Baseline characteristics of the patients and perioperative findings
During the five years of the study period, a total of 953 cases of twin cesarean deliveries were performed. Of these, either modified B-Lynch or Cho’s suture technique was applied to 147 patients because their uteri did not respond to massage and uterotonics. Modified B-Lynch suture was usually performed in case of poor uterine contraction, and Cho’s suture technique was applied to uterine atony along with placenta previa or accreta, and in some cases, in addition to modified B-Lynch suture. Table 1 shows the baseline characteristics of the patients by application status of the compression suture (patients without compression sutures versus patients with compression sutures). No statistically significant difference was found between the two groups in terms of patients’ age at delivery, parity, mode of conception, previous history of abdomino-pelvic surgery, or chorionicity. However, maternal body mass index (BMI) at delivery was significantly higher in patients with compression sutures as compared to those without compression sutures (27.5 vs. 26.9; p = 0.039), which was consistent with prior studies showing that obesity is a risk factor for developing UA [1, 16].
Table 2 shows the perioperative findings during CS of twins. Sum of birth weight was significantly larger in patients with compression sutures as compared to those without compression sutures (p = 0.018). Not unexpectedly, as a result of poor uterine contraction, the group with sutures was significantly higher than the group without sutures in estimated blood loss, hemoglobin change, and the rates of postpartum hemorrhage and transfusion within the first 24 hours of delivery. However, the rate of transfusion after the first 24 hours of delivery in the suture group was not significantly different from that in the non-suture group; both groups achieved hemostasis at an equal rate after the first 24 hours of delivery. Of 147 who were treated with compression sutures, only two patients (1.4%) proceeded to the uterine artery ligation procedure additionally to achieve hemostasis, which resulted in a success rate of 98.6% of the uterine compression suture. As expected, operation time was significantly longer in the suture than in the non-suture group due to the application of sutures; however, the mean time difference was only 8.5 minutes.
In the non-suture group, five patients (0.6%) were treated with other procedures. One patient was treated by uterine artery ligations during CS because the uterine incision was extended to the left uterine artery. To control the bleeding, multiple ligations of left uterine artery were performed. The remaining four patients had late-onset UA, and therefore, were transferred for uterine artery embolization after 24 hours of delivery.
Pregnancy outcomes following sutures
None of the patients with the compression suture developed complications related to the procedure during the first 2 months postpartum period. Among 953 patients, 371 (39%; n = 292 for the patients without sutures; n=79 for the patients with sutures) were available for the follow-up investigation after the first two months postpartum period. Of these, 308 (n = 238 for the patients without sutures; n=70 for the patients with sutures) expressed no desire to become pregnant, 22 (n = 17 for the patients without sutures; n = 5 for the patients with sutures) have tried to conceive, and 41 (n = 37 for the patients without sutures; n = 4 for the patients with sutures) became pregnant. The high rate (83% = 308/371) of giving up future pregnancies in our follow-up sample is probably a reflection of the very low birth rate in current Korean society . In addition, since our patients had twins, and most of our patients were somewhat old aged for another childbirth (see Table 1), these factors possibly contributed to low interests in subsequent pregnancies. Excluding patients who did not desire future pregnancies, the rate of subsequent pregnancy among those with sutures in our follow-up sample was 44.4% (= 4/9), which was within the range reported in the literature . However, given that our follow-up sample only included 39% of the original sample, as well as the fact that the vast majority of the follow-up sample had no interests in future pregnancy, and that the patients in our study were on the basis of the most recent 5 years’ records in our institution, our rate of future fertility should be considered only as an approximate estimate.
Table 3 shows a summary of pregnancy outcome data for the 41 patients sub-grouped according to the application status of the compression suture. As only four of the 41 patients received compression sutures, we did not carry out statistical analysis to compare patients with and without compression sutures in pregnancy outcomes. However, the mean intervals between CS and subsequent pregnancy were very similar in between the two groups (21.2 months for the patients without sutures and 22.5 months for the patients with sutures), suggesting that the adverse effect of sutures on future pregnancy may be minimal. Among the four patients with sutures, one received ART and had ongoing pregnancy, and the remaining three conceived spontaneously, of which one had a full-term delivery, and two had been in ongoing pregnancy.