Alternative techniques and materials for caesarean section
There is no clear evidence to suggest that extra-abdominal repair of the uterine incision is superior to the intra-abdominal procedure. As some indirect evidence suggests that leaving the peritoneum unsutured could be beneficial, a policy of non-closure of the peritoneum could be considered. As to closure versus non-closure of the subcutaneous tissues, there is no enough evidence to justify a change of practice in the hospital settings.
RHL Commentary by Abalos E
1. EVIDENCE SUMMARY
A large number of techniques and materials for caesarean section have been proposed to reduce the operating time, the hospital costs and to make the procedure easier for the surgeon. However, whether any of them are better or worse than the alternatives for mother’s subsequent health is generally not known and few of these interventions have been rigorously evaluated before being incorporated into practice.
Systematic reviews of randomised controlled trials evaluating different techniques and materials for caesarean section have been recently published. Some of these interventions are discussed here.
Whether suturing the uterine incison inside the abdomen or by exteriorising the uterus through the abdominal wall is evaluated in the Extra-abdominal versus intra-abdominal repair of the uterine incision at caesarean section review. Another review focuses on whether there is a need to suture the peritoneum “Closure versus non-closure of the peritoneum at caesarean section”. For the closure of the abdominal wall, the systematic review“Techniques and materials for closure of the abdominal wall in caesarean section” evaluates alternative options.
Extra-abdominal versus intra-abdominal repair of the uterine incision at caesarean section.
The review includes six trials evaluating 1221 women undergoing both, emergency and elective caesarean sections. The intervention was the exteriorisation of the uterus versus the intra-abdominal repair of the uterine scar after the delivery of the baby and the placenta. The primary outcome measures were blood loss (pre-and post-operative laboratory measures) and post-operative sepsis (defined by trial authors). Secondary outcomes include a range of variables measuring infection-related complications (such as post-operative pyrexia, wound infection), haemorrhage (blood transfusion, intra-operative blood loss), surgical complications, women and provider’s satisfaction, and costs, among others.
Neither blood loss nor post-operative sepsis showed significant differences in external or intra-abdominal repair groups. Post-operative drop in haematocrit was evaluated in 2 trials including 324 women. The weighted mean difference (WMD) was -0.47, 95% Confidence Interval (CI) ranging from -1.48 to 0.54. Two other trials (482 women) measured post-operative drop in haemoglobin, with similar results: WMD 0.02, 95%CI -0.62 to 0.65. Post-operative sepsis was reported in one trial (308 women), relative risk (RR) 0.94, 95%CI 0.19 to 4.17.
Although the authors of the review suggest a marginal but significant reduction in febrile morbidity in the uterine exteriorisation group, these results should be interpreted cautiously, as they come from only one trial published in 1978 in which more than 20% of women were excluded after randomisation, presumably due to surgical complications during the caesarean section. Furthermore, this outcome was modified post-hoc into febrile morbidity for more than 3 days, when in the type of outcome measures section this is described as post-operative pyrexia. Febrile morbidity (as per protocol definition) was reported in at least 2 other trials (Whalab 1999, Edi-Osagie 1998), not included in the meta-analysis.
All trials were conducted in developed countries, one in the seventies and the rest in the nineties. One of the trials excluded more than 20% of women after randomisation due to the need to perform surgical manipulations during the operation. Nearly one third of the women received antibiotics during surgery. In more than half this information was not available.
Sub-group analyses were added post-hoc by the reviewers taking into account whether the placenta was manually removed or spontaneously delivered, as three of the included studies pre-specified these subgroups.
Closure versus non-closure of the peritoneum at caesarean section:
The review, including 9 trials involving 1811 women, makes the following comparisons: (i) non-closure of both visceral and parietal peritoneum versus suturing both visceral and parietal peritoneum; (ii) non-closure of the visceral peritoneum only versus suturing both parietal and visceral peritoneum; and (iii) non- closure of the parietal peritoneum only versus suturing both parietal and visceral peritoneum.
