Introduction

Placenta accreta spectrum (PAS) is a life-threatening obstetric complication characterised by abnormal placental attachment to the uterine wall1. Delivery, typically via caesarean section (c-section), is associated with significant maternal morbidity, including postpartum haemorrhage, bladder injury, and the potential need for hysterectomy2,3,4,5,6. The highest risk of PAS occurs in women with a history of previous c-section and placenta previa, where a defect in the endometrial-myometrial interface allows abnormally deep anchoring of placental villi around the uterine scar2,7,8. Depending on the degree of myometrial invasion, the spectrum encompasses placenta accreta (villi attached to the myometrium), placenta increta (villi invade the myometrium), and placenta percreta (villi invade the serosa)2,7. The rising rate of c-section has paralleled an upward trend in PAS incidence, with the risk increasing progressively with each subsequent operative delivery9,10,11.The management of PAS remains challenging and calls for more innovative solutions. Interventional radiology procedures such as uterine artery embolisation, can help reduce blood loss and preserve the uterus in PAS. However, they carry a risk of failure and accessibility to these techniques is not universal12. Surgical management is more practical and includes (a) peripartum hysterectomy (either at the time of caesarean or delayed), (b) conservative management with the placenta left in situ, and (c) segmental uterine wall resection with subsequent reconstruction13. While the latter two aim to preserve the uterus, the elevated risks of secondary haemorrhage and infection associated with leaving the placenta in situ render segmental resection a more useful approach. The fundamental concept of segmental uterine resection for PAS management was introduced by Palacios-Jaraquemada in 200513 and laid the groundwork of the Triple-P procedure (involving placental localisation, stepwise devascularisation, and placental bed excision) for PAS surgery by Chandraharan et al. in 201214. Several modifications to the Triple-P procedure have emerged, such as transverse and vertical haemostatic sutures and cervical inversion techniques, enabling better haemostasis15,16,17,18. A key step in Triple-P is thorough bladder dissection, for adequate access to secure lower uterine segment haemostasis15,16,17,18. In this context, blunt dissection by lateral window technique has a lower risk of vascular disruption compared to sharp dissection19. This technique relies on identifying the lateral vesicouterine border to create a window in the anterior leaf of the broad ligament. In PAS C-section, the vesicouterine border is obscured by placental vessels and adhesions from previous C-sections, contributing to persistent bladder injury rates (8.8%,19%, 20.6% in different studies)20,21,22.

To minimize bladder injury, retrograde infusion of normal saline, also known as Cystoinflation, is a recently reported intervention for difficult caesarean deliveries3,23.The rationale for adding Cystoinflation to the Triple-P technique lies in its ability to demarcate bladder outline, thus providing clear visualization of the vesicouterine border and allowing for mobilization of the bladder in correct dissection plane. The tactile feeling of the distended bladder helps recognise and protect the bladder during blunt dissection12. Further, Cystoinflation makes the surface vessels of the placenta more prominent, aiding in their precise ligation before incision24.  Mobilising the bladder within the appropriate anatomical plane provides proper exposure of the lower uterine segment and uterine vasculature, thereby facilitating haemostatic manoeuvres during the Triple P procedure. In this study, we evaluated the efficacy of adding Cystoinflation to the modified Triple-P procedure to prevent bladder injury and secure haemostasis in women with PAS. We hypothesized that this combined approach could concomitantly reduce bladder injury, blood loss, and the need for hysterectomy, leading to improved fetomaternal outcomes and future fertility.

Results

The Enrolment of subjects in the study is shown in Fig. 1.

Fig. 1
figure 1

CONSORT diagram. Patient assignment to groups, follow up and analysis.

Full size image

From May 1, 2017 to April 30, 2019, we recruited one hundred and fifty healthy pregnant women, with 75 subjects in each group, who met the inclusion criteria. Each subject underwent short-term follow-up for three months to monitor urological complications and neonatal outcomes (the final subject’s follow-up was completed on July 31, 2019) and long-term follow-up for five years to assess future fertility in couples desiring conception (completed on April 30, 2024). There were three patients in the control group and four patients in the Cystoinflation group were lost to follow-up after discharge from the hospital. Their data were included in evaluation of primary outcome and secondary outcome during hospital stay as well as composite outcome in intention to treat analysis. The characteristics of the study population are shown in Table 1.

