Despite all the evidence corroborating the selective use of episiotomy and although routine use of the procedure is contraindicated, there are no evidences corroborating if episiotomy is necessary in any circumstance. The present clinical randomized trial was performed to compare maternal and perinatal outcomes in women submitted to a non-episiotomy protocol versus one of selective episiotomy. After signing the consent form, 241 women were randomized to a non-episiotomy protocol (the experimental group) or to a selective episiotomy group (the control group). No episiotomies were to be performed in the experimental group except under exceptional circumstances. In the control group, selective episiotomies were to be performed in accordance with the healthcare professionals’ clinical judgement. Of the 241 women enrolled, four were excluded because of a post-randomization indication for a caesarean section. Of the remaining 237, 115 were assigned to a non-episiotomy protocol and 122 to selective episiotomy. There was no difference between the two groups with respect to maternal or perinatal outcomes. The episiotomy rate was similar with two episiotomies in each group (1.7%), as was the duration of the second stage of labor, the frequency of perineal tears, severe perineal trauma, need for perineal suturing and blood loss at delivery. In conclusion, a non-episiotomy protocol appears to be safe for mother and child, and highlights the need to investigate whether there is, in fact, any indication for this procedure.

Despite all the evidence corroborating the selective use of episiotomy and although routine use of the procedure is contraindicated, the actual indications for episiotomy in modern obstetric practice still remain to be clarified [1]. According to the American College of Obstetricians and Gynecologists (ACOG), “based on the existent evidence, there are no specific situation in which episiotomy is essential, and the decision to perform an episiotomy should be based on clinical considerations” [2]. The World Health Organization (WHO) recommends an episiotomy rate of 10% as “a good goal to pursue” [3], based on the previous results of a randomized clinical trial conducted in the United Kingdom and published in 1984 [4].

A Cochrane systematic review raises the question of what indeed are the real indications for episiotomy. Suggested answers have included premature delivery, breech presentation, fetal macrosomia, shoulder dystocia, instrumental delivery (forceps or vacuum extraction), non-reassuring fetal heart rate, and rigid perineum or imminent perineal tears [5]. It has been debated, however, whether these situations do indeed represent indications for episiotomy and clearly this question merits further investigation in randomized clinical trials. Although it is obvious that routine episiotomy should be avoided, there is no solid evidence corroborating any indication whatsoever for episiotomy. The benefits of the use of the procedure in selected situations remain controversial, including in instrumental deliveries [6]. In an article proposing that the routine episiotomy era has come to an end, Scott suggests that the natural forces of labor should be allowed to gradually distend the perineum and recalls the adage already quoted by Eason and Feldman 15 years ago: “Don’t just do something, sit there!” [7, 8]. Recently, some authors suggested that episiotomy should never be performed [1, 9]. The question, however, has not yet been adequately evaluated in randomized clinical trials.

A review of the Medline, Lilacs/SciELO, EMBASE and Scopus databases and of the Cochrane Library using the keywords “episiotomy” and “vaginal delivery” in Portuguese, English and Spanish, with the filters “randomized clinical trials” and “systematic review”, failed to reveal any randomized clinical trials on the subject of not performing episiotomy during vaginal delivery. All the trials published in the literature compare selective versus routine episiotomy and they were not designed to assess the need for episiotomy in any clinical setting [5]. This is an inherent limitation for its interpretation.

The present clinical trial was performed to compare maternal and perinatal outcomes in women submitted to a non-episiotomy protocol versus one of selective episiotomy.

This study was conducted at the Women’s Healthcare Centre in the Instituto de Medicina Integral Prof. Fernando Figueira (IMIP), Recife, Brazil, between July and September 2014. The protocol is registered at ClinicalTrials.gov under number NCT02178111 and was published in Reproductive Health [1].

This is an open, randomized clinical trial conducted with clinically stable women in active labor with a live, full-term foetus (37 to 41 weeks of pregnancy) in cephalic presentation (vertex position), and with dilatation of 6 to 8 cm. Exclusion criteria consisted of pregnancy bleeding disorders (premature detachment of the placenta); indication for a caesarean section; cephalopelvic disproportion, non-reassuring fetal heart rate, dystocia; women incapable of giving their consent and women with no responsible accompanying person. In addition, women who were submitted to a caesarean section after enrolment to the study were excluded (post-randomization exclusion).

Sample size was calculated using the Open Epi software program, version 2.3 (Atlanta, GA, USA). Predicting an episiotomy rate of 1% in the non-episiotomy group versus 10% in the selective episiotomy group [3, 4], for a power of 80% and a 95% confidence level, 200 women would need to be enrolled. This number was increased by 20% to compensate for any possible losses, resulting in an estimated sample size of 240 women. A list of random numbers was generated using the Random Allocation software program, version 1.0. To compensate for any post-randomization losses, the list consisted of 250 random numbers. Opaque, sealed, consecutively numbered envelopes containing each participant’s allocation were then prepared. Each allocation envelope was only opened during the second stage, defined as when dilatation is complete, the head of the foetus is fully engaged in the pelvis and there is a spontaneous urge to push. Allocation concealment was thus assured up to this moment.