The main outcome measures were operating time, analgesic requirement, postoperative fever, endometritis, length of hospital stay, wound infection, wound dehiscence, and adhesions at the subsequent operation. There were no data evaluating the last two outcomes.
Non-closure of both layers (WMD -7.33 minutes, 95% CI -8.43 to -6.24) the visceral layer (one trial, 549 women; WMD; -6.30 min., 95% CI -9.20 to -3.40 min.) or the parietal layer (one trial, 248 women; WMD; -5.10 min, 95% CI -8.71 to -1.49 min) all seemed to reduce the operation time.
These points are not surprising as the intervention group involved the avoidance of one (or two) steps of the standard procedure performed in the control group. Post-operative fever was significantly lower (9.6% vs. 14.4%) when both layers were left unsutured (five studies, 874 women). However, the results must be interpreted with caution as there was clinical heterogeneity among the trials, with incidences ranging from 0% to 23% in the experimental arm and from 0% to 30% in the control group. When both visceral and parietal peritoneums were not sutured the result was shorter stay in hospital (six trials, 974 women). However, clinicians should evaluate the relevance of this 9-hour reduction (ranging from 6 hours to 12 hours) according to the availability of beds in their own settings. Again, they should be cautious as the post-operative hospital stay in the different trials ranged from 2.9 days to 8.3 days in the experimental group and from 2.8 days to 9 days in the control group. This could be associated with differences in hospital protocols for post-surgical care at different settings, or to different incidence of post-operative complications in these hospitals. Only two trials involving 393 women assessed the use of analgesics, showing no differences between groups. Other clinically relevant outcomes such as endometritis and wound infection were evaluated in only two and three trials, respectively, with no statistically significant differences. No economic analyses were conducted in the studies included in the review.
All adequately controlled trials that could be identified have been included in this review. The authors made attempts to contact authors of published abstracts, unpublished data or ongoing studies for details on methodology and/or results. The trials were evaluated for methodological quality and appropriateness of inclusion, and a sensitivity analysis was performed by excluding trials with inadequate allocation concealment.
Techniques and materials for closure of the abdominal wall in caesarean section:
This review was designed to evaluate the effects of alternative techniques and materials for closure of the rectus sheath and subcutaneous fat at caesarean section on morbidity and the use of healthcare resources. The authors defined short-term outcomes (related to infectious and haemorrhagic morbidity, pain, complications and duration of surgery) as well as long-term outcomes (including fertility problems or complications in future pregnancies). Health service use (length of hospital stay or re-admission to hospital) was also evaluated.
Six trials involving 1853 women evaluated the effects of closure versus non-closure of the subcutaneous tissues, and one trial of 203 women assessed the effects of blunt versus sharp needles for closing of all layers at caesarean section. No trials were found looking at different techniques or materials for closing the rectus sheath.
All but one trial were conducted in developed countries between years 1991 and 2001. Women were recruited at their first or repeated caesarean section, either electively or intrapartum. Women with transverse or vertical skin incision were included in the review. Subcutaneous drain was a comparison arm in one of the studies, in other two it was not used, and in the remaining two there were not specifications on whether or not subcutaneous drain was part of the surgical procedure. No information was provided regarding prophylactic use of antibiotics.
No differences were found in the incidence of wound infection when the subcutaneous tissue (fat and/or camper fascia) were sutured or left unsutured (five trials, 1348 women; RR 1.02, 95%CI 0.69 to 1.50). The authors of the review combined the outcomes wound haematoma and wound seroma as a single outcome, which was statistically significantly lower in the closure group (RR: 0.52, 95%CI 0.33 to 0.82). However, it is questionable if these outcomes could be combined, as their clinical implications (like the need for extra appointments to the health care system, re-admission to hospital, surgical drainage or other procedures) may differ. Moreover, seroma (the commonest complication found according to the authors) is a subjective diagnosis that may be influenced by knowledge of the allocated group.
