Episiotomy restricted to foetal indications and occurrence of severe perineal tears: systematic review and meta-analysis of randomised clinical trials

Introduction

Episiotomy, a surgical incision made in the perineum during the second stage of labour, is traditionally performed to facilitate childbirth, particularly in cases of foetal distress or shoulder dystocia, and to prevent severe perineal tears that are challenging to repair.1 However, consistent evidence against its routine use has led to a shift toward selective practice, as recommended by most contemporary guidelines.2 Leading organisations, including the American College of Obstetricians and Gynecologists, advise against routine episiotomy, but have not set a specific target rate for the practice.3 Some authors question whether episiotomy should be practised at all.4

Despite the general proscription of routine practice, episiotomy rates vary widely among countries and obstetric centres. The 2023 Leapfrog Group report on Maternity Care showed episiotomy rates in the USA of 4.6% in the prior year, down from 5.2% in 2021 and 2020 and 12.5% in 2012.5

Perineal lacerations, a negative outcome that episiotomy aims to prevent, can be anterior or posterior. Anterior lacerations can occur in the periclitoral and periurethral areas but are generally not associated with high maternal morbidity. Posterior lacerations are associated with more morbidity and are classified as:6

• First degree: involving the skin and subcutaneous tissue of the perineal and vaginal epithelium.

• Second degree: extending to the perineal body muscles, sparing the anal sphincter.

• Third degree: affecting part of the anal sphincter fibres.

• Fourth degree: including the anal sphincter complex and the anorectal mucosa.

Third- and fourth-degree lacerations, also known as obstetric anal sphincter injuries (OASIs), can result in suture dehiscence, perineal pain, loss of sphincter control, dyspareunia and urinary incontinence.7 Risk factors include episiotomy, instrumental vaginal delivery, macrosomia, prolonged second stage of labour, occiput posterior presentation, advanced maternal age and epidural analgesia.8–12 Studies show that midline episiotomy increases the risk for OASIs, while the role of mediolateral incision is still debated.13 14 Some studies suggest it may reduce the risk of OASIs in nulliparas and in instrumental deliveries with forceps.15 16 However, there are still conflicting data, and evidence quality and study designs vary.17

Furthermore, some perineal injuries, such as levator ani muscle avulsion, may be occult. In these cases, there is no visible perineal lesion, but digital rectal examination and imaging may reveal the injury.18 Although up to half of the cases may heal spontaneously, this type of lesion is associated with pelvic organ prolapse, anal incontinence and other pelvic floor disorders.19–21 Risk factors include instrumental delivery, prolonged second stage of labour and a non-occiput anterior foetal position. Additionally, while some studies correlate the occurrence of levator ani avulsion with episiotomy, others suggest that the practice has no significant impact and may help prevent such injuries.22–24

Episiotomies may be considered second-degree lacerations, since the incision involves the perineal muscle. Potential consequences of the practice include increased blood loss, including postpartum haemorrhage, a higher risk of severe tears in subsequent deliveries, dyspareunia and a negative birth experience.25–31 Moreover, the practice may increase healthcare costs.31 Conversely, non-episiotomy may be associated with higher rates of anterior and first- and second-degree perineal tears.16 Considering the potential negative effects of the practice of episiotomy, more precise indications are of value and should be established.

Given the conflicting guidelines and literature on the impact of episiotomy and target rates, this study aims to address the existing knowledge gap. While meta-analyses evaluating different episiotomy protocols exist, none of them aim to compare selective protocols to a protocol of episiotomy specifically restricted to foetal indications.2 32 33 Therefore, this study aims to assess the occurrence of severe perineal tears when an episiotomy protocol restricted to foetal indications is implemented, in comparison to the existing protocol where episiotomies are performed selectively for both foetal indications and perineal tear prevention. Additionally, the study seeks to evaluate whether this protocol effectively reduces episiotomy rates while ensuring the safety of newborns.