Table 1 Demographic features of the study population.
Full size table

In this study, the Cystoinflation and control groups demonstrated similar baseline characteristics. Specifically, there were no significant intergroup differences in mean age (Cystoinflation: 32.2 ± 3.2 years; control: 32.0 ± 2.8 years; P = .21), gestational age (P = .931), parity (P = .978), body-mass index (P = .814), socioeconomic status (P = .64), or prior caesarean deliveries (P = .94). The maximum number of prior caesarean deliveries was four in each group. Furthermore, the distribution of placenta accreta spectrum was comparable between groups with majority belonging to placenta increta and percreta (P = .31 and P = .80, respectively). Preoperative haemoglobin levels and preoperative postmicturition bladder volumes were also similar between the groups (P = .81). Details of primary and secondary outcomes are presented in Table 2.

Table 2 (a) Primary outcome, (b) secondary outcome.
Full size table

The primary and secondary outcomes of the study revealed significant differences between the Cystoinflation and the control group. The bladder injury rate was significantly lower in Cystoinflation group compared to control, 2(2.7%) versus 14(18.7%); P = .001 and the size of injury(bladder injury arm) was significantly smaller (0.75 ± 0.35 versus 6.14 ± 1.10; P < .0001).The location of injuries was higher in Cystoinflation group favouring the easier and safer approach to repair (both injuries were at the dome of the bladder in Cystoinflation group versus the control group in which six injuries were at the dome of the bladder and eight injuries were at the posterior bladder wall). The ureteric and bladder trigone injuries did not occur in this study.

The number of perforations was also eleven times lesser in Cystoinflation group compared to the control group, (P = .001). The blood loss in Cystoinflation group was significantly lesser, (1632.53 ± 325.66 cc vs. 2431.60 ± 546.75 cc; P < .0001). Due to decreased blood loss, the amount of blood transfused was also reduced in Cystoinflation group compared to control (P < .0001). Only 2 patients needed hysterectomy in Cystoinflation group compared to 12 in control group. The operative time showed significant difference between the two groups, (P = .019).

Postoperatively, urinary tract infections (UTI) and micturition problems occurred only in the bladder injury arm of both groups. In bladder injury arms, mean WBC count was higher in the control group than Cystoinflation group (6.79 ± 2.41versus 4.93 ± 1.33; P = < 0.0001) with more culture-positive cases of UTI (P = .017) and fever (P = .001) respectively. The post-micturition bladder volume had no statistical intergroup differences (P = .802). Dysuria, urgency, and urinary incontinence (calculated as compost variable were significantly more common in the control group compared to the Cystoinflation group. There were two cases of vesicovaginal fistula in control group in which primary repair failed compared to none in control group although result was not significant.

In Cystoinflation group, mean duration of urinary catheterization (P < .001) and hospital stay (P < .001) were also significantly shorter compared to the control group.

Regarding neonatal outcome, APGAR score of the new-born was significantly higher in Cystoinflation group; P value < 0.0001, and neonatal intensive care unit admissions were 6.5times decreased, 2(2.7%) versus 13(17.3%); P value = 0.003. Table 3 shows delayed outcomes.

Table 3 Delayed outcome.
Full size table

The data reveals a significantly higher conception rate in the Cystoinflation group compared to the intervention group (62.8% vs. 28.6%; P = .001). While miscarriage, preterm delivery, and recurrence rates were lower in the Cystoinflation group, these differences did not reach statistical significance (P = .461, 0.143, and 0.284, respectively).

Discussion

Cystoinflation, or bladder retro-fill, is a technique used to prevent bladder injury during difficult caesarean section. In this study, the incorporation of Cystoinflation into Triple-P procedure, a conservative surgical technique for placenta accreta spectrum (PAS) cases, significantly reduced bladder injuries, blood loss, and the hysterectomy rate, thereby enhancing the procedural efficacy. Cystoinflation helps identify bladder margins23and indicates the correct site for creating the lateral window in the broad ligament, demonstrating promising results in reducing bladder injuries. Furthermore, Cystoinflation effectively minimized the size of bladder injuries by producing fluid jets when perforation is pinpoint, thereby guiding the surgeon to the correct dissection plane23. In this study, the bladder injury rate was reduced seven-fold in the Triple-P procedure by incorporating Cystoinflation, strongly supporting our hypothesis. The theoretical urological complications such as urinary tract infections, urinary retention, and other symptoms were not increased. This aligns with the primary objectives of the Triple-P approach, which seeks to optimize patient care and improve surgical outcomes in conservative management25,26,27,28,29.