Although the authors of the review concluded that the methodology of the seven trials appeared to be generally satisfactory, there is a high probability of selection, attrition and/or detection bias. One of the trials (164 women) has possibly not concealed the allocation sequence (randomised from a “list” in theatre). The rest have attempted allocation concealment through sealed opaque envelopes, not mentioning if they were consecutive and numbered, so it is not possible to determine whether a woman was excluded after an envelope was opened. Assessment at follow-up was blinded in two trials, not blinded in other two, and not documented in the others. One of the trials excluded nearly 40% of the women after randomisation, as one of the inclusion/exclusion criteria was applied afterwards. All trials followed-up women until hospital discharge. After discharge, in two studies women “were encouraged to return to hospital only if they developed problems”. In another, women were followed-up through a retrospective chart review. The trials that systematically and prospectively tried to follow-up women failed to find more than 20% of them. One of the studies (451 women) did not report results in an intention-to-treat basis. The authors of the review stated that some of the trials reports were unclear about the number of women assessed for each outcome, and that assumptions were made when conducting the review.
No conclusions can be made when blunt needles versus sharp needles are used for closure of the abdominal wall. The only trial included is too small for any reliable estimate.
2. RELEVANCE TO UNDER-RESOURCED SETTINGS
2.1. Magnitude of the problem
An alarming increase in the incidence of caesarean sections has been observed in both developing and developed countries over the last 20 years. In 1998 approximately 15% of women in the United Kingdom underwent delivery by either elective or emergency caesarean section (1). In 2001 the rate was 21% (2). Caesarean section is the most common major surgical procedure in the United States (3),where rates increased, from 5.5% in 1970 to 24.4% in 2001 (4,
5). Rates vary considerably within and between countries. In Norway wide variations in caesarean section rates (between 6% and 20%) have been reported between obstetric departments(6).Rates around the world vary between 5% to over 20% of all deliveries(7). During a two-year period in an urban area of India, total caesarean section rates in the public, charitable and private sectors were 20%, 38% and 47% respectively (
(8).Figures on caesarean section rates in Latin America are not easily available, as systematically collected national records are not maintained in many of the countries of the region. Chile recorded an incidence of 37.2% in 1994 (9), an increase of one-third compared with 27.7% in 1986. Brazil reported national rates of 32% in 1986 (10), reaching levels of over 70% in some units (11,
Reports from a Venezuelan hospital indicate a rate of 44.8% (16).
As the operation is conducted so frequently, any attempt to reduce morbidity, even with relatively modest differences for a particular outcome, is likely to have significant benefits in terms of costs and health benefits for the women. As for the review on "Antibiotic prophylaxis for caesarean section", the above described reviews have also critical relevance for under-resourced settings.
2.2. Applicability of results
Many of the trials included in the reviews were conducted in developed countries. Thus the universal applicability of the intervention at secondary care settings needs to be evaluated. There is little information about ability and skills of the attending physicians, as well as whether the hospitals were in the public or private sector or whether they were teaching hospitals or not. No information is given in many trials about the concomitant use of "Prophylactic antibiotics during the procedure", which may have implications for the reductions in postoperative fever rates.
Regarding the results, it is difficult to specify the main outcomes since different caesarean section techniques and procedures were used in the trials covered in all three reviews. Febrile morbidity, post-operative infection rates, antibiotic use, wound complications, haemorrhage, pain, need for analgesia and length of stay in hospital reflect short-term maternal morbidity. However, a surrogate measure for maternal morbidity, i.e. duration of operation, is the most often recorded outcome for trials evaluating Closure vs. non-closure of the peritoneum at caesarean section. Because it is an easy-to-measure outcome and is expressed as a continuous variable, a smaller sample size is needed to demonstrate a given statistically significant difference. The use of surrogate outcomes may not be appropriate as they assume a correlation between maternal morbidity and the outcome measured. This relationship may be misleading. A longer operation time may not necessarily be associated with post-operative maternal morbidity if, for example, the technique being used causes less tissue damage or introduces less foreign material into the wound, thus reducing the likelihood of morbid complications.