Methods

Results

The initial search identified 771 publications, from which 320 duplicate articles were removed. After title/abstract screening, four publications were considered potentially relevant. On full-text review, two publications were excluded for not meeting the inclusion criteria.39 40 Amorim et al compared a restrictive episiotomy protocol with selective protocols but did not specify that the restrictive protocol should be limited to foetal indications.39 Sleep et al met the inclusion criteria for both intervention and control groups; however, the comparative analysis included both primiparous and multiparous individuals, and it was not possible to obtain data exclusively for patients in their first vaginal delivery from the authors. Additionally, Sleep et al assessed severe perineal tears as an outcome but did not use the classification of degrees 1, 2, 3 and 4 as required by our study criteria.40

Consequently, two RCTs were included: Dannecker et al, with a sample size of 109 participants, and Sagi-Dain et al, with a sample size of 676 participants, comprising a total of 785 participants.41 42 The key characteristics of the included articles are presented in table 1. The PRISMA flow diagram provides a visual representation of the study selection process (figure 1).

Dannecker et al were assessed as having an overall high risk of bias, whereas Sagi-Dain et al presented some concerns for bias overall.41 42 Both studies demonstrated a low risk of bias concerning the randomisation process, missing outcome data and selective reporting of results. However, Dannecker et al was rated as high risk overall due to the lack of blinding among personnel and the failure to report using an appropriate method to estimate the effect of adherence to the intervention.41 In contrast, while Sagi-Dain et al also lacked personnel blinding, the study mitigated this limitation by presenting both intention-to-treat and per-protocol analyses, justifying a classification of ’some concerns' for bias in this domain.42 Additionally, both studies raised some concerns regarding outcome measurement; despite providing explicit criteria for classifying perineal tears, the outcome assessors were not blinded (figure 2).

Funnel and radial plots were generated for the outcome ‘severe perineal tears’, the primary outcome assessed in the present study. Although only two studies were included, the funnel plot appeared symmetrical, with both falling within the 95% confidence contour. Additionally, the radial plot showed both studies closely aligned with the regression line, suggesting consistency of effects (online supplemental figures 1 and 2).

The Selective Episiotomy Protocol and the Restrictive Episiotomy Protocol groups did not vary in terms of mean maternal age, birth interventions and neonatal birth weight (table 2). Although mean gestational age showed a statistically significant difference between the groups, the birth weight did not. Demographic characteristics are summarised in table 2.

The incidence of severe perineal tears was lower in the Restrictive Episiotomy Protocol group, though the difference did not reach statistical significance (pooled RR 0.50, 95% CI 0.21 to 1.21, I² = 0%; table 2 and figure 2). Fourth-degree tears did not occur in either group, so the assessment for severe perineal tears included only third-degree tears.

The occurrence of episiotomies was significantly lower in the Restrictive Episiotomy Protocol group compared with the Selective Episiotomy Protocol group (pooled RR 0.61, 95% CI 0.49 to 0.76, I²=0%; table 2 and figure 3). No difference was found for either minor perineal tear rates or intact perineum rates (pooled RR 1.07, 95% CI 0.95 to 1.20, I²=86% and pooled RR 1.01, 95% CI 0.83 to 1.24, I²=82% table 2 and figure 3).

Considering neonatal outcomes, no statistical difference was noticed between the Restrictive Episiotomy Protocol and the Selective Episiotomy Protocol groups. No significant difference was found in our meta-analyses of the rate of Apgar score <7 at the first minute (pooled RR 0.80, 95% CI 0.45 to 1.44, I²=0%, table 2), the rate of Apgar score <7 at 5 min (1.01, 95% CI 0.06 to 16.02, table 2) or the rate of low umbilical cord pH (pooled RR 1.20, 95% CI 0.78 to 1.85, I²=38%, table 2). Maternal and neonatal outcomes are summarised in table 2.

