Objective
We sought to test the hypothesis that an extraperitoneal cesarean section (ECS) technique reduces postoperative pain without increasing intraoperative and postoperative complications.
Study Design
In a single-center, single-blinded prospective trial we randomized 54 patients with an indication for primary or first repeat cesarean section at term pregnancy to an ECS (n = 27) or transperitoneal cesarean section (TCS) (n = 27) procedure. Patients with suspected abnormal placentation, a history of >1 cesarean section, or major abdominal surgery were excluded. The primary endpoint of the study was maximum abdominal pain measured by numeric rating scale ranging from 0–10.
Results
Patients after ECS had significantly less maximum surgical site pain than patients after TCS. Median peak pain scores on postoperative day 1 were 4.00 (interquartile range, 3.00–5.00) for ECS and 5.00 (interquartile range, 4.00–7.00) for TCS, respectively ( P = .031). Analgesic requirements, intraoperative nausea, and postoperative shoulder pain were significantly less after ECS. Overall operative time was significantly shorter in ECS, with no difference in delivery time. No bladder injury occurred in either group. There were no differences in estimated blood loss and neonatal outcome. Urogenital distress, urinary tract infection, and bowel dysfunction did not differ at discharge from hospital and 6 weeks after.
Conclusion
An extraperitoneal approach to cesarean section appears to reduce postoperative pain, usage of analgesics, and intraoperative nausea without an increase in significant complications.
Click Supplementary Content under the article title in the online Table of Contents
Cesarean section is the most frequent major abdominal surgery among women worldwide. Delivery of a baby by cesarean section commonly leads to postoperative abdominal pain and may cause a variety of surgical complications. With high cesarean section rates, obstetricians are challenged to reduce perioperative and postoperative morbidity by adapting their surgical skills and techniques.
The standard technique for cesarean section in medical centers around the world is a low transverse cesarean section modified by the Misgav Ladach method. This approach entails opening the parietal peritoneum and delivering the baby transperitoneally through a transverse incision in the lower uterine segment. Exposure of the peritoneal cavity to blood, amniotic fluid, vernix, and mechanical irritation is thought to potentially aggravate pain, nausea and vomiting, dysfunction of intestines, voiding disturbances, adhesions, and infertility.
It has been speculated that an extraperitoneal approach to cesarean section may protect intraperitoneal organs and the peritoneum by avoiding potential harmful effects and thus reduce morbidity.
Anecdotal reports of extraperitoneal sections date back to the year 1821. In the era before antibiotics, extraperitoneal cesarean section (ECS) was thought to protect the peritoneal cavity from spreading intrauterine infection. There is very limited modern literature on the extraperitoneal procedure in the English language and the majority of the literature peaks in the 1940s and 1950s, focusing on technical details. In 1949, Cacciarelli introduced a method applying a paravesical approach, which was later modified by Waters to a technique with a supravesical access. Wallace et al and Hanson in 1984 and Zabransky and Grossmannová in 1985 and 2001 aimed at a reduction of septic morbidity. They concluded that the extraperitoneal section was beneficial but had never been introduced as a routine method due to a lack of training.
The more recent observational studies of ECS reported quick postoperative recovery and reduced need for analgesics. In 2009 the technique was relaunched combining a horizontal Pfannenstiel-like skin incision with a vertical fascial incision and a left paravesical dissection of the space inferior and lateral to the vesicouterine fold. This procedure is advertised as “outpatient cesarean section” with no supporting evidence for that statement.
Possible complications of ECS do not appear to differ from those of transperitoneal cesarean section (TCS). In the extraperitoneal procedure, injuries to the bladder have been reported in <1% ; injuries to the ureters are anecdotic. Endometritis remains a rare condition in any cesarean section in medically well-developed countries. Delivery time is reported to be equal among different techniques.
