Cesarean delivery is one of the most frequently performed surgical procedures worldwide. In Latin America, the median rate of cesarean deliveries is 33%, and it is higher in private hospitals.1 According to data derived from DATASUS 2006,2 the cesarean delivery rate in Brazil was 42.7% in 2004, while in the United States of America the practice of cesarean deliveries reached a record high of 29% of all deliveries.3 Given the ease of surgical performance, some countries show an elevated cesarean delivery rate, accounting for 70% of all deliveries.1
Many variations in surgical techniques for cesarean delivery have been proposed, aimed at reducing surgical time, making the surgery easier and more efficient, lowering costs, decreasing the risk of adverse effects and postoperative morbidity, as well as length of hospital stay.4–8 The details of the surgical technique and its variation are important and were evaluated in randomized controlled trials.4–8
Temporary removal of the uterus from the abdominal cavity (exteriorization) has been postulated as a valuable technique for repair of the uterine incision (hysterorrhaphy) after delivery of the newborn and placental removal, either spontaneous or manual.4 However, the lack of consensus about the optimal site for performing hysterorrhaphy is reflected by variations in surgical practice. There are few randomized controlled trials and results are discordant. Some studies suggest that uterine exteriorization is associated with a reduction in rates of infection and postoperative morbidity,9 in addition to decreasing the occurrence of perioperative bleeding and reduction in postoperative hematocrit.10–12 In contrast, several studies have found no association between surgical technique and intra- or postoperative maternal morbidity. One randomized controlled trial involving 194 patients even showed a slight increase in postoperative pain in women undergoing uterine repair with the exteriorization technique (mean pain score of 4.4 compared with 3.7, P=.046).13 A systematic review of the Cochrane Library, updated in 2004, included 1,221 women in six randomized controlled trials comparing the effects of extra-abdominal with in situ repair of the uterine incision.4 There was no statistically significant difference between both groups regarding the majority of outcomes analyzed, except for febrile morbidity and length of hospital stay. With extra-abdominal closure of the uterine incision, febrile morbidity was lower (relative risk 0.41, 95% confidence interval 0.08–0.39). Nevertheless, these results are based on few and small studies. There is still no conclusive evidence about the surgical technique that offers the most advantages. In this context, we performed the current study with the aim of comparing the differences between these two surgical techniques regarding the rate of nausea, vomiting, and blood loss in the intraoperative period; duration of surgery; number of sutures used for hysterorrhaphy; pain assessment using the visual analog scale (VAS); number of daily doses of analgesics required in the postoperative period; rate of nausea and vomiting in the postoperative period; length of hospital stay; rates of surgical site infection and endometritis.
MATERIALS AND METHODS
A randomized single-blind controlled trial was performed in pregnant women with indication for cesarean delivery (elective or emergency) who were seen at the Maternity Hospital of the Maternal-Infant Institute of the Instituto Materno-Infantil Prof. Fernando Figueira (IMIP), from August 2005 to August 2007. Sample size was calculated based on the assumption that the rate of postpartum febrile morbidity in cesarean deliveries is around 11% for uterine closure within the abdominal cavity. Assuming a reduction of about 50% for repair with the uterus exteriorized9 resulted in 612 women. Eventual losses were predicted, increasing this number by about 10%, yielding a total of 670 participants. A power of 80% and an alpha error of 5% were adopted. Included were women with an indication for cesarean delivery (emergency or elective), with gestational age greater than 24 weeks, who agreed to participate in the study. Criteria of exclusion were repeat cesarean deliveries (history of two or more cesarean deliveries), chorioamnionitis, history of previous surgery (gynecologic or abdominal), third-semester hemorrhage or active gastrointestinal bleeding or both, and inability to consent. The study had been previously approved by the Research Ethics Committee of the institution. Only women who agreed to participate in the study and signed a free informed consent form were included.
The study was performed on Tuesday, Thursday, and Friday day shifts. The physicians on call were responsible for identifying potential candidates for study participation. The researcher was contacted to apply the criteria of eligibility and approach the pregnant patients, explaining the aims of the project and allocating each patient to a research group. For allocation of the participants, a computer-generated (Epi-Table, Epi-Info 6.04b, U.S. Centers for Disease Control and Prevention [CDC], Atlanta, GA) list of random numbers was used. Sequentially numbered, sealed, opaque envelopes were prepared, containing the research form and assigning each woman to a group. Neither the participants nor the researchers involved had any knowledge of the contents of the envelope, ensuring blinding of allocation. Only after applying the criteria of eligibility and obtaining consent from each participant were the envelopes opened. Confidentiality of the allocation group was partially revealed and should have been known by the assistant physician responsible for performing the cesarean delivery. The patients were blinded to the technique adopted for uterine repair during the entire study (single-blind), although the technique was known to the researcher responsible for data collection who had access to the medical charts.