Although not the primary objective of our study, we observed that both studies assessed maternal postpartum pain and found no statistically significant differences in this outcome.41 42 Both studies also included follow-up analyses. Sagi-Dain et al reported results at a 2-month follow-up within the same article, while Dannecker et al provided data from a follow-up with a mean time of 7.3 months, presented in a separate publication.43 Both studies evaluated perineal pain, dyspareunia, and urinary and anorectal incontinence, and no statistically significant differences were identified.42 43 Furthermore, Sagi-Dain et al reported no significant differences in perineal recovery, perineal infection or the time to resumption of sexual activity.42 Dannecker et al also examined urethral pressure profiles, anal manometric findings and pelvic floor contraction strength (by evaluating levator ani contraction), similarly finding no significant differences between groups. 43 GRADE assessment revealed moderate certainty for episiotomy occurrence, low certainty for severe perineal tears occurrence, 1 min APGAR <7 and very low certainty for minor tears occurrence and 5 min APGAR <7.

Discussion

In this systematic review and meta-analysis, the episiotomy protocol restricted to foetal indications (Restrictive Episiotomy Protocol group) was not associated with worse outcomes (severe perineal tears, episiotomy occurrence, minor perineal tears, intact perineum and neonatal outcomes) when compared with the protocol for episiotomy following foetal indications (Selective Episiotomy Protocol group).

Our findings suggest a potential trend toward fewer severe perineal tears when episiotomies are performed under a restrictive protocol compared with a selective approach. Although this result was not statistically significant and the level of certainty was low, severe perineal tears were 51.8% less frequent in the Restrictive Episiotomy Protocol group. Therefore, while current evidence is insufficient to draw definitive conclusions about the protective effect of restrictive episiotomy protocols on perineal outcomes, these findings indicate that the Selective Episiotomy Protocol is not associated with a significantly reduced incidence of severe perineal tears, contrary to previous suggestions.3 4

Moreover, our study reveals that protocols that suggest episiotomy restricted to foetal indications successfully reduce rates of episiotomy, with moderate level of certainty. The statistically significant difference between protocols suggests adherence to the suggested protocol by the attending obstetric personnel, but it is important to acknowledge the ongoing challenge of shifting the entrenched beliefs and practices of healthcare professionals regarding episiotomy. Sagi-Dain et al stated that they faced problems in convincing the obstetric personnel to avoid episiotomy performance, because of a belief that avoiding episiotomy would lead to severe perineal tears.42 Even though a fully successful shift toward the restricted protocol has not been achieved, the fact that rates of episiotomy were significantly lower in the restrictive group suggests feasibility in clinical practice. Nevertheless, episiotomy rates remained relatively high, although services with lower episiotomy rates, as low as 1.7%, have shown safe outcomes.39

Minor perineal lacerations were more frequent in the restrictive episiotomy protocol group, but these differences were not statistically significant, heterogeneity was high and the level of certainty was very low. It is worth noting that episiotomy itself may be classified as a second-degree laceration since it involves a tear through the perineal muscles. Additionally, a study found no difference in sexual function in parturients who suffered second-degree tears when compared with episiotomy.30 Conversely, intact perineum rates were higher in the restrictive group, though this difference was also not statistically significant. Additionally, the authors did not explicitly state whether assessments for levator ani muscle avulsions were conducted when classifying an intact perineum. However, the follow-up study by Dannecker et al assessed pelvic floor contraction strength by evaluating levator ani muscle contractions and reported no significant differences between the groups.43

