Attitudes and knowledge regarding episiotomy use and technique in vacuum extraction: A web-based survey among doctors in Sweden

Objectives: Correct episiotomy use and technique may prevent obstetric anal sphincter injuries. We aimed to explore the attitudes, use, and technique regarding episiotomy among doctors in Sweden, and their willingness to contribute to a randomized controlled trial of lateral episiotomy or no episiotomy in vacuum extraction in nulliparous women. Study design: A web-based survey was sent to members of the Swedish Society of Obstetrics and Gynecology (n = 2140). The survey included 31 questions addressing personal characteristics, use of epi-siotomy, a two-dimensional picture on which the respondents drew an episiotomy, and questions regarding attitudes towards episiotomy and participation in a randomized controlled trial. We calculated the proportion of supposedly protective episiotomies (fulﬁlling criteria of a lateral or mediolateral epi-siotomy and a length (cid:1) 30 mm). We compared the results between obstetricians, gynecologists, and residents using Chi-square and Kruskal-Wallis tests for differences between groups, and logistic regression to estimate the odds ratio (OR) of drawing a protective episiotomy. Results: We received 432 responses. Doctors without a vacuum delivery in the past year were excluded, leaving 384 respondents for further analyses. In all, 222 (57.8%) doctors reported use of episiotomy in<50% of vacuum extractions. We obtained 308 illustrated episiotomies with a median angle of 53 (cid:1) , incision point distance from the midline of 21 mm, and length of 36 mm, corresponding to a lateral epi-siotomy. Few doctors combined these parameters correctly resulting in 167 (54.2%) incorrectly drawn episiotomies. Residents drew shorter episiotomies than obstetricians and gynecologists. Doctors ranked episiotomy the least important intervention to prevent obstetric anal sphincter injuries in vacuum extraction. Doctors contributing to an ongoing randomized controlled trial of lateral episiotomy or no episiotomy in vacuum extraction were more able to draw a protective episiotomy (OR 3.69, 95% conﬁ-dence interval 1.94–7.02). Conclusions: Doctors in Sweden reported restrictive use of episiotomy in vacuum extraction and depicted lateral type episiotomies, although the majority were incorrectly drawn. Preventive episiotomy was ranked of low importance. Our results imply a need for education, training, and guidelines to increase uptake of correct episiotomy technique, which could result in improved prevention of obstetric anal sphincter injuries. (cid:3) 2021 The Authors. Published by Elsevier B.V. ThisisanopenaccessarticleundertheCCBYlicense(http://


Introduction
Episiotomy should be used selectively and restrictively [1][2][3].The clinical application of this wording is unclear, although operative vaginal delivery could be such a selected situation [3,4].A rate below 60% is proposed to be selective in a Cochrane review of selective versus routine episiotomy in vaginal birth [5].In this review, routine episiotomy [6z0-100%) was not protective against obstetric anal sphincter injuries (OASIS) in spontaneous vaginal delivery [5].Further research is needed to clarify if routine episiotomy is useful in women with operative vaginal delivery [5].In observational studies, a lateral or mediolateral episiotomy may prevent OASIS in vacuum extraction (VE) in nulliparous women [6,7].There may be a correlation between episiotomy rate and preventive effect, as seen in a meta-analysis and the EURO-PERISTAT project [6,8].The ambivalence towards episiotomy is reflected in the episiotomy rates ranging from 17.1% to 97.2% in operative vaginal deliveries in Europe [8].
Compared to most countries, the use of episiotomy in VE is low in Denmark, Iceland, and Sweden (17.1%-47.9%)where the OASIS rate is high (12.5-15.5%)[8] A low use of episiotomy may lead to lack of skills regarding the correct episiotomy technique to prevent OASIS [6,8].The trigonometric properties, angle, incision point, and length of an episiotomy, may all be of importance [9][10][11].An angle of approximately 60°to the midline is preferable [7,9,10,12,13].The incision point's importance is more debated and defines the main difference between a lateral and a mediolateral episiotomy [14].The preferred episiotomy length is not established [4,15], but an increasing length could decrease the risk of OASIS [9] Consensus on different types of episiotomies has been published by Kalis et al and include definitions of lateral, mediolateral, and midline episiotomies [15].
How doctors perform episiotomies has been reported from countries with a frequent use of episiotomy [14,[16][17][18].To our knowledge, episiotomy techniques have not been explored in a country with a low episiotomy rate.The aim of our study was to explore the use, technique, and attitudes regarding episiotomy among doctors in obstetric care in Sweden.We also wanted to investigate the willingness to contribute to a randomized controlled trial (RCT) of lateral episiotomy or no episiotomy in VE in nulliparous women.