For the experimental group, the healthcare professionals were instructed to base their management on the principle that episiotomy is unnecessary even in situations in which the literature suggests that it may confer some benefit. Therefore, no episiotomy was to be performed in this group except under exceptional circumstances, i.e. those in which clinical judgement would deem the procedure absolutely necessary. For the control group, the healthcare professionals should perform episiotomy selectively according to their clinical judgement, i.e. in line with the routine procedure at the institute.

The primary maternal outcomes were episiotomy rate; duration of second stage; frequency of spontaneous lacerations (degree of the laceration and the location: anterior or posterior); frequency of instrumental delivery; frequency of perineal trauma (any type: episiotomy or tearing); postpartum blood loss; need for perineal suturing, and the number of suture threads. The primary perinatal outcomes were 1^st and 5^th minute Apgar scores; need for neonatal resuscitation and umbilical cord blood pH. The secondary maternal outcomes evaluated were frequency of severe perineal trauma; complications with perineal suturing (edema, ecchymosis, hyperemia, secretion, hematoma, dehiscence, perineal pain and infection) identified during postnatal consultation; intensity of postpartum perineal pain assessed according to a visual pain scale [10, 11], and maternal satisfaction. The secondary perinatal outcomes were neonatal morbidity (neonatal asphyxia and the duration of neonatal hospitalization) and admission of the newborn infant to the neonatal ICU. In the cases in which episiotomy was performed, the indications for the procedure and the characteristics of the women submitted to it were also analyzed.

The women were accompanied during delivery according to a humanized childbirth care model based on evidence and on the WHO recommendations [3]. This model favors non-supine positions, non-pharmacological pain relief methods, continuous support during delivery and techniques for protecting the perineum, all in accordance with the woman’s preferences and the routine at the institute.

The duration of second stage of labor was recorded and blood loss was measured during the first hour postpartum. For this latter purpose, blood was collected using specially adapted plastic bags. All these bags as well as the sponges and gazes used were then weighted and the blood loss (expressed in milliliters -ml) was measured by discounting the previously determined dry weight. After delivery, perineal conditions were evaluated by visually inspecting the vulvar introitus and perineal region. Whenever present, lacerations were classified as first degree (involving the skin and/or vaginal mucosa), second degree (affecting the perineal muscle), third degree (affecting the anal sphincter muscle) and fourth degree (in addition to the sphincter, the rectal mucosa is also affected) [7]. Third and fourth degree tears were classified as severe perineal damage.

The umbilical cord blood pH was measured in a sample collected from the fetal end of the cord after it was cut, with pH values ≥7.2 being considered normal.

Perineal pain was evaluated between 24 and 48 h following delivery by asking the woman if she had any discomfort in the perineal region. A visual analogue scale was used to estimate pain intensity, assessed as a score that ranged from 0 to 10, with zero indicating a complete absence of pain and 10 reflecting the worst pain imaginable [10, 11]. This variable was later recoded as “any perineal pain” (scores of 1 to 10). Next, maternal satisfaction was evaluated and classified as very satisfied, satisfied, fairly satisfied, dissatisfied or very dissatisfied in accordance with the woman’s selection from a range of faces in a faces scale [9].

The presence of complications with perineal suturing (edema, ecchymosis, hyperemia, secretion, hematoma, dehiscence, perineal pain and infection) was evaluated during hospitalization (until 48 h of delivery) and at the postpartum appointment scheduled 10 days after of vaginal delivery. Women were told they could return to the hospital before that date to the hospital if they felt it was necessary for suspecting any complications. At these times all the women had the perineal region routinely examined by a health professional (certified nurse-midwife or an obstetrician doctor) blinded to the group assignement.

A statistician blinded to the group assignment conducted the data analysis using the public domain Epi Info statistical software program (Atlanta, GA), version 7.2. The intention-to-treat principle was applied, i.e. each woman was analysed in the group to which she had originally been allocated. Student’s t-test or the Mann–Whitney test was used, as appropriate, for the numerical variables. The Pearson’s chi-square test or Fisher’s exact test (when one of the expected values was <5) was used for the categorical variables. A significance level of 5% was adopted. All p-values were two-tailed. Risk ratios were calculated as measures of relative risk (RR), together with their 95% confidence intervals (95% CI). A standard value of 1.0 was attributed to the reference category.

An overall episiotomy rate of less than 2% was found in the two groups evaluated in the present study, showing that it is possible to implement a non-episiotomy protocol for a group of women with full-term pregnancies and foetuses in the cephalic position without any statistically significant differences in relation to the selective practice of the procedure. This is the first randomized clinical trial comparing the intention to never perform episiotomy to the current recommended practice of selective episiotomy. Further studies should be conducted, particularly in institutions and regions in which episiotomy rates remain high (although under the concept of “restrictive” use), in an attempt to determine whether there really is any indication for episiotomy in modern obstetrics. Until the results of these studies are available, much lower episiotomy rates should be achieved in this country and in other countries, and the end of episiotomy is a goal that should be pursued within a humanized childbirth care model.