To date there has been no prospective, randomized study comparing traditional TCS with ECS. The aim of the present study was to compare TCS with ECS with respect to a variety of intraoperative and postoperative complications and fetal outcome. It was hypothesized that ECS would lead to less postoperative abdominal pain.
Materials and Methods
This prospective, randomized, single (patient) blinded study was conducted at a single academic tertiary care center (Department of Obstetrics and Gynecology, Medical University of Graz, Graz, Austria). Patients were invited to participate if they were age ≥18 years and presented with an indication for an elective cesarean section (primary or first repeat cesarean section) at ≥37 weeks of gestation. Patients with suspected abnormal placentation (placenta accreta, increta, or percreta), placenta previa, previous vertical uterine incision, a history of >1 cesarean section, a history of major abdominal surgery, or fetal macrosomia (estimated fetal weight >4500 g) were excluded. Eligible patients were approached by members of the research team when they were scheduled for cesarean deliveries. Informed consent was obtained from all participants; the trial was conducted in strict accordance with the protocol and was approved by the local institutional review board. Randomization and treatment allocation was done before entering the operating room. Spinal anesthesia with a uniform protocol based on ropivacaine and fentanyl was suggested to all patients. Surgeries were done by attending physicians and residents. Resident physicians as primary surgeons were assisted by attending physicians until the closure of the oblique muscles’ aponeurosis.
The primary endpoint of the study was maximum abdominal pain (peak pain) on postoperative day (POD)1. Postoperative abdominal pain was measured by numeric rating scales (NRS) ranging from 0–10. Nurses were blinded to treatment randomization and documented the level of perceived pain every 8 hours until discharge. The postoperative care–early ambulation without dietary restrictions–and analgesic protocol were standardized and did not differ between surgical groups. Maintenance therapy with intravenous nonsteroidal antiinflammatory drugs (diclofenac, 75 mg every 12 hours) was offered on POD1 and POD2, followed by oral diclofenac (50 mg every 8 hours) on POD3. Systemic opioid analgesics (piritramide, 15 mg every 6 hours) were offered for breakthrough in case of a pain measurement of ≥5 on NRS. Patient-controlled epidural analgesia and patient-controlled analgesia were not utilized.
Secondary endpoints were intraoperative nausea and vomiting, postoperative thoracic and shoulder pain, urogenital dysfunction, and bowel dysfunction. Staff physicians blinded to treatment randomization evaluated the secondary endpoints and symptom severity on the day of discharge by means of a questionnaire with NRS corresponding to the 0-10 pain scale. A rating of “0” corresponded to the best result whereas a rating of “10” corresponded to the worst result of any particular question. Urogenital distress was determined using a Urogenital Distress Inventory (UDI). The UDI reflects 3 aspects of urogenital dysfunction: irritative symptoms, obstructive discomfort, and stress symptoms. The 3 original subscales ranging from 0–100 were adapted for consistency to the 0-10 NRS.
The category bowel dysfunction included any intestinal problems or discomfort and did not particularly distinguish between symptoms of constipation or diarrhea. Consistently, the evaluation followed the 0-10 NRS. Other secondary outcomes were as follows.
All surgical complications were recorded and blood loss was estimated by the difference in hemoglobin on the day of surgery and POD1. Neonatal outcome was defined by Apgar scores, arterial umbilical pH values, and neonatal weight. Delivery time (skin incision–delivery), operative time (skin incision–skin closure), need for analgesics, and length of postoperative hospital stay were recorded.
Surgical procedures
For both techniques a Foley catheter was placed preoperatively and a single dose of a first-generation cephalosporin was given intravenously before the skin incision.