Cesarean deliveries were performed by residents trained and apt to perform uterine repair both with the uterus exteriorized and within the abdominal cavity, under guidance of staff obstetricians. The cesarean technique chosen was the Misgav-Ladach method,14 usually performed in the service, and the placenta was removed by traction. Uterine incision was closed with a continuous single layer of chromic catgut number 1. The visceral or parietal peritoneum were left unsutured.
Anesthetic technique was spinal anesthesia performed with a 27G needle, using 12 mg of 0.5% hyperbaric bupivacaine associated with morphine (70 micrograms). No vasoconstrictor was used routinely, apart from ephedrine (10 mg) if blood pressure decreased 20% from baseline values. All patients received 1g cefazolin intravenously immediately after cord clamping and 2 g dipyrone intravenously at the end of the cesarean delivery. For postoperative pain relief, dipyrone was used in doses adjusted as necessary, starting at 500 mg/dose every 6 hours and paracetamol (starting at 500 mg/dose every 6 hours) for patients allergic to dipyrone. If pain persisted, even with maximum doses of dipyrone, a combination of codeine (30 mg) and paracetamol was used.
When entering the study, patients were accompanied during surgery to assess surgical time, presence of nausea and vomiting, number of sutures used for hysterorrhaphy, and estimate of blood loss. In the postoperative period, the number of daily doses of analgesics used by the patient during hospital stay, need for supplemental analgesia, rate of nausea and vomiting in the postoperative period, pain scores at 6 hours, pain in the first and second days after cesarean delivery (measured by the visual analog scale), length of hospital stay, surgical site infection, and endometritis were assessed. Surgical time was calculated in minutes from skin incision to the last suture of the skin. To estimate blood loss, the difference between the initial hematocrit and the final hematocrit15 was considered, as well as the difference in postoperative and preoperative weight of surgical drapes and towels. The length of hospital stay began at the time of cesarean delivery and was measured in days. For unidimensional assessment of pain, the Face Pain Scale-Revised16 was used. This is a 10-cm scale anchored by two extremes of pain (none and worst pain imaginable), expressed both in numbers and face correspondents. The patient was asked about her pain perceptions at that moment. She was asked to indicate variation in pain intensity according to the face. This result was expressed as a number corresponding to the perception of pain (ordinal number), ranging from zero to 10, and this was the VAS result. Pain was considered mild/none when scores were between zero and five, moderate/severe when the scores were between six and ten. To define site of surgical infection, we used the CDC criteria (1999)17: incisional (in the surgical wound and either superficial or deep) or in organs and spaces (intracavitary or deep infection, with or without abscesses). The diagnosis of postoperative endometritis was based on the presence of fever (starting at 38°C [100.4°F], in the absence of any other cause, in two or more postpartum days, or temperature of 38.7°C [101.7°F] or higher during the first 24 hours postpartum) (U.S. Joint Commission on Maternal Welfare). Uterine sensitivity, purulent or malodorous lochia, and leukocytosis with a shift to the left represent adjuvant criteria. For diagnosis of surgical site infection or endometritis or both, the patients were reassessed at 7 and 30 days after undergoing cesarean delivery.
Statistical analysis was performed with Epi Info 3.1 (CDC). A univariable analysis was performed to test randomization, using the χ2 test of association for categorical variables and Student t test for numerical variables. An intention-to-treat analysis was used. To test the association between the independent variable (surgical technique) and the dependent variables the χ2 test of association was used, adopting a significance level of 5%. The strength of association was determined using risk ratio (RR) and 95% confidence intervals (CIs). The number needed to treat for benefit or harm was calculated when a significant difference was observed between the groups. The group of uterine exteriorization was considered as “Intervention,” and in situ repair group was considered as “Control.”
