Outcome after a high number (4–10) of repeated caesarean sections

Authors


Dr K. Juntunen, Family Federation of Finland, Oulu Clinic, Medipolis, Kiviharjuntie 11, 90220 Oulu, Finland.

Abstract

Objective  To evaluate outcomes in caesarean sections repeated several times.

Population  Sixty-four women had had four or more (up to 10) caesarean sections.

Methods  The outcomes of these operations N= 149, study group) were compared with other caesarean sections.

Results  A quarter of the women in the study group complained of low abdominal pains during the late pregnancy. A thin or fenestrated isthmic myometrial layer was observed in 55% of their operations. Abnormal placentation with an increased risk of major operative complications occurred more often in the study group. No differences in post-operative complications or perinatal outcome emerged between the groups.

Conclusion  Thus, no definitive upper limit for the number of caesarean sections per individual woman can be given.

INTRODUCTION

One of the main indications for caesarean delivery is repeat caesarean section, which has been shown to involve an increased risk of placenta praevia and some other specific complications.1,2 Until now, no one has recommended a definite upper limit for the number of caesarean sections to be performed on an individual woman.

We report here the operative outcomes and clinical findings of patients who had had a high number of repeated caesarean sections.

METHODS

The patient files of 64 women who had undergone four or more caesarean sections in the Department of Obstetrics and Gynecology, University of Oulu, Finland, were evaluated. These women had had a total of 341 caesarean sections. The number of these patients' first to third operations was 192 and the number of their fourth or subsequent caesarean sections (median 5.2, range 4–10) was 149 (the study group) during the years 1982–2002. All of their 4th to 10th pregnancies and deliveries were managed by the same hospital. The total number of deliveries in the area during this same period was approximately 73,000.

In order to characterise the general outcome of a typical caesarean section during the whole study period, we analysed a control case for every 4th to 10th operation of the study group by selecting the next caesarean section of the same category (elective/emergency, with the study cases excluded) in our clinic (control group, N= 149).

In the primary operations, a Pfannenstiel incision was made in 65% of the study cases, while after four to five operations a midline vertical incision was used. All caesarean sections consisted of an isthmic transverse uterine incision, manual evacuation of the placenta, exteriorisation of the uterus after delivering the infant, resection of the membranous or fibrotic borders of the isthmic uterine wound if necessary, double-layer suturing of the myometrium, careful drying of the pelvic cavity and closure of the visceral and parietal peritoneum. Different visual estimates of the condition of the isthmic myometrium were used: (1) membranous or lacerated isthmic layer (fenestration), through which the fetal membranes could be seen, (2) non-transparent but very thin layer of muscular fibres, with an estimated thickness <2 mm, (3) normal or almost normal layer with clearly recognisable muscular tissue.

Mantel–Haenzel's χ2 test for categorical variables and Student's t test for means of continuous variables were used. Odds ratio and 95% confidence interval were used to establish the proportional rate of differences between the case and control groups.

RESULTS

In the study group, the incidence of pregnancy-associated complications, such as pre-eclampsia, preterm delivery, intrahepatic cholestasis of pregnancy, gestational diabetes or antepartum fetal distress, was similar to that seen in the controls. Instead, the incidence of chronic illness, such as hypothyreosis, diabetes mellitus, essential hypertension or collagenous disease, was significantly higher in the study group (data not shown).

In the third trimester, the patients in the study group reported abdominal pains more often than the control patients (Table 1). The operative finding of thin or membranous isthmic myometrium was also made more often in their operations. Forty percent of the patients (6/15) in the study group with a peri-operative finding of a membranous isthmic scar had no subjective symptoms. Out of the 34 pregnancies in the study group with ‘scar pains’, all but five presented with a thin or membranous isthmic myometrium at operation. The mean gestational ages at the time of the operation in the study group (38.0 weeks) and the control group (39.0 weeks) did not differ significantly, although the risk of preterm delivery was increased in the study group (odds ratio 2.1, confidence interval 1.2–3.6).

Table 1.  Pre-operative pains at lower abdominal area and operative conditions, findings and complications according to the number of caesarean sections (CS).
 4th to 10th CS (N= 149)Controls CS (N= 149)Odds ratio95% Confidence IntervalP
Pains at lower abdominal area (%)22.80.744.15.9–327.30.00001
<2 mm isthmic myometrial layer (%)55.02.060.418.4–198.30.00000
Membranous, transparent or lacerated myometrial layer (%)10.10  0.00007
Intraperitoneal adhesions (%)18.22.78.12.7–23.80.00001
Cranial bladder attachment (%)43.67.49.95.0–19.90.00000
Blood loss ≥1000 g (%)12.110.71.10.6–2.30.71
Major peri-operative complication (%)4.70.77.30.9–60.10.03153
Placenta praevia verified at operation (%)5.40.78.41.0–68.00.01782
Abruptio placentae (%)3.40  0.02413
Difficult placental evacuation (%)5.40.78.41.0–68.00.01782
Post-operative infections (%)14.114.80.90.5–1.80.86928