These findings suggest that a restrictive episiotomy protocol may be associated with better perineal outcomes, but the small sample size and variability between studies limit the strength of this conclusion. However, considering that episiotomy is associated with risks such as infection and has negative impacts on people’s mental health and self-esteem, reducing episiotomy rates has inherent benefits. A longitudinal cohort study of 202 participants found that the birth experience of individuals with severe perineal lacerations or episiotomy was similarly negative compared with those with minor tears.28 Additionally, a systematic review and meta-analysis of 18 studies, including 3133 participants, found that perineal pain and dyspareunia occur after spontaneous vaginal birth regardless of trauma or episiotomy, but rates are lower with an intact perineum. Episiotomy had the highest incidence of perineal pain, with significant dyspareunia persisting at 12 months postpartum.29 Moreover, restrictive episiotomy protocols may reduce healthcare costs.34

The risk of bias assessment for perineal tears identified a high risk of bias in Dannecker et al and some concerns for bias in Sagi-Dain et al.41 42 This bias stems from the difficulty in blinding physicians performing and assessing episiotomies, as they would inevitably know whether an episiotomy was performed. This awareness introduces potential bias, as many clinicians still believe avoiding episiotomy could result in more severe tears, possibly leading to overclassification of tears as severe.42 These biases are particularly challenging to assess and address. Furthermore, the funnel and radial plots did not suggest evidence of publication bias or small-study effects. However, the limited number of included studies may constrain the interpretability of these findings.

Regarding neonatal safety, the fact that there were no statistically significant differences between groups suggests that there is no increased neonatal risk in pursuing a restrictive episiotomy protocol in comparison with a selective episiotomy protocol. It is important to point out that, although this is not the scope of the present study, there are no meta-analyses of RCTs suggesting that episiotomies should be performed at all to prevent neonatal complications. Further studies comparing protocols restricted to foetal indications to protocols of no episiotomy for any reasons would be of interest to address this matter.

The present study has several limitations. Most notably, only two RCTs met the strict inclusion criteria, substantially limiting the sample size and reducing the overall statistical power of the analysis. The lack of access to individual participant data precluded adjusted risk ratio calculations for relevant confounders such as gestational age, foetal presentation and birth weight. Additionally, low-frequency outcomes—such as low umbilical cord pH—could not be evaluated with sufficient statistical precision. The absence of standardised protocols across studies further contributes to variability in outcome assessment. The limited number of included trials also restricted the possibility of conducting subgroup or sensitivity analyses. Moreover, while a 51.8% relative reduction in severe perineal tears was observed in the restrictive episiotomy group, this finding did not reach statistical significance, possibly due to underpowering. One of the included trials was rated as having a high overall risk of bias, raising concerns about internal validity. Furthermore, the lack of blinding of clinicians and outcome assessors may have introduced classification bias, particularly for subjective outcomes such as perineal tear severity. Finally, important outcomes like levator ani muscle avulsion were not systematically assessed, limiting a more comprehensive evaluation of pelvic floor trauma.

Nonetheless, our results reinforce the trend towards increasingly restricting episiotomy practices, which were once routine but are now performed selectively in line with recent evidence. A Cochrane review from 2017 found a risk ratio for severe perineal tears of 0.7 (95% CI 0.52 to 0.94) when comparing selective episiotomy to routine episiotomy.2 More recently, Pereira et al, comparing selective protocols with more restrictive protocols (but not specifically restricted to foetal indications), found no statistically significant difference between these approaches.33

Therefore, considering maternal outcomes and cost-effectiveness, restrictive episiotomy protocols, which may be associated with lower episiotomy rates without increasing severe perineal tear incidence or compromising foetal health, are beneficial. New studies with larger sample sizes are necessary to confirm the findings of this study, particularly regarding the association between episiotomy protocols and severe perineal laceration rates.

In conclusion, this study reveals that restricting episiotomy practice to foetal indications may reduce episiotomy rates without compromising perineal outcomes or neonatal safety. Additionally, although our study revealed that there might be an association between better perineal outcomes and more restrictive episiotomy protocols, there was no statistical significance to this result; therefore, evidence is not yet definitive. Hence, further investigation is required and future research should aim to provide stronger evidence for severe perineal tear prevention and more precise episiotomy indications.