During the study period, 1,293 women were invited to participate, and 937 were considered eligible for the study. Of these eligible women, 670 agreed to participate in the study and were randomly assigned to undergo exteriorized (n=335) or in situ uterine repair (n=335). The major cause of refusal to participate was fear of uterine exteriorization (n=267). Among the 670 women included, an analysis of 637 was carried out. Of these women, 325 were from the exteriorized group (group 1), and 312 were from the in situ group (group 2). There was a loss of 33 women, due to evasion from the service or refusal to remain in the study. Exteriorization of the uterus was not possible in six patients randomized for this technique because of difficulties encountered in removing the uterus so it remained in its usual place. In contrast, in 26 patients from the in situ group the uterus had to be exteriorized (19 due to increased bleeding and seven due to technical difficulties), and the surgeon made the decision. However, an intent-to-treat analysis was conducted. For analysis of the variables “endometritis” and “surgical site infection,” information given by 581 women was considered because diagnosis of these morbid conditions could not be made in 56 cases (two deaths and 54 patients who did not return for reassessment at 7 and 30 days) (Fig. 1).
Baseline characteristics of both groups (exteriorized and in situ uterus) were similar regarding age (24.7±6.1 and 25.6±6.3 years, respectively), parity (predominantly primigravida, 60% and 57.7%), gestational age (37.5±3.1 and 37.6±3.0 weeks), and body mass index (BMI) (29.8±5.0 and 29.2±4.9 kg/m2) (Table 1). The most frequent indications for cesarean delivery were dystocia and cephalopelvic disproportion, followed by severe oligohydramnios, severe preeclampsia/eclampsia, abnormal presentation, chronic fetal distress, acute fetal distress, twinning, fetal malformation, and previous cesarean delivery (Table 1). There was no statistically significant difference in the rate of nausea (21.5% compared with 22.8%) and vomiting (7.1% compared with 9.0%) during surgery between the groups. Blood loss estimated by the hematocrit was similar in both groups, and it was less than 800 mL in 61.2% and 57.4% of patients, respectively. When blood loss was estimated by the difference between the bloody and dry weights of surgical drapes and towels, it was 1,000 mL or less in 42.8% of women in group 1 and 42.6% in group 2. There was also no statistically significant difference blood transfusion requirement between the groups (0.3% compared with 0.0%) (Table 2). In contrast, a statistically significant difference was observed between the groups (P=.026) when the number of sutures used for hysterorrhaphy was evaluated. In 18.2% of cases in which the uterus was exteriorized, only one suture was employed, compared with 11.9% of cases with in situ uterine repair (P=.03). The risk ratio for using more than one suture was reduced in group 1 (RR 0.93, 95% CI 0.87–0.99). Regarding length of surgical procedure, a statistically significant difference was observed (P=.02), with surgery lasting less than 45 minutes in 44% of patients in the exteriorized group compared with 35.3% in the in situ uterine group (Table 2). The risk ratio for surgical time greater than 45 minutes was 0.86 (95% CI 0.76–0.98).
An analysis of the number of daily analgesics consumed postoperatively and the need for supplemental analgesia showed no difference between the groups. The majority of patients received up to two doses of analgesics (68.3% compared with 69.6%) and needed no supplemental analgesia (96.9% compared with 98.1%) (Table 3). In the postoperative period, the rate of nausea was 4.3% in the exteriorized uterus group compared with 3.2% in the in situ uterus group, while vomiting occurred in 1.5% of patients in the exteriorized uterus group compared with 1.6% of patients in the in situ uterus group. No statistically significant differences were shown between the groups regarding these two complaints (Table 3). In pain assessed by the VAS, we observed a statistically significant difference (P=.03) at 6 hours postoperatively, where group 1 had a higher rate of moderate-to-severe pain (32.6%) compared with group 2 (23.1%). The risk of moderate-to-severe pain at 6 hours was 41% greater for group 1. There was no difference between the groups regarding pain scores in the first or second day postpartum (Table 3).
There was no statistically significant difference between the groups concerning the rate of surgical site infection (7.1% compared with 8.7%) and endometritis (1.7% compared with 2.1%). These variables were assessed in only 581 patients, who returned for a hospital visit 7 and 30 days after the procedure (Table 3). Two maternal deaths occurred in the exteriorized uterus group, and no significant difference was noted between the groups (Fisher exact test, P=.26). The causes of maternal death were peripartum cardiomyopathy and acute pulmonary edema.
In the current study, we found no significant differences in the two techniques for uterine repair at cesarean delivery, except for duration of the surgical procedure and number of sutures used. Time was shorter and sutures were fewer in number when the uterus was exteriorized. Pain 6 hours postoperatively was less intense in the in situ uterine repair group.