Intraperitoneal adhesions were more common in the study group, as was also high and tight attachment of the bladder flap over the isthmic area. Mean blood loss was slightly greater in the study group (642 g vs 605 g), but blood loss >1000 g occurred with similar frequency in both groups. The major operative complications, seen in 4.7% (seven operations) and defined as difficult delivery of the infant due to a scarred abdominal wall, deep myometrial ruptures into the broad ligament or towards the uterine cervix, transient uterine atonia with bleeding >1500 g, emergency hysterectomy due to haemorrhage or a transient peri-operative anaesthesiological complication, were more common in the study group than in the controls.

The placental findings and problems are also presented in Table 1. The incidence rates of placenta praevia and tightly attached placenta were eightfold in the study group compared with the controls. Severe complications due to placenta praevia or placenta accreta occurred in five of the 149 operations in the study group. Emergency hysterectomy was necessary in four of them in spite of the availability of radiological embolisation, extensive vascular ligation procedures and intensive peri- and post-operative care.

The incidences of infectious post-operative complications did not differ between the groups. The percentage of pregnancies without any post-operative complications was 89% after the 4th to 10th operations and 82% in the control group. Primary healing of the abdominal wound was similar in all groups, and all mothers survived. Neonatal outcomes did not differ in the groups.

DISCUSSION

One to two percent of caesarean sections are complicated by a significant intra-operative problem.3 Emergency operations and repeated sections have been found to pose a specific risk.3,4 In mothers with four or more caesarean sections, the perinatal outcome was good and full-term pregnancy was possible in most cases. With the exception of placental problems, the increase of peri- and post-operative complications did not have much clinical significance. Kirkinen,5 in 1988, reported almost identical results from the same clinic concerning patients with three caesarean sections. That study covered mothers whose last caesarean section had been performed during 1978–1986, approximately 15 years earlier than in the present study. Since the late 1980s, early mobilisation and early per oral nutrition have been increasingly used after caesarean sections. This may improve the outcome by, for example, reducing the risk of post-operative thromboembolism. Similarly, regional anaesthesia allows more time for safe preparation of the abdominal wall.

An important point in the care of our patients was neither to allow them to go into labour nor to extend the period with ruptured membranes before the operation. This practice possibly helped to avoid increasing technical difficulties at operation and decreased the rate of infectious complications or even the risk of uterine rupture. Our choice of using a vertical abdominal wall opening after three to four previous Pfannenstiel incisions may not have been absolutely necessary, but it did, however, allow more room and good visualisation for safe preparation of the isthmic area and atraumatic delivery of the baby. This—along with our previous observation that the urinary bladder in these patients may often be attached high cranially to the parietal peritoneum—may have enabled us to perform the operations with no accidental cystotomies or bowel injuries. The peritoneal opening was thus made near the umbilicus to avoid the bladder fundus. The primary healing of the wound was good and infectious post-operative complications did not increase compared with the control operations.

Twenty-three percent of the women with four or more caesarean sections had had abdominal pains in the late pregnancy, and this finding was almost always associated with a thin or membranous isthmic layer. Myometrial thinning usually, although not always, progressed with the increasing number of previous caesarean sections. Our technique included resection of the thin, fibrotic borders of the uterine incision before a double-layer closure of the myometrium. Thus, a considerably better myometrial condition was seen in some patients at a later operation.

In our series with 149 operations, the incidence of placenta praevia was 5.4%, and emergency hysterectomy was indicated in four of these eight cases. In the normal population, the incidence of placenta praevia is less than 0.5% of deliveries, but becomes two- to threefold in women with previous caesarean section.6–8 This risk also increases with greater parity, independent of the number of prior caesarean sections. The likelihood of placenta praevia is eight- to ninefold among women with parity greater than four and with more than four caesarean sections.2 Our results are basically in concordance with the previous findings on placental complications. Pre-operative ultrasonography will help to diagnose a high risk situation in which all available therapeutic facilities must be provided before the operation.

To conclude, even after three or more caesarean sections, full-term pregnancy is achievable in spite of the tendency towards a thin and scarred isthmic myometrium and maternal subjective sensations of pain in this area. The risk of abnormal placentation is increased with an increased risk of major operative complications. The post-operative complication rate, however, is low. According to our findings, no absolute upper limit for the number of caesarean sections per individual woman can be presented.

Accepted 27 February 2004

Ancillary