The technique of the extraperitoneal procedure was based on the descriptions by Norton, Cacciarelli, and Fauck. After Pfannenstiel incision the rectus muscles were bluntly separated in the midline to expose the preperitoneum and the bladder. The Foley catheter was blocked preoperatively to better visualize and to ease preparation of the bladder. Sweeping aside the preperitoneal fat with a sponge stick identifies the left medial umbilical fold containing the obliterated umbilical artery. This structure may be termed lateral umbilical ligament in some older publications and is the medial border of the paravesical space and the lateral border to the bladder. This avascular space can be bluntly dissected; a retractor may be placed lateral to the left medial umbilical fold with traction made medially and inferiorly. With these maneuvers and blunt dissection, the ligament and the posterior surface of the bladder are separated from the lower uterine segment. At this point the tongue of the peritoneum of the vesicouterine fold becomes visible superiorly. This tongue-like fold courses downward and medially and can be followed throughout its course. By inserting the fingers or a sponge stick beneath the peritoneal fold, the bladder can be further pushed medially (to the right) and inferiorly until the space is large enough to deliver the infant. The dissection is performed bluntly and essentially bloodlessly. The exposed lower uterine segment is now incised and the incision is extended digitally towards both sides ( Figure 1 ). The baby is delivered and the placenta is removed as it is done in TCS. The uterine incision is closed with a single-row continuous absorbable suture; the abdominal wall is closed in the conventional Pfannenstiel manner ( Video Clip ).
The TCS is performed as described previously. In brief, blunt dissection of the abdominal wall and parietal peritoneum, omission of the bladder flap, blunt uterine incision expansion, and spontaneous placental removal were applied. The uterus was not exteriorized and the visceral and parietal peritoneum were not closed. Running intracutaneous suture closure of the skin was used in all cases.
Study design, statistics
This study was designed as a prospective, randomized, single (patient) blinded, single institutional investigation. The primary endpoint of the study is maximum abdominal pain (peak pain) on POD1 measured by NRS ranging from 0–10. Prior to the study a mean peak pain measurement of 5.0 with SD of 1.8 was recorded among 30 traditionally (transperitoneally) operated patients at our institution. A reduction to a mean peak pain measurement of 3.5 was expected for the patients with ECS. On this base (calculated by difference of means [1.5] and effect size [0.83]) it was estimated that at least 48 completed patients had to be included into the study to achieve a 5% level of significance and a power of 80%. We increased the sample size to 54 to allow up to 10% loss to follow-up for the primary endpoint. Patients with a primary cesarean section and with 1 previous cesarean section were randomly assigned to TCS or ECS following the biased-coin method using an Internet-based randomization service for clinical trials. Primary analysis was performed according to the intention-to-treat principle.
The Shapiro-Wilks test and visual data inspection were used to examine for normal distribution. Normally distributed continuous variables and categorical data were compared using the double-sided Student t test and the χ 2 test, respectively. Variables that were not normally distributed as assessed by the Kolmogorov-Smirnov test were compared using the Mann-Whitney U test. The nonparametric methods were performed particularly for ordinal NRS data. Data are presented as mean ± SD and median with range and/or interquartile range (IQR), respectively. P values < .05 were considered statistically significant. All statistical analyses were performed using software (SPSS, version 19.0; IBM Corp, Armonk, NY).
Results
A total of 54 patients were included in the study from September 2011 until March 2012, of whom 27 were randomly assigned for ECS and 27 for TCS, respectively ( Figure 2 ). Surgeries were performed by 1 attending (n = 13) and 2 resident (n = 6 and 8) physicians for ECS and by 2 attending (n = 2 and 3) and 4 resident (n = 5, 5, 6, and 6) physicians for TCS, respectively. Attending and resident physicians’ experience was a minimum of 10 and 3 years, respectively.