There are relatively few randomized controlled trials that compare surgical techniques for uterine repair during delivery. Reviewing MEDLINE, Lilacs/SciELO, EMBASE, and databases and records of Cochrane clinical trials, we found only 11 randomized controlled trials with this purpose, apart from the systematic review available in the Cochrane Library.4 This systematic review was last updated in 2004 and included six studies covering 1,221 women. In this review, no statistically significant differences were found when comparing the two techniques for uterine repair, except for the lower rate of postpartum fever and shorter hospital stay in the extra-abdominal technique. The reviewers concluded that evidence is not sufficient to definitely support any method as the most advantageous for uterine repair. Further research is required, especially in low-income countries.
In our study, we observed a shorter surgical time, approximately 2.5 minutes, when the uterus was exteriorized for repair. The risk ratio for duration of surgery greater than 45 minutes was 0.86 (95% CI 0.76–0.98). It produced a number needed to treat for benefit (shorter surgical time) of 12 (95% CI 6–100). A systematic Cochrane review evaluated this outcome in six studies and found no significant difference in the duration of surgery, according to type of repair of the uterine incision.4 However, all studies analyzed involved a lower number of participants compared with our study, which may increase the probability of type II statistical error.9–11,13,18
In a controlled trial conducted and published in 2007, a statistically significant decrease was observed in duration of uterine repair in the exteriorized uterus group in comparison with the in situ uterus group.19 Despite this shorter time for uterine repair in the exteriorized group, there was no statistically significant difference when the total surgical time (from skin incision to skin closure) was considered. This result was surprising because there was no difference regarding estimation of blood loss or other reference to events that might have justified this outcome. It may be assumed that there was difficulty in placing the uterus back into the cavity after uterine repair, although the authors do not explain this finding.
In our study, we did not calculate the duration of uterine repair alone, but only total surgical time, observing that 44% of the cesarean deliveries performed with uterine exteriorization lasted up to 45 minutes, whereas this occurred in only 35.3% of cesarean deliveries performed with the uterus in situ (P=.024). The mean surgical time found in our study was longer than that described in the majority of studies, ranging from 29 to 45 minutes,10–12,20 which may be justified because medical residents performed the cesarean deliveries in our service. Although these physicians had been previously trained in both techniques, they still lacked the skill and ability of more experienced surgeons.
Evaluation of the number of sutures used for uterine repair revealed a statistically significant difference. A decreased number of sutures was used in the exteriorized technique, with a risk ratio of 0.93 (95% CI 0.87–0.99) for using more than one suture. It produced a number needed to treat for benefit (decreased number of sutures) of 16 (95% CI 9–∞). This variable was not studied in most randomized controlled trials available,9–13,18–20 and we chose to evaluate this aspect because procedures were performed in a philanthropic maternity hospital, supported by Sistema Único de Saúde (SUS, single health system in Brazil). It may be important to use the minimum number of sutures to reduce hospital costs.
When analyzing the rate of nausea and vomiting in the intraoperative and postoperative periods, there was no statistically significant difference between the groups. Intraoperative analysis is in accordance with the results of the systematic review of the Cochrane Library,4 corroborating our assumption that these symptoms were not directly influenced by the technique used. In contrast, a small randomized controlled trial, recently published and involving 80 patients, found a lower rate of intraoperative nausea and vomiting in the in situ uterus group compared with the exteriorized uterus group (38% compared with 18%; P=.04).19 We would like to highlight that, although the study was well-conducted and attempted to evaluate these effects by using vasopressor and antiemetic drugs, it may be criticized for involving a small number of patients (40 in each group). Worthy of note is the high rate of nausea and vomiting observed by those authors, likely due to differences in the anesthetic technique employed or, perhaps, the greater emphasis that was placed on the investigation of these effects.
