Elective caesarean delivery is the only true primary prevention strategy. Caesarean delivery after the onset of labour is not protective of injuries to the pelvic floor. Some functional loss to the pelvic floor might be avoided by elective caesarean section but not by emergency caesarean section [57, 69].

Elective cesarean section can certainly prevent mechanical trauma to the anal sphincter but not neurological trauma [70], but this has not been demonstrated for the urethral sphincter. The surgical and anaesthetic risks of a caesarean section and the risks associated with repeat caesarean sections for future pregnancies need to be considered in making an informed decision.

Studies have shown an increased risk of long-term urinary incontinence [71], and surgery for POP and/or SUI [37], following vaginal delivery; most epidemiological studies suggest that caesarean section provides only partial protection. This is for both emergency and elective caesarean section. Eight or nine caesarean sections need to be performed to avoid one case of urinary incontinence [71]. A 12 year large cohort study by Mcarthur et al has concluded that caesarean section is not protective of urinary incontinence unless all the women had only caesarean deliveries. Even after exclusive caesarean deliveries the prevalence of urinary incontinence was as high as 40 % [109].

With regards to POP, patients delivered exclusively by caesarean section have a significantly reduced risk of objectively measured POP 12 years after delivery [73] and a reduced risk of symptoms by 20 years [71]. Yet the lifetime risk of undergoing a single operation for POP and UI is estimated to be 11.1 % [74], which suggests that the development of PFD may be attributable to factors beyond vaginal versus caesarean deliveries. An intervention such as primary elective caesarean delivery for all births could potentially cause harm to a proportion of women who would otherwise have been delivered vaginally and not have experienced pelvic floor disorders.

Nevertheless, there may be a role for elective caesarean delivery in women with non-modifiable risk factors as part of an alternative primary prevention strategy.

Pregnancy and birth trauma are risk factors for urinary incontinence. The incidence of stress urinary incontinence during pregnancy has been reported in the range of 19.9–70 % in nulliparous women [75–77] while in the postpartum period after one vaginal delivery it ranges from 0.7 to 35 % [78, 79].

Antenatal pelvic floor muscle training has been shown to reduce the incidence of postnatal SUI in the short term [80–82]. However, a 6-year [83] and 8-year [84] follow-up showed no significant improvement. By 8 years there was no difference in the quality of life between the study and control group. This was mainly because of the poor compliance with PFMT. Only 38 % of the women performed pelvic floor muscle training (PFMT) twice per week or more. There was no difference in outcome between those who performed regularly compared with those who performed less frequently. However the National Institute of Clinical Excellence (NICE) [85] recommends PFMT for all women in a first pregnancy for prevention of SUI based on data from two randomized controlled trials (RCTs) [81, 82]. An RCT on antenatal pelvic floor exercises to prevent and treat urinary incontinence by Ko et al. [86] on 300 women, concluded that there was a significantly lower incidence of self reported urinary incontinence in the PFMT group than the control group. It also showed that women who delivered vaginally experienced more postpartum leakage than those who delivered by caesarean section.

Antenatal perineal massage has been proposed as a method of decreasing the incidence of perineal trauma. Daily antenatal perineal stretching massage by the woman or her partner has been found to prevent perineal trauma and perineal pain [87]. Antenatal perineal stretching massage has a protective effect in nulliparous women and is associated with an overall reduction in the incidence of trauma requiring suturing as well as lower incidence of episiotomy. These findings were significant for women without previous vaginal birth only. No differences were seen in the incidence of first- or second-degree perineal tears. No significant differences were observed in the incidence of instrumental deliveries, sexual satisfaction, or incontinence of urine, faeces or flatus for any women who practiced perineal massage compared with those who did not.