Two patients with primary intention for ECS had accidental opening of the peritoneal cavity followed by a transperitoneal procedure. Patients’ characteristics (age, parity, body mass index, history of cesarean section) and indications for cesarean section were similar in the 2 groups ( Table 1 ). Spinal anesthesia was accomplished in 51 cases and conversion to general anesthesia was required in 3 cases due to insufficient intraoperative pain management. Three patients (2 ECS, 1 TCS) could not be contacted after 6 weeks and were lost to long-term follow-up.
| Variable | CS | P value | |
|---|---|---|---|
| Extraperitoneal n = 27 | Transperitoneal n = 27 | ||
| Age (y), mean ± SD | 29.9 ± 4.9 | 31.4 ± 7.2 | .384 |
| BMI (pregestational), mean ± SD | 22.4 ± 2.4 | 24.2 ± 3.9 | .057 |
| Gestational age (wk), mean ± SD | 38.8 ± 1.2 | 39.2 ± 0.9 | .083 |
| Parity, n (%) | .392 | ||
| Nulliparous | 16 (59) | 13 (48) | |
| Multiparous | 11 (41) | 14 (52) | |
| History of CS, n (%) | 7 (26) | 10 (37) | .309 |
| Indication, n (%) | .084 | ||
| Breech presentation | 11 (41) | 9 (33) | |
| History of CS | 7 (26) | 10 (37) | |
| Psychological | 2 (7) | 2 (7) | |
| Preeclampsia | 3 (11) | 0 (0) | |
| Fetal malformation | 2 (7) | 0 (0) | |
| Other | 2 (7) | 6 (22) | |
Main outcome measures
Postoperative pain score registration was thorough and its chronological assignment is shown in Table 2 .
| Postoperative day and time | Cesarean section | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Extraperitoneal | Transperitoneal | ||||||||
| n | NRS | NRS mean | NRS peak | n | NRS | NRS mean | NRS peak | ||
| Day 1 | 00:00-08:00 | 25 | 3.00 (2.00-5.00) | 3.00 (2.33–3.67) | 4.00 (3.00–5.00) | 26 | 4.00 (3.00-6.00) | 3.67 (2.67–5.00) | 5.00 (4.00–7.00) |
| 08:00-16:00 | 25 | 3.00 (2.00-4.00) | 24 | 3.00 (2.00-5.00) | |||||
| 16:00-24:00 | 27 | 3.00 (2.00-4.00) | 27 | 3.00 (2.00-5.00) | |||||
| Day 2 | 00:00-08:00 | 27 | 2.00 (1.00-3.00) | 1.67 (1.00–2.67) | 3.00 (2.00–4.00) | 27 | 2.00 (1.00-4.00) | 2.00 (1.33–3.33) | 3.00 (2.00–4.00) |
| 08:00-16:00 | 25 | 1.00 (0.00-3.00) | 24 | 2.00 (1.00-3.00) | |||||
| 16:00-24:00 | 27 | 1.00 (1.00-2.00) | 25 | 2.00 (1.00-4.00) | |||||
| Day 3 | 00:00-08:00 | 27 | 1.00 (0.00-2.00) | 1.00 (0.00–2.00) | 1.00 (0.00–2.00) | 26 | 2.00 (0.00-3.00) | 1.00 (0.00–2.00) | 2.00 (0.00–3.00) |
| 08:00-16:00 | 18 | 1.00 (0.00-1.00) | 21 | 1.00 (0.00-2.00) | |||||
| 16:00-24:00 | 23 | 1.00 (0.00-1.00) | 25 | 1.00 (0.00-2.00) | |||||
Patients after ECS had significantly less maximum surgical site pain than patients after TCS. In the intention-to-treat analysis median peak pain scores on POD1 were 4.00 (IQR, 3.00–5.00) for ECS and 5.00 (IQR, 4.00–7.00) for TCS, respectively ( P = .031) ( Figure 3 ).
During ECS patients suffered significantly less from intraoperative nausea and vomiting compared to TCS (median score 0.0 [range, 0.0–1.0] vs 3.0 [range, 0.0–9.0], P < .001). Postoperative shoulder pain was noted significantly less in the ECS compared to the TCS group (median score 0.0 [range, 0.0–3.0] vs 0.5 [range, 0.0–9.0], P < .001). Evaluation of bowel dysfunction and UDI scores at discharge from hospital did not differ between the 2 groups ( Table 3 ).