Accurate estimate of the amount of blood lost during cesarean delivery is very difficult because there is a high rate of blood splash during surgery and the methods available for determining blood loss are not very precise. In our study, we evaluated blood loss using gravimetric analysis, weighing the surgical drapes and towels used during surgery. In addition, blood loss was estimated by the difference between preoperative and postoperative hematocrit (immediately before cesarean delivery and in the first 48 hours postoperatively), attempting to associate the volume obtained in each assessment method with the surgical technique used. Using both methods to estimate blood loss, we found no statistically significant difference between the groups, which is in agreement with other studies.4,19
Other studies were conducted with the purpose of evaluating whether there was a difference in blood loss associated with cesarean delivery and whether this difference was related to placental removal (spontaneous or manual) or site of uterine repair. A statistically significant difference was observed, with less blood loss in spontaneous compared with manual placental removal, and there was no relation to site of uterine repair.10,20,21 Likewise, we found no statistically significant difference in blood transfusion requirement between the techniques used, in agreement with the results of two randomized controlled trials in which this effect was evaluated.11,13
For a more direct and accurate pain assessment, we used a VAS to measure pain reported by the patient at 6 hours and on the first and second days postpartum. In our study, we found higher pain scores in the group with extra-abdominal uterine repair at 6 hours after surgery (P=.026), whereas pain scores were similar between groups in the first and second days postoperatively. Uterine exteriorization was associated with 41% greater risk of moderate-to-severe pain (VAS greater than 6). Analyzing the number needed to treat for harm, we found that approximately 11 (95% CI 50–6) cesarean deliveries have to be performed with uterine exteriorization to produce one case of moderate-to-severe pain.
In a systematic review of the Cochrane Library, a meta-analysis was performed to assess intraoperative pain, and no statistically significant difference was observed between the groups.4 However, a randomized controlled trial included in this review assesses not only intraoperative but also postoperative pain up to 6 weeks postpartum.13 In that study, there was a statistically significant difference in intraoperative pain. The authors attributed the rate of pain to adequacy of the anesthesia used rather than the surgical technique. The pattern of pain continued into the late postpartum period (6 weeks), with similar percentages of residual pain in both groups.
In contrast, a randomized controlled study conducted by Nafisi in 200618 to assess postoperative pain in patients, demonstrated lower pain scores in the group with in situ uterine repair on both the first and second postoperative nights. We believe that our results are more consistent because pain due to traction associated with uterine exteriorization would be more intense in the intraoperative period, persisting for the first hours postoperatively, its effect waning gradually.
On the other hand, we found no difference in need for analgesia between the groups. We emphasize that analgesics were given in the postoperative period upon patient request (prescribed if necessary), reflecting the actual need for pain relief. The number of days of hospitalization did not differ between the groups in our study. This was in disagreement with the systematic review, in which a longer hospital stay was observed in the exteriorized uterus group.4 Similar pain between the groups in the first and second postoperative days could explain the similar length of hospital stay, as well as blood loss. These factors could delay hospital discharge of these patients.
Regarding the rate of surgical site infection, as well as the presence of endometritis, there was no statistically significant difference when the patient was observed 7 and 30 days after cesarean delivery, using the criteria recommended by the CDC.12 This result may seem surprising because, in the Cochrane systematic review, the risk of postpartum febrile morbidity was lower in the group with uterine exteriorization.4 Nevertheless, for this outcome, only the study results of Hershey and Quilligan,9 published in 1978 and including 386 women, were regarded. In our study, covering a much larger number of patients, greater than all the randomized controlled trials included in the Cochrane systematic review, sample size was calculated based precisely on the incidence of postpartum febrile morbidity described, with sufficient power to demonstrate significant differences between both techniques.
In conclusion, there is no striking difference between the two techniques used for repair of the uterine incision in cesarean deliveries. The technique should be chosen at the convenience of the surgeon. The shorter surgical time in uterine exteriorization does not seem to provide additional benefits that justify the choice of this technique. We believe that pain reduction at 6 hours postoperatively does not justify choosing the in situ technique because there was no need to increase the number of analgesic doses or use supplemental analgesia.
In an article published in 2005 on evidence-based surgery for cesarean delivery, those authors concluded that there is no evidence to justify a universal recommendation.22 It is worthy of note that further studies should be conducted to evaluate diverse outcomes that were analyzed in this study and in other randomized controlled studies to elucidate whether there are advantages in using either technique.
Personal preference for performing a determined technique is the main guide for choice because in either technique some data justify a choice. It is important to train residents in both techniques because some situations obligate the surgeon to use one or the other. Perhaps, in the population studied (a public hospital in the northeast of Brazil), this answer, although irrelevant in certain populations, may have an important impact on public health: potential decrease in maternal morbidity and lower hospital costs (decreased number of sutures, shorter hospital stay, decreased need for analgesics and hospital readmissions in case of endometritis and surgical site infection). However, it is necessary to definitely establish whether there is any difference between the techniques that may be beneficial to patients or whether the choice of technique should be made by each surgeon.