In our study, we observed that major haemorrhage in PAS can occur at different stages: Haemorrhage may occur from vesicouterine surface vessels during bladder flap separation, subsequent to iatrogenic injury to placental and uterine blood vessels during uterine incision and placental separation, and as a result of vascular injury during the excision of adherent myometrium. Cystoinflation effectively stretched peritoneal vessels apart, making them prominent and easy to ligate in vesicouterine pouch, thereby preventing blood pooling. Previous PAS research has also demonstrated that surface vessels embedded within the empty bladder wall start floating on the surface of a full bladder, facilitating their ligation24. The ability to maintain a clear surgical field helps avoid major vascular injuries during uterine incision and additional interventions.

After delivery, uterine elevation through the incision helps reduce bleeding by kinking the uterine blood vessels30. Ligation of the uterine arteries, the principal blood supply to the uterus, before placental removal and excision of the adherent uterine wall further minimizes blood loss. Cystoinflation helps identify the correct plane for adequate bladder dissection from the uterus up to the cervix, facilitating safe haemostatic suturing without the risk of needle passing through the bladder wall. Appropriate bladder mobilization also enables complete repair of uterine wall tears and the application of vertical haemostatic sutures to the lower uterine segment. Figure 2 illustrates the steps of modified triple P technique where use of Cystoinflation improved bladder safety and facilitated haemostatic manoeuvres.

Fig. 2
figure 2

Conservative management of placenta accreta spectrum. Cystoinflation in Triple P technique. Haemostatic manoeuvres facilitated by meticulous bladder mobilisation using Cystoinflation. Steps where Cystoinflation is not required.

Full size image

In this study, a notable outcome is the reduction in blood loss, as haemorrhage remains a significant risk in procedures involving PAS. However, this benefit was not consistently observed in another study31. This discrepancy may be attributable to differences in the surgical approach. We delivered both the baby and placenta through a single incision and used Cystoinflation and blunt bladder dissection for safer bladder mobilization to facilitate haemostatic manoeuvres of the Modified Triple-P technique. In contrast, the other study employed separate incisions for fatal and placental delivery and tried Cystoinflation for controlling PPH and preventing bladder injury during massive blood loss31.

The mean operative time in the Cystoinflation group was significantly shorter compared to the control group, despite the additional time required for bladder filling. We observed that the prevention of bladder injuries and quicker haemostasis facilitated by Cystoinflation in the intervention group contributed to the reduced operative time and hospital stay, contrasting with the control group that required time for confirming and repairing bladder injuries, thus prolonging operative time. This finding is supported by a similar study, although the authors did not report a significant decrease in bladder injury rates32. The lower success in preventing bladder injury may be due to differences in dissection techniques. We used blunt dissection via the lateral window technique in both groups, whereas dissection method was not specified in the other study.

Routine visceral peritoneal closure is unnecessary in c-sections and may increase adhesions. However, in PAS, overrunning the bladder peritoneum helps achieve haemostasis by managing bleeding from placental sinuses started after lower segment manoeuvres, which may not respond to repeat diathermy. Therefore, it is useful to inspect the bladder peritoneum before abdominal closure and considering continuous suture above the bladder margin for haemostasis in PAS.

Our study demonstrated a significantly lower incidence of urological injuries and postoperative complications (2.7%) compared to our control group and certain earlier studies that did not report the use of the Cystoinflation technique(up to 32.4% and 39.5%)3,32. Assuming our control group effectively removed biases related to population diversity, surgeon experience, and operating room facilities, the difference in outcomes suggests that Cystoinflation may be a contributing factor. Further research is required in this direction including diverse populations and surgical setups to find out real effectiveness of Triple P combined with cystoinflation.