Five studies [88–92] evaluated the effectiveness of perineal massage. Two of those studies failed to analyse by intention to treat [88] and two of the studies failed to randomize [89]. Two large and similar trials and a pilot study [90] evaluated massage with sweet almond oil for 5–10 min daily from 34 weeks until delivery. The Labrecque and colleagues’ trial [91] reported that in women with first vaginal births, 24.3 % of those that had been randomized to massage had intact perineums compared with 15.1 % of controls who did not massage. A simultaneous trial with subsequent or second births showed no significant benefit of massage [91]. A similar trial by Shipman et al. [92] in nulliparous women found a 6.2 % increase of intact perineums in the massage group compared to the controls. The weighted risk difference from all those trials was – 0.08, i.e., one case of perineal trauma that required suturing would be avoided for every 13 women who did antenatal perineal massage [44].

Some accoucheurs ease the perineum back over the crowning head, whereas others believe that manual stretching increase the local tension and causes laceration. There are no randomised controlled trials to provide evidence for or against perineal stretching massage in the second stage. One study from Turkey on 396 nulliparous women who delivered vaginally, concluded that perineal massage during labour decreases the rate of episiotomy and lacerations [93]. Another randomized controlled trial by Stamp et al. [94] on 1340 women concluded that the practice of perineal massage in labour does not increase the likelihood of an intact perineum or reduce the risk of pain, dyspaerunia, or urinary and fecal problems. This trial was underpowered and did not assess the outcome of third degree tears. A Cochrane review showed there was a significant effect of warm compresses and perineal massage during the second stage of labour on reduction of perineal trauma and suturing [95] but suggested that further research is required to see if it prevents OASIS. The procedure has shown to be acceptable to women and midwives.

Randomised controlled trials have evaluated birthing position for the second stage of labour. Two studies [60, 61] assessed squatting or other unsupported upright position compared with recumbent positions for the second stage of labour. Seven randomised controlled trials [96–102] comparing upright and recumbent position, showed that women delivering in upright position were less likely to have an episiotomy, had more lacerations and required repair. A Cochrane review on position in the second stage of labour for women without epidural anaesthesia [103] stated that there was no difference in second degree perineal tears and fewer episiotomies were performed in the birthing stool or squatting position. The rates of episiotomy were also lower for women using birth chair, but second degree perineal tears were increased. However, when considering all women, there was a higher risk of second degree tears in the upright positions except when the birthing cushion was used. The current evidence on the effectiveness of various delivery positions in the prevention of perineal trauma still remains inconclusive. So, it is suggested that women should be encouraged to deliver in whichever position is most comfortable for them.

Whirlpool baths have a place mainly in low risk midwifery units, but there was only one trial of Jacuzzi whirlpool baths in labour albeit not for birth. This showed less perineal trauma in the group assigned to the Jacuzzi [104].

Perineal trauma after forceps and vacuum delivery has been compared in many RCTs [44, 105, 106]. Eason et al. [44] compared seven randomized controlled trials and showed that women delivered by forceps had more anal sphincter trauma than women delivered by vacuum extraction. Compared to vacuum extraction, forceps delivery was associated with almost twice the risk of developing faecal incontinence [105]. Vacuum delivery therefore causes significantly less maternal trauma compared to forceps. The weighted risk difference for anal sphincter trauma was −0.06 (95 % CI −0.10,−0.02). Obstetricians would need to deliver 18 women by vacuum than forceps to prevent one case of anal sphincter tear.