Material and methods
We conducted a web-based survey among members of the Swedish Society of Obstetrics and Gynecology.An invitation was sent to members with a registered e-mail address in 2019 (n = 2140).Answers were considered a consent to participate and handled anonymously.One reminder was sent.Participants were instructed to respond on a computer to be able to draw an episiotomy with the computer marker.This survey focused on doctors who in Sweden make the decision to perform an episiotomy at VE.
The survey consisted of 25 closed answer questions and extra space in the end for free text.The questions addressed the characteristics of the respondent.Clinical profile was self-defined as obstetrician, gynecologist, resident/unspecified (such as working in both fields which is common in smaller hospitals).Other characteristics were gender (female, male, non-binary), age (<30, 30-39, 40-49, 50-59, 60 years, where age < 30 and 30-39 were collapsed into one group due to the low prevalence < 30 years; n = 11), years in practice within obstetrics and gynecology including residency (0-5, 6-9, 11-15, 16-20, >20), role at VE and forceps delivery (supervising, independent, trainee, do not perform), and number of VE and forceps deliveries the past 12 months (0, 1-5, 6-10, 11-20, 21-30, >30).The respondents were asked to estimate the proportion of VE in which they would perform episiotomy (0-100%), list perceived indications for episiotomy, and rank the estimated importance of different preventive measures against OASIS (1 = most important to 5 = least important) in VE in nulliparous women.Subsequently, they were asked to draw the episiotomy they would perform in a clinical situation on a two-dimensional schematic picture of a crowning fetal head with a vacuum cup attached, using the marker on a computer screen (Fig. 1).They were asked to name the technique (lateral, mediolateral or midline episiotomy).Two questions addressed if the respondent would consider participating in an RCT of lateral episiotomy or no episiotomy in VE in nulliparous women, as a doctor or a trial participant, and what an acceptable number needed to treat (NNT) with episiotomy would be to avoid one OASIS [19].
The drawn episiotomies were translated into coordinates in a diagram with the grid origin at the posterior fourchette.The epi-siotomy angle was calculated as compared to the midline.The length was calculated in millimeters with the distance between the posterior fourchette and anus as reference (40 mm).The incision point was calculated as the straight distance in millimeters from the origin at the posterior fourchette to the incision point.The episiotomies were categorized as lateral, mediolateral, midline or unclassifiable.Lateral episiotomies had an incision point 10 mm from the posterior fourchette with an angle of 45-80°to the midline [14,15].Mediolateral episiotomies had an incision point < 10 mm from the posterior fourchette with an angle of 50-80°to the midline [14,15].Midline episiotomies had an incision point < 10 mm from the posterior fourchette with an angle < 30°to the midline [14].All other episiotomies were considered unclassifiable.Episiotomies with the above defined properties of a lateral or mediolateral episiotomy and a length 30 mm, were considered protective.Other episiotomies were considered nonprotective.

Statistical analyses
Analyses were performed using SPSS (IBM SPSS Statistics for Macintosh, Version 26.0, Armonk, NY: IBM Corp).Descriptive statistics are presented for each clinical profile (obstetrician, gynecologist, or resident/unspecified profile), with Chi 2 tests for differences in proportions between clinical profiles and Independent Samples Kruskal-Wallis test for differences in median angle, distance, and length between clinical profiles.The odds of drawing a protective episiotomy were calculated using univariate logistic regression for different doctor characteristics: clinical profile, gender, age, years of practice, role at VE, number of VE past 12 months, opinion about episiotomy effect on OASIS at VE, rate of episiotomy in VE, and willingness to participate in an RCT.Covariates were assessed in a directed acyclic graph (Fig. 2).The effect of clinical profile was further analyzed using a multivariate backward stepwise conditional logistic regression to identify a relevant model and reduce the risk of overadjustment.The ranking of preventive measures against OASIS at VE were dichotomized into most and second most important (rank 1-2) versus other ranks (rank 3-5).We calculated the proportion of doctors who considered each mea-

Ethical approval
The study was approved on 12 September 2018 by the Regional Ethical Review Board of Stockholm (2018/2291-32).