Ureteric injury, present as surgical ascites and diagnosed via ultrasonography or cystoscopy is reported in up to 2% of PAS surgeries—particularly with parametrical involvement33.While some advocate for preoperative ureteric stent insertion, particularly when Doppler studies suggest ureteral invasion34,35, others propose subtotal hysterectomy36. Conversely, a higher incidence of urological complications with DJ stents in PAS cases has also been reported37, and subtotal hysterectomy carries the risk of incomplete placental removal in placenta previa38. Although cystoinflation has no direct role in preventing ureteric injury, and we did not use ureteric stents, no ureteric injuries occurred in this study. This may be attributed to improved visualisation through stepwise haemostasis and careful blunt dissection with the pulp of the finger directed toward the uterus, maintaining a safe plane away from the ureters and bladder, both methods optimised by Cystoinflation. Further investigation is warranted to validate this observation.

We observed a high incidence of urinary tract infections (UTI) in bladder injury cases, which may be due to infection at the injury site and prolonged catheterization (14 to 21 days), a known risk factor for urinary tract infections39. The incidence of other postoperative urological complications, including micturition problems, and fistula formation, was also more frequent in the bladder injury cases. These findings align with studies involving full bladder techniques, suggesting that Cystoinflation contributes to better long-term urinary outcomes by preventing bladder injuries3,23.

While the theoretical risk of urinary retention with Cystoinflation exists40,41, it was not observed in this study. This may be due to the appropriate filling volume of saline (300 cc; the normal bladder capacity reported in literature is 300-400 cc)40,41,42, the short duration of bladder filling only during bladder mobilization, and immediate bladder emptying after completing dissection.

Previous literature indicates that neonatal admissions and adverse neonatal outcomes are independently associated with PAS disorders43. In contrast, the Cystoinflation group did not show a significant increase in neonatal respiratory distress or admissions, indicating a positive impact on foetal well-being. The absence of compromised foetal well-being despite the ligation of surface vessels can be attributed to the lack of direct connection between these vessels and the foetal blood supply, as well as minimized blood loss before delivery44.

In this study, both groups showed improved outcomes compared to international studies that did not use the Triple-P procedure, with the Cystoinflation group achieving better results, thus aligning with our hypothesis13,15. Longitudinal follow-up addressed concerns about residual ovarian function post-uterine devascularization, indicating that arterial ligation procedures do not impact pelvic vascular supply or future fertility due to extensive collateral circulation2,45,46,47,48.The literature suggests minimal long-term effects on fertility, reporting an 88.9% conception rate and a recurrence rate of 22.8–28.6% in future pregnancies, with occasional cases of spontaneous uterine rupture, uterine synechiae, and Asherman’s syndrome41. Our study showed a conception rate of 62.8% in the Cystoinflation group and 28.6% in the control group, both lower than the previously reported 88.9%. This discrepancy may be due to differences in subject characteristics between studies and could also suggest a suppression of reproductive function in the bladder injury cases within our study population, warranting further investigation. Once conception was achieved, there were no significant differences between groups in miscarriage rates, preterm labour, or recurrent PAS. The recurrence rate of PAS was 22.2% in both groups, consistent with literature reports.

Strengths and limitations

This study is among the earliest to observe the effect of Cystoinflation in improving the outcomes of the Triple-P technique for the conservative management of PAS. A key strength is the inclusion of only visually confirmed PAS cases during surgery and an adequate follow-up period, which reduces bias in the results. While ureteric injuries have been documented in the literature, no such incidents were observed in our study. This absence may suggest the possibility of an insufficient sample size. Inability to blind surgeons is another limitation, as inflated bladder can be taken as distended lower uterine segment if bladder is not carefully identified by surgeon during filling. The optimal bladder filling volume remains undetermined; insufficient filling may prove ineffective, while overfilling risks overstretch and urinary retention. Further, the ideal substance for Cystoinflation (methylene blue, CO2, or saline) is still unknown. Additionally, Cystoinflation requires detailed counselling and trained staff. Despite these limitations, Cystoinflation significantly reduced bladder injuries and hysterectomy rates in PAS cases. This single-centre study’s findings need validation through larger, multicentre trials.