A recent trend has been towards hands off (not to support the perineum) approach for normal vaginal delivery during crowning. There is lack of evidence for this change of practice. In a recent survey of midwifery practice in England, 50 % of the midwives preferred a hands off/poised approach [107]. A recent Cochrane review has suggested no difference in OASIS rates between ‘hands on’ or ‘hands poised/off’ [95]. One of the studies (Decosta et al. [108]) was small and did not give any estimable effect. Mayerhofer et al. [109] reported only on third-degree tears. McCandlish et al. [4] reported on both third- and fourth-degree tears. There was considerable heterogeneity in the studies. The perineal techniques of the studies were different. The outcome measures were different. The terms “hands on,” “hands off,” “standard care” and “perineal support” meant different things across the studies and were not always defined sufficiently. In the McCandlish et al. study [4], “hands off” not only meant no hand on the perineum and infant’s head until the head was born, but also no manual assistance for the birth of the shoulders, while Mayerhofer et al. [109] defined “hands off” as no hands on the perineum or fetal head until the head was born, but made no distinction between “hands on” and “hands off” for the assistance of the birth of the shoulders. In the Albers’ study [110] “hands off” only meant no hands on the perineum until crowning of the head. Although the standard care or “hands on” manual support techniques are poorly described in most of the studies, it is clear that all studies implied a slow and controlled delivery of the head. Another randomized trial by Albers’ et al which compared hands off the perineum technique and warm compress did not show any advantage or disadvantage in reducing the obstetric genital trauma [111]. Evidence from Finland suggests that their lower OASIS rate (0.6 %) is a result of the more frequent use of perineal support and episiotomy, compared with other Nordic countries (OASIS rates 3.6–4.2 %) [112]. In Norway, recent implementation studies of the ‘hands on’ method have shown a 50 % reduction in OASIS rates [113, 114], supporting a return to the use of the traditional method of perineal support as a method of prevention [115].

During the second stage of labour, women may be encouraged to bear down throughout a contraction. A prolonged second stage, in which strong voluntary pushes are encouraged, has been implicated in denervation injury [69]. Parnell et al. [116] and Thompson [117] found no difference in perineal trauma between women who only pushed spontaneously and those who were directed to push throughout the contraction.

The role of episiotomy has already been discussed elsewhere. Whether episiotomy should be used as a form of prevention is still controversial. However, a quasi randomized study by Coats et al. [67] in nulliparous women showed that midline episiotomies had 12 % anal sphincter tears compared with a 2 % anal sphincter tears in mediolateral episiotomies. A prospective study of almost 300,000 vaginal deliveries reported that the selective use of mediolateral episiotomy did protect against damage to the anal sphincter complex [43].

Perineal hyaluronidase injection has been used during the second stage of labour for reducing perineal trauma. Four randomised controlled trials involving 599 woman were included in a Cochrane review [118]. Two trials compared hyaluronidase with placebo injection and three trials compared hyaluronidase with no intervention. In comparison to the group that received no intervention, the group, which received hyaluronidase during the second stage of labour, had a lower incidence of perineal trauma but no difference when compared with placebo injection. The role of perineal hyaluronidase during second stage is yet to be established.

EPI-NO® – a new birth and vaginal training device – was developed to reduce the number of episiotomies and increase the incidence of an intact perineum by training with the device. A German doctor, Wilhelm Horkel, designed it. It consists of an inflatable silicone balloon connected to a hand pump. EPI-NO® is designed to dilate the vagina with the aim of adaptation of vagina and perineum to the delivering fetus. Furthermore, women can train their pelvic floor muscles and are able to develop a feeling for pushing process during labour. First results of a German trial demonstrated not only a significant decrease of perineal trauma (42 %) and much lower episiotomy rates (33 %) but also a significant reduction of analgesics, patient anxiety of birth and shortening the duration of second stage of labour after training with EPI-NO® [119].

Similar experience was reported by Kok et al. [120] who found a significant reduction in episiotomies and a tendency towards lower rates of injured perineum (90 % vs. 96.6 %). Similar results with significantly higher rates of intact perineum and a lower rate of perineal tears were observed in an Australian trial [121]. Another prospective randomized multicentre trial concluded that training with EPI-NO® is safe for both mother and child, easy to use, helps to avoid unnecessary episiotomies and increases the likelihood of having an uninjured perineum [122]. Further improvement of these results by means of combining EPI-NO® and perineal massage should be evaluated. A randomized controlled pilot study of EPI-NO® on levator trauma by Shek et al. [123] showed a weak trend towards a lower incidence of levator avulsion and irreversible overdistension in women allocated to EPI-NO® group than in those who actually used the device. It also concluded that a larger sample study size was needed to determine the efficacy on EPI-NO® on levator muscle. Shek et al. [123] also performed a subgroup analysis in women who delivered by prelabour or first-stage caesarean section. No significant differences in peripartum changes of hiatal areas and pelvic organ descent between the control and EPI-NO® groups were found.