Results
We sent 2140 surveys of which 409 bounced due to unknown email address and 432 were completed (response rate 25.0%) (Fig. 3).We excluded 48 respondents who reported no performed VE during the past year and thus considered not actively working in the labor ward.It resulted in 384 included respondents (Fig. 3).
The obstetricians, gynecologists, and residents differed in most characteristics, although the reported use of episiotomy in VE was similar (Table 1).In total, 222 (57.8%) of 384 doctors reported that they use episiotomy in<50% of VE and 77 (20.1%) reported using episiotomy in 75% of VE (Table 1).
In total, 312 respondents drew their preferred episiotomy at VE on the attached picture.Four were deemed invalid and excluded, resulting in 308 valid pictures (Fig. 4).The median angle was 53°, the median incision point distance from the midline 21 mm, and the median length 36 mm, corresponding to a lateral episiotomy (Table 2).Residents drew shorter episiotomies than obstetricians and gynecologists (Table 2).The ranges were wide for all three episiotomy trigonometric properties across all clinical profiles (Table 2).In total, 176 (57.1%) pictures illustrated a lateral episiotomy, although 84 (47.7%) of these were called a mediolateral episiotomy (Table 3).
Of the 308 drawn episiotomies, 167 (54.2%) did not fulfill the criteria of a protective episiotomy (Table 4).Willingness to participate, or already participating in, an RCT assessing lateral episiotomy vs. no episiotomy in VE in nulliparous women showed threefold odds to draw a protective episiotomy (Table 4).Few doctors rated correct episiotomy as an important measure to prevent OASIS in VE.Most doctors rated slowing down the birth of the head (''slow birth") as an important measure to prevent OASIS in VE (Fig. 5).
In their role as a doctor, 240 (62.5%) of 384 were positive to, or already involved in, an RCT assessing lateral episiotomy vs. no episiotomy in VE in nulliparous women (Table 5).The greatest fear regarding the RCT was causing unnecessary harm to the perineum: either an unnecessary episiotomy or an unnecessary OASIS.Most doctors stated that up to 14 episiotomies to prevent one OASIS was acceptable (Table 5).

Discussion
In this survey among doctors in obstetric care in Sweden, we found that most doctors depicted a lateral type of episiotomy, but it did not fulfill the criteria of a protective episiotomy.Most doctors believed that episiotomy decreased the risk of OASIS at VE in nulliparous women, but that other measures to prevent OASIS were more important.Episiotomy was reportedly used restrictively.Most doctors were willing to contribute or were already contributing to an RCT regarding the effect of lateral episiotomy in VE in nulliparous women.Still, they required a low NNT to consider routine episiotomy.

Interpretation of results
The doctors who were willing to participate, or already participating, in the ongoing RCT were three times more likely to draw a protective episiotomy.Doctors   have had educational workshops on episiotomy technique prior to the initiation of the RCT, which may have improved their technique [17,20].An interest in the RCT could also be associated with a better knowledge regarding episiotomy technique.The self-reported number of VE or rate of episiotomy did not consequently affect the correctness of drawn episiotomies.Nevertheless, residents drew shorter episiotomies as previously reported [21].
The doctors in our study believed that a correct episiotomy was preventive against OASIS but that it was the least important preventive measure.Most doctors reported a low individual episiotomy rate, reflecting the restrictive attitude toward episiotomy.Since junior doctors largely adopt practices from senior colleagues, learning how and when to use episiotomy in a restrictive setting could be problematic if an incorrect technique is customary [22,23].Moreover, if other preventive measures like correct perineal protection or the Couder maneuver are not well known, this is also negative for the prevalence of OASIS in VE [24,25].
In a meta-analysis, 18.3 episiotomies were required to prevent one OASIS at VE in nulliparous women and fewer if episiotomy was performed in at least 75% of VE [6].More recently, we reported that 27 episiotomies were required to prevent one OASIS at 31.5% episiotomy rate [26].Most of the doctors in our survey would only accept a lower number needed to treat.We hypothesize that the reluctance to perform episiotomy is due to a perceived lack of predictable effect and concern for long-term consequences.This may lead to a negative circle of poor episiotomy technique, modest preventive effect, disbelief in the intervention, and consequently decreased use and skills.

Strengths and limitations
To the best of our knowledge, this is the first study of Swedish doctors' perception of use, technique, and attitudes towards episiotomy in VE.The main difference from previous studies is the low episiotomy use setting [14,[16][17][18].We also assessed selfestimation of episiotomy use and attitudes towards episiotomy, which add perspectives to episiotomy practices in Europe.The number and mix of responding doctors correspond well to the true around-the-clock staffing in labor wards [27].It allowed us to compare practices between doctors with different levels of experience, while previous studies included mostly consultants with a mean of 14-18 years' clinical experience [14,18], or had few participants [16,17].The main limitation is the low response rate entailing a possible response bias.Non-responders could represent doctors with no current practice in obstetrics, or doctors with no opinion or another opinion than in our results.To increase validity, we only included doctors with experience from VE in the past year.It is possible that a higher proportion of doctors that favor episiotomy participated in the survey.The self-reported use of episiotomy in VE (43%) exceeds the registered national average (32%) [28].It is also possible that more doctors who are positive toward the ongoing RCT participated in the survey.Then again, the ongoing RCT is conducted in eight hospitals covering 40% of all births in Sweden, while doctors reporting that they already participate in the RCT constituted 25% of respondents in the survey.
Other limitations were that our survey could not be based on a validated instrument but was based on a previously published survey [29].Moreover, the drawing of an episiotomy on a computer screen is dependent on a precise conversion of picture coordinates into measurements in millimeters and degrees.Using selfillustrations is an established method to assess episiotomy technique [14,16,17], while actually cutting an episiotomy on a bespoke model would be closer to real-life [30].The drawn picture may not exactly mirror what doctors would do in clinical practice.Finally, our definition of a protective episiotomy length of 30 mm was deducted from Stedenfeldt et al who have shown a correlation between length and protective effect, as well as stud-ies on the Episcissors-60 Ò with a 50 mm blade [9,31,32].We are aware that other definitions of protective length may prevail and could change the results.

