In summary, as outlined by Dahlke et al, there is good evidence to support the following practices at CS:

  • prophylactic antibiotics prior to skin incision
  • cephalad-caudad blunt uterine incision extension
  • spontaneous delivery of the placenta
  • oxytocin infusion for PPH prevention
  • single-layer uterine closure for women not desiring future fertility and closure of the subcutaneous tissue when the thickness is <2 cm.1

Further research is still required surrounding various CS techniques, especially in regards to long-term outcomes for double-layer uterine closure and peritoneal closure. In regards to prophylaxis, ideal antibiotic prophylaxis for obese women, who now comprise a significant proportion of our population and are at higher risk of infectious morbidity, requires further research to ensure this is optimised, whether it be with a higher dose or more broad spectrum regime. VTE and PPH prophylaxis (with respect to the use of Tranexamic Acid) require further attention, given their importance in relation to maternal morbidity and mortality. Finally, as clinicians, we should endeavour to perform and teach this common operation using an evidence-based approach, and be willing to change our technique when the evidence shows us we ought to, for the wellbeing of the women we serve.

Caesarean section (CS) is the most commonly performed major operation for women worldwide. In Australia in 2011, 95 894 women gave birth by CS, which equates to approximately one in three deliveries.2 This rate has increased from 18 per cent of births in 1991, to the current rate of over 30 per cent.3 Compared with vaginal birth (VB), mothers who undergo CS have higher rates of morbidity, including the serious morbidities of major puerperal infection and haemorrhage leading to hysterectomy4 and mortality.5 In the past, Berghella et al summarised the evidence base for performing various steps in a CS6 and this advice has recently been updated.7 The purpose of this article is to present a snapshot of the current evidence-based guidance for making the myriad surgical decisions required during the performance of a CS, either elective or emergency, and to touch on some of the important preoperative considerations. We should endeavour to standardise the performance of the procedure and avoid unnecessary steps, in order to minimise morbidity and mortality for the women and babies under our care.

Preoperative considerations

Prophylactic antibiotics

There is strong evidence to support the use of a single dose of ampicillin or a first-generation cephalosporin, such as cephazolin, for all women undergoing CS, as this leads to a significant reduction in infectious morbidity (>60% less endometritis, 25% less wound infections in the elective setting and 65% less in emergency CS and an overall reduction in fever and urinary tract infection [UTI]).8 Timing of administration should ideally be 15–60 minutes prior to skin incision, as this has been shown to further reduce endometritis risk and total morbidity from infection without any impact on neonatal sepsis or neonatal intensive care unit (NICU) admissions.9 Outcomes are not improved with multiple doses and, until recently, no data supported using a more broad-spectrum regime.10

A large prospective multicentre randomised controlled trial (RCT) involving 2000 women has just been published, looking at the addition of azithromycin 500mg IV. This study found that for the azithromycin group, the rate of composite infectious outcomes was halved (6.1% versus 12%).11 The trial group was, however, high risk, with only those having had an emergency CS being included, and with greater than 70 per cent of women having a BMI of over 30.12 The question remains as to whether a higher dose of cephazolin for obese women would lead to the same reduction in infectious morbidity, as there are data to indicate that 2 g does not reach the minimal inhibitory concentration for this group of women.13 Further research is needed in this area.


This is an area largely devoid of evidence to guide our practice.14 However, venous thromboembolism (VTE) remains the second leading cause of direct maternal death in Australia,15 and is an important cause of the excess mortality associated with CS.16 As such, an assessment of the VTE risk of each woman should be consistently undertaken. A number of guidelines have been developed based largely on expert opinion, such as those by McLintock et al.17 As a minimum for all women undergoing CS, it is recommended that compression stockings, early mobilisation and adequate hydration be implemented.18 In addition, chemical thromboprophylaxis is recommended for all women having had an emergency CS for ≥5 days or until fully mobile. As such, an assessment of the VTE risk of each woman should be consistently undertaken. A number of guidelines have been developed based largely on expert opinion, such as those by McLintock et al.19

Vaginal preparation

Vaginal preparation preoperatively with povidone-iodine solution has been the subject of a Cochrane review, which included four trials (n=1198). The review found that endometritis rates were significantly reduced by this practice (5.2% versus 9.4%, RR 0.57), with the reduction being particularly significant for women with ruptured membranes (1.4% versus 15.4%, RR 0.13).20 Given this evidence, we should consider making this simple and cheap intervention part of our routine practice, particularly in the emergency setting.21

Use of an indwelling catheter

Interestingly, there are some data that support not placing an indwelling catheter (IDC) prior to CS, as the rate of UTI is reduced significantly (0.5% versus 5.7%) and there is no difference in the occurrence of urinary retention. However, to date, the available studies have been underpowered to detect a difference in the uncommon, but important, outcome of urinary tract injury.22

Skin preparation and hair removal

Until recently, there has been minimal good-quality evidence to guide our practice with regard to skin preparation, as pointed out in a Cochrane review conducted in 2014.23 Since then, two further RCTs have been performed comparing chlorhexidine-alcohol with iodine-alcohol; one which found no difference in surgical site infection24 and the second published recently, involving 1147 patients, that found a lower rate of surgical site infection with chlorhexidine-alcohol preparation (4.0% versus 7.3%, RR 0.55, P=0.02).13 In light of this, the current evidence would support the use of chlorhexidine-alcohol preparation.