A key point of ongoing debate within the PAS community concerns the differentiation of true PAS from cases of low-lying placenta or previa associated with scar dehiscence which effects validity of the conservative management. Placenta accreta spectrum (PAS) represents an invasive pathological condition of placenta characterized by active neovascularization and direct trophoblastic invasion of the myometrium, with potential extension beyond the uterine serosa into adjacent pelvic structures2. In contrast, scar dehiscence is a non-invasive mechanical defect in the uterine wall. In cases of scar dehiscence, the placental tissue, remains confined within the anatomical boundaries of the uterus and does not exhibit invasive behaviour49.

In this study, we used prenatal imaging including color Doppler ultrasound to identify hypervascular, aggressive nature of PAS. Key features that support diagnosis of PAS include obliteration of the uteroplacental interface, the presence of bridging vessels traversing different layers of the uterine wall, large turbulent placental lacunae, and, in severe cases, hypervascularity at the serosal surface50. Cases with significant neighboring organ invasion are rare and therefore excluded.

Conclusion

The incorporation of Cystoinflation into the modified Triple-P procedure represents a beneficial addition in the management of placenta accreta spectrum disorders. By effectively reducing bladder injuries, operative time, blood loss, and postoperative complications, this technique enhances surgical outcomes and safety. Future research and collaborative efforts are warranted to validate the efficacy and long-term benefits of these modified techniques in managing placenta accreta spectrum disorders.

Materials and methods

This single-centre, randomized controlled trial, part of the Cystoinflation project, was conducted at Lady Willingdon Hospital after approval from the Ethical Committee of King Edward Medical University (KEMU), Pakistan (IRB# 216/RC/KEMU, March 27, 2017). The trial was retrospectively registered on Clinical Trials.gov (NCT04302545) on March 10, 2020. This study adhered to the Declaration of Helsinki, CONSORT 2010 guidelines, and Good Clinical Practice guidelines.

Recruitment of participants took place over a two-year period, from May 1, 2017, to April 30, 2019. Each participant underwent short-term follow-up for three months to monitor urological complications and neonatal outcomes, with the final subject’s follow-up completed by July 31, 2019. Additionally, long-term follow-up was conducted for another five years to evaluate future fertility in couples seeking conception, with the final assessment concluding on April 30, 2024.The modified Triple-P procedure served as the control condition in both groups, with Cystoinflation implemented exclusively in the intervention group.

Sample size

We performed a sample size calculation using G*Power for a Fisher’s Exact Test to compare bladder injury rates between two independent groups. Based on prior hospital data for women with Placenta Accreta Spectrum (PAS), we estimated expected bladder injury rates of 2% in the Cystoinflation group and 13% in the control group. Assuming equal group sizes, the analysis indicated that 68 participants per group (total: 136) were required to achieve 95% power at a 0.05 significance level. For a 10% drop out rate, the total sample size was calculated to be 150 (75 participants per group).

In this study, the observed bladder injury rate was 2.67% in the intervention group and 18.7% in the control group, yielding an actual statistical power of 0.89.

Subjects and selection criteria

Subjects were selected in two phases. The initial phase involved a thorough clinical evaluation of pregnant women with placenta previa and a history of previous operative delivery admitted in antenatal ward. This included a detailed medical history, physical examination, and preoperative testing, which encompassed blood grouping, a complete blood count, urinalysis, fasting blood sugar levels, hepatitis screening, ultrasound imaging, and Doppler flow studies of the placenta. All the subjects had to meet the following inclusion criteria.

  • Healthy pregnant women aged 18 to 45 years, with a gestational age of 36 weeks or more, verified by a dating scan.

  • A haemoglobin level ≥ 11.5 g/dL and six cross-matched blood donors arranged before surgery.

  • History of placenta previa in present pregnancy and prior C-section.

  • Prenatal diagnosis of placenta accreta spectrum (PAS) based on ultrasound and Doppler flow studies using PAS imaging criteria (loss of retroplacental hypoechoic space, reduced myometrial thickness, and abnormal vascularity of the lower uterine segment with subserosal bulging) holding sensitivity and specificity around 0.83 for both ultrasound and MRI12.

Exclusion criteria included.