Conclusion
Most doctors in this survey would perform a lateral episiotomy.More than half of the doctors drew on a computer screen an incorrect, non-protective episiotomy due to variation in angle, incision point, and length.Most doctors believed that episiotomy was preventive of OASIS, but the least important measure, and would not use episiotomy routinely.Most doctors would consider participating or were already participating in an RCT assessing lateral episiotomy vs. no episiotomy in VE.These doctors were also more likely to draw a protective episiotomy.Our results suggest a negative circle of low use, poor technique, and modest effect, as well as the usefulness of education and training.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.cology for sharing email addresses.We thank ImproveIT for development of the electronical picture from the model image by Drs Fodstad, Staff, and Laine [14], and the distribution of surveys.We also thank Dr Sylvan for language editing and translation of the survey.
Contents lists available at ScienceDirect European Journal of Obstetrics & Gynecology and Reproductive Biology j o u r n a l h o m e p a g e : w w w .e l s e v i e r .c o m / l o c a t e / e u r o

Fig. 1 .
Fig. 1.Two-dimensional schematic picture of a crowning fetal head with a vacuum cup attached.

Fig. 2 .Fig. 3 .
Fig. 2. Directed acyclic graph of the effect of clinical profile on protective episiotomy.

Fig. 4 .
Fig. 4. Two-dimensional schematic picture of a crowning fetal head with a vacuum cup attached with depicted episiotomies.

5 .
Proportion of Swedish doctors ranking measures as important to prevent OASIS at vacuum extraction in nulliparous women.
at hospitals participating in the RCT

Table 1
Swedish doctors' self-reported characteristics and use of vacuum extraction, forceps, and episiotomy.

Table 2
Episiotomy trigonometric properties in illustrations of preferred technique among Swedish doctors.
Angle (to the midline), Distance (from posterior fourchette to start of incision) and Length (from start to end of incision) are presented as continuous variables with median and range, and as number of doctors with column proportions.a132 of 151 obstetricians drew a valid picture, 1 invalid.b 81 of 99 gynecologists drew a valid picture, 2 invalid.c 95 of 134 residents/unspecified drew a valid picture, 1 invalid.

Table 3
Swedish doctors' definition of illustrated episiotomy vs. consensus type of episiotomy.

Table 4
Non-protective and protective episiotomy and associated self-reported characteristics among Swedish doctors.Model based on backward stepwise conditional logistic regression including gender, number of VE past 12 months, rate of episiotomy in VE, and willingness to participate in RCT.
a Doctors with a valid picture and at least one VE the past 12 months (n = 308).b Row percentages.c

Table 5
Swedish doctors' perceptions about a randomized controlled trial of lateral episiotomy or no episiotomy in vacuum extraction in nulliparous women.Column percentage.NNT = numbers needed to treat, RCT = randomized controlled trial, OASIS = obstetric anal sphincter injury, epi = episiotomy. a Helena Kopp Kallner is associate professor at Karolinska Institutet in Sweden, senior consultant in Ob/Gyn and head of the clinical research department and clinical trials unit at the department of Ob/Gyn at Danderyd Hospital.She defended her thesis "Improving access to medical abortion" at Karolinska Institutet 2012.She has ongoing research in both Ob/Gyn and is PI of several clinical studies.She is president of the working group on family planning in Sweden and member of the board of the Swedish Society of Obstetrics and Gynecology and the board of European Society for Contraception and Reproductive Health Sophia Brismar Wendel is MD, PhD, Associate Professor affiliated to Karolinska Institutet, Stockholm, Sweden, and senior consultant in obstetrics and gynecology and labor ward lead at Danderyd Hospital, Stockholm, Sweden.She defended her thesis ''HPV and progression markers in cervical intraepithelial neoplasia -clinical and diagnostic impact" at Karolinska Institutet in 2009.Her current research regards complications at childbirth.She is member of the board of the Swedish Society of Obstetrics and Gynecology, PI of the ongoing EVA trial, and steering group member of several other multicenter RCTs.V. Ankarcrona, S. Hesselman, H. Kopp Kallner et al.European Journal of Obstetrics & Gynecology and Reproductive Biology 269 (2022) 62-70