When hair removal is necessary preoperatively, it is recommended to do so via clipping rather than shaving, as this is associated with fewer surgical site infections.

The procedure

Abdominal entry techniques

The Joel-Cohen technique, as compared to a Pfannenstiel incision, is associated with less fever, pain and analgesic requirements, reduced blood loss and operative time and a shorter hospital stay.25 26 No long-term data are available comparing the two, but given these short-term benefits, one should use the Joel-Cohen technique when feasible (it may not be practical in the setting of significant adhesions from prior surgery where sharp dissection is usually required). The Joel-Cohen technique involves making a straight transverse incision 3cm below the level of the anterior superior iliac spines, with the subcutaneous tissues opened only in the middle 3cm and the rectus sheath incised transversely in the midline.27 The subcutaneous tissue and rectus sheath is then extended laterally with blunt dissection, which is also used to separate the rectus muscles vertically and laterally and open the peritoneum.28

Development of a bladder flap is considered part of the Joel-Cohen technique, and this has been the subject of three RCTs.29 30 Omitting bladder flap development is associated with shorter incision to delivery intervals with no change in the rate of other complications; however, trials are currently underpowered to assess risk of bladder injury.31

Uterine incision

A lower segment transverse uterine incision is performed where possible. This should begin with a small incision made sharply with a scalpel. Blunt, as opposed to sharp, expansion of the uterine incision results in reduced blood loss and postpartum fall in haemoglobin.32 Furthermore, cephalad-caudad versus transverse blunt expansion results in significantly less unintended extensions of the uterine incision (3.7% versus 7.4%), and a reduction in estimated blood loss (EBL) over 1500 mL (0.2% versus 2.0%).33

Delivery of the placenta

There is strong evidence to support routinely practising spontaneous delivery of the placenta (using cord traction and fundal massage) as compared with manual removal. Spontaneous delivery has been associated with lower rates of endometritis, less blood loss, lower falls in haematocrit and shorter hospital stays.4,16

Preventing postpartum haemorrhage Available evidence supports the routine use of intravenous oxytocin or carbetocin for the prevention of postpartum haemorrhage (PPH) at CS. Various dosage regimens for oxytocin have been used in trials, which show benefit from oxytocin infusion with unknown benefit from oxytocin bolus.34 Carbetocin appears equivalent to oxytocin, with the exception of requiring fewer additional oxytocic agents.35 The cost effectiveness of carbetocin has yet to be established.

Trials assessing tranexamic acid as a preventative agent for PPH at elective CS have been undertaken and show a significant reduction in blood loss (100–200 mL), an EBL >1000 mL and the need for additional uterotonic agents.36 Data are currently lacking on potential serious adverse effects, such as VTE, which should be available before any change in routine practice.

Exteriorisation of the uterus and closure of the uterine incision

Exteriorisation of the uterus has not been associated with any significant differences in febrile complications, surgical time, blood loss, intraoperative nausea, vomiting or pain, compared with intra-abdominal repair, and so either approach is considered reasonable, depending on practitioner preference.37

The debate regarding single- or double-layer closure of the uterus continues. Short-term outcomes are comparable; however, operating time may be reduced with a single-layer closure.5,18 Evidence derived from case-control and cohort studies suggests that a double-layer closure reduces the future risk of uterine rupture,38 39  and there is some RCT evidence showing that a single-layer closure and a locked first layer are both associated with thinner residual myometrial thickness.

Presently, RCT data are insufficient to draw conclusions regarding single- versus double-layer closure and future uterine rupture risk. A single-layer closure is considered appropriate for a woman with no desire for future fertility,40 41 otherwise it would seem prudent to continue the practice of a double-layer, non-locking, closure.

Peritoneal closure

Non-closure of the peritoneum has been shown to have short-term benefits for women, including less postoperative fever, shorter operating times and reduced hospital stays.42 Closure of the visceral peritoneum is also associated with increased urinary frequency, urgency and stress incontinence in the short term.43 There are some long-term data that do suggest reduced intraabdominal adhesion formation when peritoneal closure is performed; however, the evidence regarding this remains limited and inconsistent at present.44 45 Currently, on balance, the evidence remains in favour of non-closure.46

Rectus sheath closure

There have been no RCTs to evaluate the optimal technique for closure of a transverse incision of the rectus sheath. Generally speaking, the sheath is closed with a continuous technique using a slowly absorbable suture, such as a 0 or 1 vicryl or polydioxanone (PDS) suture, with PDS maintaining greater tensile strength in vivo for longer than vicryl.

Subcutaneous tissue

Closure of subcutaneous tissue where the thickness is ≥2 cm reduces wound disruption (RR 0.66) and seroma formation (RR 0.42) and hence should be routine practice.47 There is no evidence to support closure when the thickness is ≥2 cm and the addition of a subcutaneous drain has not been shown to provide any additional benefit.48 49

Skin closure

Subcutaneous sutures, as compared with staple use, have been found to significantly reduce wound separation rates, with no significant differences in wound infection, haematoma, seroma or readmission rates, or cosmetic outcome.50 Suturing takes on average seven minutes longer.51 It should be noted that the higher wound separation rate was associated with staple removal on or before day four,52 so both techniques are likely equivalent with later removal of staples.