  • A history of urinary tract problems,

  • Intrauterine fatal death,

  • Anaemia (Hb < 11.5 g/dL),

  • Bleeding disorders, or other medical issues that could elevate surgical risks.

Those who met the criteria were provided with detailed counselling regarding Cystoinflation. The second phase of selection occurred during the caesarean section, where adherent placenta (Accreta, Increta or percreta) was confirmed by direct observation. Written informed consent was obtained from all participants before surgery, should adherent placenta be diagnosed during C-section, for assignment to either study group as subjects.

Perioperative criteria for adherent placenta (PAS)

Perioperatively, adherent placenta was classified by direct observation according to FIGO consensus guidelines15. FIGO diagnostic criteria classify PAS into.

Grade 1 Placental adherence to the myometrium without invasion. Manual detachment at delivery provokes haemorrhage, necessitating intervention. At surgery, there is no serosal bulge or hypervascularity.

Grade 2 Invasion into the myometrium is evident with external signs like bluish discoloration, uterine bulge, and marked vascularity. However, the placenta does not penetrate the serosal surface.

  • Grade 3: This includes:

  • 3a: Penetration up to the uterine serosa.

  • 3b: Invasion of the bladder.

  • 3c: Extension into adjacent pelvic structures.

The subjects were excluded if PAS showed deep invasion into adjacent organs beyond the peritoneal surface (3b and 3c). These cases are rare involving invasion of bladder muscularis and mucosa (usually suspected by haematuria preoperatively or gut) and resection of adherent portion of the organ by concerned surgeon).

Randomization

Participants were randomly assigned to either the control group (Modified Triple-P technique without Cystoinflation) or the intervention group (Modified Triple-P technique with Cystoinflation) by parallel assignment. The Head of Department delegated the surgeries to experienced obstetricians in the department with expertise in the Triple-P technique, blunt bladder dissection, and Cystoinflation. The principal investigator generated random numbers using the STAT TREK random number generator. These numbers were recorded on sealed envelopes by study team members, with each envelope containing the assigned group and placed in sequential order in a file. The file was securely kept by the operating theatre charge nurse in a file storage locker within the operating theatre until the assigned surgeon needed to know the group assignment. The charge nurse then opened the top envelope, informed the surgeon of the group assignment just before surgery, and then placed the envelope at the bottom. The assigned random number for each participant was recorded on all relevant documents by the study team manager in the operating theatre, including the hospital admission file, consent form, trial entry form, and outcome form. In charge nurse then replaced the file in the storage locker again. The team manager was responsible to get the entry form signed by the surgeon at the end of surgery and then placing all study documents into the study file within the lead investigator’s office.

Blinding

Surgeon could not be blinded in this study. However, Patients, as well as all medical and paramedical staff outside the operating room, were kept unaware of group assignments. Operating room attendants were trained to simulate Cystinflation in both groups using phrases such as, “I will fill the bladder, you will remove the clip, and we will ensure adequate bladder filling for this patient” while the patient was conscious before anaesthesia. The staff conversations were intended for perception of the procedure in both groups without introducing any extraneous sounds that could potentially distract the surgical team.

Post-operative follow-up was conducted by study team medical officers blinded to the patient’s group. To ensure objective statistical analysis, the lead investigator provided the statistician with groups replaced by alphabets in data file, effectively blinding them to the intervention details.

Preoperative preparation

Given the risk of continued blood loss, six units of blood were arranged for each subject before surgery to ensure surgical safely. Prophylactic antibiotics were administered 30 min before the incision.

Anaesthesia and catheterization

Participants underwent catheterization under general anaesthesia. One of the fully trained theatre staff nurses inserted the catheter into the urethra using aseptic measures, while the surgeon held the distal end of the catheter high in his hand during the catheterization procedure. The surgeon then retained the catheter by inflating the bulb with 10 cc of distilled water. After draping the subject, the surgeon fixed the urinary port end to the sheet in front of the thigh and attached it to the urine bag for drainage. The abdomen was opened in all cases by phanensteil incision and abdominal wall was opened layer by layer to reach the peritoneum. Figure 2 illustrates steps of modified Triple P technique with cystinflation.

