoperative aortic balloon was recommended.
The depth and extent of placental invasion and severity of pelvic adhesion were determined intraoperatively and reassessed according to the International Federation of Gynecology and Obstetrics classification for the clinical diagnosis of PAS disorders.16 Placenta accreta was diagnosed when manual removal of the placenta was required. Placenta increta was diagnosed when part of the placenta was cut with scissors, and the remaining part of the uterine wall was thinner than the adjacent parts. Placenta percreta was diagnosed when the placental tissue had penetrated through the serosa of the uterus with hypervascularity, or a clear surgical plane could not be identified between the bladder and uterus. The extent of placental invasion was graded based on the involved area as: grade 1, no invasion; grade 2, estimated extension <1/6 of the placenta; grade 3, estimated extension 1/6–1/3 of the placenta; and grade 4 or extensive invasion, estimated extension >1/3 of the placenta. The location of the placenta was classified as mainly anterior, mainly posterior, or central. EBL was evaluated by recording the volume of blood in the vacuum pump and calculating the difference in the weight of the gauze and surgical drape pre- and postoperatively, with each extra gram of weight being estimated as approximately 1 mL of blood. Severe pelvic adhesion was determined when the border of the uterus was unclear and required elaborate dissection to be separated from the surrounding organs.
Data were analyzed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as means and standard deviations, whereas categorical variables are expressed as absolute numbers and percentages. Several variables including maternal age, number of previous CS, gestational week at surgery, emergency surgery, preoperative aortic balloon, depth and extent of placental invasion, main location of the placenta, pelvic adhesion, and cervical involvement, were studied using the Spearman correlation analysis to determine the factors correlated with EBL and were further evaluated with a multiple linear regression analysis to determine their independence in relation to EBL. An independent t-test was used to compare the EBL in severe cases from the balloon and non-balloon groups. Differences were considered significant when p values were <0.05.
Elective CS was planned after 34–37 weeks of gestation or after necessary preparations when the patient was transferred from a local hospital after 37 weeks. Before surgery, the hemoglobin levels were increased to >100 g/L by blood transfusion. The possibility of intractable blood loss and hysterectomy were discussed with the family, and written consent was obtained. Combined lumbar anesthesia was performed in patients without any risk factors, based on the preoperative sonography, and general anesthesia was performed in the others. The study complied with the Declaration of Helsinki, and all human subjects provided written informed consent with guarantees of confidentiality.
The incision was made on the previous CS scar to avoid a cruciate incision unless the original transverse incision was very low. An incision sleeve was used for better exposure of the surgical field in both transverse and longitudinal incisions. The following eight-step procedure was then performed:
- Open the uterovesical serosa and try to separate the bladder from the lower uterine segment. If the procedure is difficult owing to severe adhesion and bleeding, this procedure can be performed post-delivery.
- Cut the uterine wall while avoiding the placenta, if possible. Otherwise, decisively cut the placenta by hand to deliver the baby as soon as possible. Exteriorize the uterus from the pelvic cavity and grasp the lower uterine segment with one hand to block bleeding. Detach the bulk of the placenta with the other hand and then clamp the lower uterine segment on both sides with ovum forceps (Figure 1). Bleeding usually attenuates dramatically.
- While the assistant extracts the uterus, use the left hand to replace one of the ovum forceps and push the bladder sufficiently down. Ligate the uterine artery and simultaneously compress the lateral part of the lower uterine segment. Do this thrice at 2–3-cm intervals in the upward direction on each side (suture A) (Figure 2). Bleeding will attenuate remarkably.
- Perform several full-thickness sutures on the anterior lower uterine segment (suture B) (Figure 2).
- Perform several full-thickness horizontal sutures on the posterior lower uterine segment (suture C) (Figure 3).
- Scrape the remaining placenta and cut around the uterine incision if the uterine wall has been penetrated or is very thin (Figure 2), which will also decrease the uterine volume and help with hemostasis.
- Perform another long-step suture on each end of the uterine incision (suture D) (Figure 2). Additional local hemostatic sutures on the uterine wall may be needed. In most cases, complete hemostasis can be achieved. However, bleeding from the corpus may occur due to uterine inertia and coagulopathy after massive blood loss.
- Perform vertical compression suture in the uterine corpus (suture E) (Figure 2). Compression on the uterine body to tighten the suture is crucial. Usually, five stitches for a term-pregnancy uterus and three stitches for a smaller uterus are needed. These sutures will compress the upper part of the uterus and simultaneously allow drainage. The tiny space on the top of the uterus is used to prevent enclosure of the bowel as the uterus later shrinks.
All possible bleeding sites in the entire uterus are subsequently compressed with sutures. The uterine incision is then closed and embedded carefully. Finally, the abdominal wall is closed after extensive hemostasis in the vesico-uterine pouch.
Large-step sutures are recommended to achieve rapid hemostasis and avoid potential ischemic necrosis in the uterus. Timely blood infusion is also important. Otherwise, hypoperfusion and coagulopathy after instant blood loss can lead to uterine inertia and intractable bleeding in the uterine corpus. The vertical compression suture in the uterine corpus is the only effective procedure to stop this vicious cycle.