The long-term morbidity outcomes are difficult to evaluate within the context of randomised controlled trials. Adhesions may be asymptomatic in most cases and to show a difference among the symptoms (pain, dyspareunia) or other morbidity (such as secondary infertility) a large number of women would need to be evaluated. Direct and indirect costs to both the health system and users were not evaluated in the trials.
2.3 Implementation of the intervention
Future research should be aimed at evaluations of costs and women's views on post-operative discomfort associated with the surgical technique. Clinically relevant short-term as well as long-term benefits or complications of the above mentioned procedures need to be evaluated in large pragmatic randomised controlled trials. A multi-centre trial of techniques of caesarean section is in progress (Caesar study – National Perinatal Epidemiology Unit, Oxford, United Kingdom).
Sources of support: Centro Rosarino de Estudios Perinatales. Rosario. Argentina.
- Percentage of births performed by Caesarean section. Office of National Statistics, London, United Kingdom (www.statistics.gov.uk last visited 18 December 2003) Social trends 31 .
- Thomas J, Paranjothy S. Royal College of Obstetricians & Gynaecologists, Clinical Effectiveness Support Unit. The National Sentinel Caesarean Section Audit Report. London. RCOG Press;2001.
- Taffel SM, Placek PJ, Moien M, Kosary CL. 1989 US caesarean section rate steadies: VBAC rate rises to nearly one in five. Birth 1991;18:73-77.
- Sandmire HF. A Guest editorial: Every obstetric department should have a caesarean birth monitor. Obstetrical and gynecological survey 1996;51:703-704.
- Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park M, Sutton PD. Births: final data for 2001. National Vital Statistics Reports 2002;51 (2):1-102.
- Annual Report 1999-2000. Bergen Medical Birth Registry of Norway 2002;.
- Lomas J, Enkin M. Variations in operative delivery rates. In: Chalmers I, Enkin M, Keirse MJNC (eds). Effective care in pregnancy and childbirth, 1989, Oxford University Press, Oxford. Oxford University Press, Oxford;1989.
- Sreevidya S, Sathiyasekaran BWC. High caesarean rates in Madras (India): a population-based cross-sectional study. BJOG: an international journal of obstetrics and gynaecology 2003;110:106-111.
- Murray SF and Serani Pradenas F. Health sector reform and rise of caesarean birth in Chile. The Lancet 1997;349:64.
- Notzon FC. International differences in the use of obstetric interventions. JAMA 1990;263:3286-3291.
- Barros FC, Vaughan JP, Victora CG, Huttly SRA. Epidemic of Caesarean sections in Brazil. The Lancet 1991;338:167-169.
- Barros FC. Caesarean sections in Brazil (letter). The Lancet 1996;347:839.
- Loffredo LC and Simoes MJ. Birth weight and patterns of deliery care in a municipality of Sao Paulo State, Brazil, 1986. Revista Saude Publica 1990;24:80-83.
- Faundes A, Cecatti JG. Which policy for Caesarean section in Brazil? An analysis of trends and consequences. Health policy and planning 1991;8:33-42.
- Potter JE, Berquo E, Perpetuo IHO, Leal OF, Hopkins K, Souza MR, et al. Unwanted caesarean sections among public and private patients in Brazil: prospective study. British Medical Journal 2001;323 (7322):1155-1158.
- Faneite-Antique PJ. Estadísticas perinatales. Revista obstétrica y ginecólogica Venezuel 1986;46:76-82.
This document should be cited as: Abalos E. Alternative techniques and materials for caesarean section : RHL commentary (last revised: 3 Decemberr 2004). The WHO Reproductive Health Library; Geneva: World Health Organization.