Intervention group–cystoinflation

In the intervention group, the bladder was retro-filled either after opening the peritoneal cavity or before this step if the uterus was adherent to the rectus muscles and bladder could not be delineated. The house officer in the operating theatre filled the bladder over approximately two minutes with 300 cc of normal saline through the bladder catheter using a bladder wash syringe. Meanwhile, the principal surgeon identified the lateral margins of the gradually distending bladder among vessel bundles and adhesions during the filling process. The urine port was then clamped with an artery clip. After the surgeon completed bladder mobilisation, The house officer attached the urinary port of the catheter with the tubing of the urine bag and removed the artery clip, allowing the bladder to drain freely into a urine bag23.

Control group-no cystoinflation

The subjects assigned to the control group underwent bladder separation with the urinary catheter placed on free drainage.

Modified Triple-P procedure

The modified Triple-P procedure was applied in both groups as a controlled condition. The steps of the procedure included:

  1. 1.

    Blunt dissection of the bladder: Bladder was separated using the lateral window technique19. A lateral window was created in the anterior flap of the broad ligament next to the bladder margin, avoiding nearby vessels. This allowed the surgeon to insert the index finger between the bladder and uterus below the adhesions for blunt dissection. The portion of the bladder over the surgeon’s finger was separated by lifting the finger upwards toward the vesicouterine peritoneum while applying pressure on the lower uterine segment. The surgeon then diathermized the superficial vessels on the peritoneum and peritoneum was incised with a fine dissecting scissor keeping away from the bladder margins. The adhesiolysis was performed until the bladder surface could be easily be pushed down. The surgeon then pushed the bladder downward with Doyn’s retractor to improve visualization in the surgical field. The house officer then removed the artery clip in Cystoinflation group, allowing the bladder to drain freely into a urine bag.

  2. 2.

    Delivery of the Baby: The triple- P technique recommends high incision above the upper border of placenta14. In the modified technique, incision was given according to location of placenta, a high up lower segment incision in lower uterine segment in central placenta previa and a routine lower segment incision in other types. A 10 cm transverse uterine incision was made. The placenta was separated from the uterine wall to reach the membranes, amniotomy was performed, and the baby was delivered.

  3. 3.

    Uterine elevation: the uterus was exteriorized along with the placenta and lifted up by assistant surgeon. Uterine exteriorization, also known as uterine elevation, temporarily kinks the uterine arteries, thereby reducing haemorrhage51,52. Bilateral uterine artery ligation was done at least 3 cm below the incision to ligate the main ascending uterine vessel. The non-adherent portion of the placenta typically begins to separate once the uterus contracts post-delivery, leaving only the adherent part attached. Placenta was kept in situ till uterine vessels were ligated.

  4. 4.

    Adherent Placenta Separation and Haemostasis: A crucial step in reducing blood loss is to avoid attempting placental removal before ligating the uterine blood vessels. In this study, after bilateral uterine artery ligation was performed, gentle cord traction was employed to help identify the exact site of placental invasion and confirm abnormal adherence. Non-adherent placentas were separated by following the cleavage plane between the uterine wall and the placenta. The adherent placenta, along with the myometrium, was excised. The uterine cavity was then carefully examined, remaining pieces of placenta were removed manually and any torn areas where the placenta had invaded were approximated.

    A crucial step to reduce blood loss is to avoid attempting placental removal before ligating the uterine blood vessels17. In this study, after bilateral uterine artery ligation was performed, Placenta was gently separated by creating cleavage plane between uterine wall and placenta. The adherent placenta along with the myometrium was excised or separated by gentle traction. The uterine cavity was then carefully examined, and torn areas where the placenta had invaded were approximated.

  5. 5.

    Hemostasis of lower uterine segment: A cross-stitch knot was applied within a 3 cm square area, positioned 3 cm from the corner of the incision. Continuous squeeze sutures were then placed vertically along the entire lower margin of the incision, extending up to the isthmus and the depth of placental attachment. The stiches were continued until the other corner of the incision was reached, where it concluded with another cross-stitch knot at the opposite corner leaving 1 cm at the incision margins for subsequent closure of the uterine wall in two layers using Vicryl #117. Haemostasis of the upper margin of the incision was similarly achieved using continuous squeeze sutures to the involved depth. Thorough haemostasis of the posterior uterine wall was ensured by overrunning any bleeding surfaces and repairing torn areas resulting from the adherent placenta, prior to a two-layered closure of the uterine incision.