1. Villar J, Valladares E, Wojdyla D, Zavaleta N, Carroli G, Velazco A, et al; WHO 2005 global survey on maternal and perinatal health research group. Cesarean delivery rates and pregnancy outcomes: the 2005 WHO global survey on maternal and perinatal health in Latin America. Lancet 2006;367:1819–29.
3. Morales KJ, Gordon MC, Bates GW Jr. Postcesarean delivery adhesions associated with delayed delivery of infant. Am J Obstet Gynecol 2007;196:461.e1–6.
4. Jacobs-Jokhan D, Hofmeyr G. Extra-abdominal versus intra-abdominal repair of the uterine incision at cesarean section. The Cochrane Database of Systematic Reviews 2004, Issue 4. Art. No.: CD000085.
5. Bamigboye AA, Hofmeyr G. Closure versus non-closure of the peritoneum at caesarean section. The Cochrane Database of Systematic Reviews 2003, Issue 4. Art. No.: CD000163.
6. Anderson ER, Gates S. Techniques and materials for closure of the abdominal wall in caesarean section. The Cochrane Database of Systematic Reviews 2004, Issue 4. Art. No.: CD004663.
7. Alderdice F, McKenna D, Dornan J. Techniques and materials for skin closure in caesarean section. The Cochrane Database of Systematic Reviews 2004, Issue 2. Art. No.: CD003577.
8. Mathai M, Hofmeyr GJ. Abdominal surgical incisions for caesarean section. The Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No.: CD004453.
9. Hershey DW, Quilligan EJ. Extraabdominal uterine exteriorization at cesarean section. Obstet Gynecol 1978:52:189–92.
10. Magann EF, Dodson MK, Allbert JR, McCurdy CM, Martin RW, Morrison JC. Blood loss at time of caesarean section by method of placental removal and exteriorization versus in situ repair of the uterine incision. Surg Gynecol Obstet 1993;177:389–92.
11. Wahab MA, Karantzis P, Eccersley PS, Russell IF, Thompson JW, Lindow SW. A randomised, controlled study of uterine exteriorization and repair at caesarean section. Br J Obstet Gynaecol 1999;106:913–6.
12. Ezechi OC, Kalu BK, Njokanma FO, Nwokoro CA, Okeke GC. Uterine incision closure at caesarean section: a randomised comparative study of intraperitoneal closure and closure after temporary exteriorization. West Afr J Med 2005;24:41–3.
13. Edi-Osagie EC, Hopkins RE, Ogbo V, Lockhat-Clegg F, Ayeko M, Akpala WO, et al. Uterine exteriorization at caesarean section: influence on maternal morbidity. Br J Obstet Gynaecol 1998;105:1070–8.
14. Holmgren G, Sjöholm L, Stark M. The Misgav Ladach method for cesarean section: method description. Acta Obstet Gynecol Scand 1999;78:615–21.
15. Gross JB. Estimating allowable blood loss: corrected for dilution. Anesthesiology 1983;58:277–80.
16. Hicks CL, von Baeyer CL, Spafford PA, van Korlaar I, Goodenough B. The Faces Pain Scale Revised: toward a common metric in pediatric pain measurement. Pain 2001;93:173–83.
17. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Centers for Disease Control and Prevention (CDC). Hospital Infection Control Practices Advisory Committee. Guideline for prevention of surgical site infection, 1999. Am J Infect Control 1999;27:97–132.
18. Nafisi S. Influence of uterine exteriorization versus in situ repair on post-cesarean maternal pain: a randomized trial. Int J Obstet Anesth 2007;16:135–8.
19. Siddiqui M, Goldszmidt E, Fallah S, Kingdom J, Windrim R, Carvalho JC. Complication of exteriorized compared with in situ uterine repair at cesarean delivery under spinal anesthesia: a randomized controlled trial. Obstet Gynecol 2007;110:570–5.
20. Baksu A, Kalan A, Ozkan A, Baksu B, Tekelioğlu M, Goker N. The effect of placental removal method and site of uterine repair on postcesarean endometritis and operative blood loss. Acta Obstet Gynecol Scand 2005;84:266–9.
21. Magann EF, Washburne JF, Harris RL, Bass JD, Duff WP, Morrison JC. Infectious morbidity, operative blood loss, and length of the operative procedure after cesarean delivery by method of placental removal and site of uterine repair. J Am Coll Surg 1995;181:517–20.
22. Berghella V, Baxter JK, Chauhan SP. Evidence-based surgery for cesarean delivery. Am J Obstet Gynecol 2005;193:1607–17.