This is a retrospective case-control study. A total of 360 cases of CSP were identified in the Department of Gynecology, First Affiliated Hospital of Chongqing Medical University, between January 2016 and December 2021. All patients underwent hysteroscopic curettage as treatment.

The inclusion criteria were: complete clinical and follow-up data; elevated human chorionic gonadotropin (hCG) level; history of cesarean section; ultrasound demonstrating an empty uterine cavity and cervical canal; gestational sac implanted in the lower segment of the anterior wall of the uterus; stable vital signs; patients were fully informed of the various treatment methods and had voluntarily chosen hysteroscopic surgery. There were no contraindications for the operation based on preoperative physical examination and ancillary examinations (complete blood count, coagulation tests, liver and kidney function, electrolytes, blood glucose, electrocardiogram). The exclusion criteria were: evidence of a systemic infection or severe local infection; heart, lung, liver, or kidney complications; coagulation dysfunction; presence of trophoblastic disease; presence of malignant tumor; the patient chose a treatment method other than hysteroscopy.

Two cases with hydatidiform mole were excluded from the final analysis. Fourteen cases were treated with medications with expected outcomes, while 6 cases were excluded from uterine curettage, vaginal surgery, and laparoscopic surgery. The final cohort for analysis therefore consisted of 338 cases.

All patients provided informed consent prior to the operation. The information given to patients explained the risks associated with massive bleeding, including the possibility of requiring emergency surgery or potentially a hysterectomy. The cervix was dilated using oral or vaginal misoprostol 1 hour prior to surgery. Preparations were made for blood transfusion, fluid replacement, and emergency surgery. Antibiotics were given perioperatively to prevent infections.

The surgical instrument used was an intraluminal hysteroscope (StrykerHD1288, Kalamazoo, MI, USA), and the perfusion fluid consisted of 5% glucose. Local anesthesia with lidocaine block was used. All procedures were performed by the same team of doctors. After routine disinfection, hysteroscopy was performed to determine the position, shape, and size of the pregnancy, followed by curettage. A follow-up hysteroscopy was performed to determine if there was residual bleeding. If an active hemorrhage was detected during the operation, a water bag cervical tube was used to compress the bleeding site.

According to the literature, the standard criterion for postpartum hemorrhage is when the amount of bleeding during and within 24 h after the operation reached 500 mL. Patients with $\ge$500 mL of hemorrhage were allocated to Group A, while patients with 500 mL of hemorrhage were allocated to Group B (see Fig. 1).

Fig. 1.

Flow chart used for selection of the study population. CSP, cesarean scar pregnancy; D&C, dilatation and curettage.

The following data was collected for each patient: age, days of amenorrhea, frequency and interval time of cesarean section, number of induced abortions, hCG level before treatment, gestational sac length, CSP classification, blood flow signal around the gestational sac, and amount of hemorrhage.

The Expert Consensus on the Diagnosis and Treatment of Scar Pregnancy after Cesarean Section issued by the Chinese Medical Association in 2016 defines several types of CSP. Type I refers to the part of the gestational sac located at the scar site, with a muscular layer thickness $>$3 mm. Type II refers to the part of the gestational sac located at the scar, with a muscular layer thickness $\le$3 mm. In Type III, the gestational sac is located completely at the scar, protrudes outward toward the bladder, and the thickness of the muscular layer at the scar is $\le$3 mm. Type III also includes a unique form of lumpy CSP, characterized by cystic or solid-like echoes. These cases typically involve a combination of pregnancy remnants and blood clots following a CSP abortion. No blood flow signal and a resistive index (RI) $\ge$0.4 were classified as group L, while those with a blood flow signal and RI 0.4 were classified as group H.

Univariate and multiple logistic regression analysis were used to examine the relationship between variables and the risk of postoperative hemorrhage during hysteroscopy for CSP. SPSS 25.0 statistical software (IBM Corp., Chicago, IL, USA) was used for analysis. Measurement data were expressed as $\overline{x}$ $\pm$ s, and p 0.05 indicating statistical significance. The 95% confidence interval (CI) was calculated using the total score function in GraphPad Prism 9.0 software (GraphPad Software, Inc., San Diego, CA, USA), where the standard deviation is calculated as the square root of [(incidence of hemorrhage within 1 week of amenorrhea days) $\times$ (incidence of hemorrhage within 1 week of amenorrhea days)/total number of cases within 1 week of amenorrhea days]. Measurement data that conformed to a normal distribution was expressed as the mean $\pm$ standard deviation, with groups compared using two independent sample t-tests. Measurement data that did not conform to a normal distribution was expressed as the median (Q25, Q75), and compared using the Mann Whitney U-rank sum test. Count data was analyzed using the Chi-square test.

The factors of age, number and years of previous cesarean sections, number of induced abortions, hCG level prior to treatment, and CSP classification were not significantly associated with the risk of hemorrhage during hysteroscopic treatment of CSP (Table 1). However, the number of amenorrhea days, gestational sac length, myometrial thickness of the uterine scar, and blood flow signal around the gestational sac were all significantly associated with the risk of hemorrhage (p 0.05).

As shown in Table 1, the amenorrhea period with different bleeding volumes was 49.00 (43.00, 59.00) vs. 57.00 (50.00, 74.25), with statistical differences (p = 0.001). The diameter of gestational sac (mm) was 30.50 (21.25, 42.00) and 45.00 (35.00, 52.75), respectively, with a p value of 0.001, indicating statistical differences. The thickness of uterine scar (mm) was 2.40 (1.30, 4.00) and 0.95 (0.00, 1.93), respectively, with a p value of 0.001, indicating statistical differences (see Table 2).

The amenorrhea days were odds ratio (OR) 1.036, with a 95% CI (1.015–1.058), and a p-value of 0.001. The gestational sac length was OR 1.027, with a 95% CI (1.005–1.049), and a p-value of 0.016. The myometrial thickness at the uterine scar was OR 0.392, with a 95% CI (0.268–0.573), and a p-value of 0.001. The blood flow signal around the gestational sac was OR 4.103, with a 95% CI (1.737–9.691), and a p-value of 0.001. The significant risk factors for greater hemorrhage after hysteroscopic treatment were fewer amenorrhea days, higher Type L blood flow signal around the gestational sac, and shorter gestational sac length. On the other hand, a thinner myometrial layer was a protective factor.

Risk maps for hemorrhage during hysteroscopic treatment for CSP were constructed based on the period of amenorrhea and length of the gestational sac. The risk of hemorrhage from CSP increased after 49 days of amenorrhea, with the incidence of hemorrhage $\ge$500 mL being 17.46% after 70 days of amenorrhea. The incidence of hemorrhage $\ge$500 mL was 28.57%. Hysteroscopic treatment of CSP with a gestational sac length $>$30 mm increased the risk of hemorrhage $\ge$500 mL, with an incidence of 11.53% (Figs. 2,3).

Fig. 2.

Bleeding after hysteroscopic treatment of CSP according to days of amenorrhea.

Fig. 3.

Bleeding after hysteroscopic treatment of CSP according to the length of the gestational sac.

The error line in Fig. 2 represents the 95% CI. The risk of bleeding was found to increase when the duration of amenorrhea was $>$49 days, and especially when it was $>$70 days.

Bleeding after hysteroscopic treatment of CSP increased when the gestational sac length was $>$30 mm (Fig. 3), the blood flow signal RI around the gestational sac was 0.4, and the uterine scar showed a thicker myometrium.