Figure 3 depicts the crucial steps in the Triple P technique and highlights the role of Cystoinflation in ensuring their successful execution.

Fig. 3
figure 3

Triple p technique and cystoinflation.

Full size image

Bladder peritoneum and posterior surface check

The peritoneum and posterior surface of the bladder were carefully inspected by holding and elevating the bladder with Allis forceps to check for bleeding vessels or injuries, which were then diathermized. An intraperitoneal drain was inserted, followed by layered closure of the abdominal wall.

Outcomes

Primary outcome measures

The primary outcomes included the bladder injury rate, size, site, and depth of the injury, Bladder injury was determined by a retrograde dye test with methylene blue whenever suspicion of bladder injury arose. The size, site, and depth of the injury were carefully measured and recorded using a sterilized scale. The primary outcome is shown in Table 1.

Secondary outcome measures

Secondary outcomes included maternal blood loss, and hysterectomy rate. operative time, foetal Apgar scores, postoperative neonatal outcomes, urological complications, and duration of hospital stay. The estimated blood loss was calculated by the increase in the weight of sponges used during the procedure, with 1gram equating to 1 cc of blood. Hysterectomy was performed if the patient experienced a fall in blood pressure due to uncontrolled bleeding that persisted despite blood transfusion.

The staff nurse documented the duration of the surgery in minutes, measured from the incision to the final closure of the skin. Foetal outcome was assessed by Apgar scores (new-born’s heart rate, respiratory effort, muscle tone, reflex response, and colour immediately after birth, with a maximum score of 2 points for each criterion) at 1 min and 5 min, and the number of admissions to the neonatal unit. The delayed diagnosis of a urological injury was made based on the presence of surgical ascites, followed by biochemical screening which revealed a high level of creatinine in the ascitic fluid, approximating the creatinine level in urine when compared to blood36.Other Urological complications were recorded as frequency (urinating more than 8 times per day), urgency (compelling desire to micturate after passing urine), urethral incontinence (involuntary urine loss through the urethra), and vesicovaginal fistula (loss of urine through the vagina). The catheter was removed after 24 h in the no bladder injury arms of both groups. In the bladder injury arms, the catheter was kept in situ for 7 days if only muscularis disruptions were repaired, for 10 days if perforations smaller than 1 cm were repaired, and for 14 days if perforations greater than 1 cm were repaired. Subjects were discharged after twelve hours of catheter removal if there were no urinary complaints. A mandatory postoperative visit at the end of 3 months was conducted, during which all subjects were evaluated for postoperative micturition problems, and retrograde cystography was performed to rule out fistula. Subjects with fistula were referred to a urologist for further management. Outcomes were recorded on a proforma by assigned medical officers and entered into an SPSS20 datasheet.

Delayed outcome measures

The study follow-up was extended to a period of five years to evaluate reproductive outcomes, including conception rates, incidences of miscarriage, occurrences of preterm delivery, and recurrence of Placenta Accreta Spectrum (PAS). The subjects with uterus in situ, no BTL and a desire to conceive were requested to meet the study team by sending massages through Emails and by cell phones.

Data collection

Data was collected on a study proforma including patient characteristics and primary, secondary, and delayed outcomes measures.

Adverse events reporting

Adverse events were systematically recorded throughout the study. All complications, including urinary tract infections, urinary retention, postoperative bleeding, and other urological issues, were documented and categorized by severity. Any unexpected complications related to Cystoinflation (e.g., bladder overdistension or retention) were also to be monitored and reported.

Statistics

Statistical analysis was performed using IBM SPSS Statistics 20 (SPSS Inc, Chicago, IL, USA). The normality of continuous data was analysed by the Shapiro-Wilk test and normal data were compared by the T-test, non-normal data with the Mann-Whitney U test, and categorical data with the chi-square test. The bladder injuries in both groups were recorded as frequencies and percentages, size of bladder injury was analysed within bladder injury to evaluate the difference more clearly. Other parameters were evaluated across both groups as a whole. The correlation of Cystoinflation with outcome data was analysed using regression analysis.