Midwifery 124 (2023) 103763

Contents lists available at ScienceDirect

Midwifery
journal homepage: www.elsevier.com/locate/midw

Perineal massage and warm compresses – Randomised controlled trial

for reduce perineal trauma during labor
Silvia Rodrigues a,∗, Paulo Silva b, Fátima Rocha b, Ligia Monterroso c, João Neves Silva d,
Natacha Quintal de Sousa b, Ramon Escuriet e
a
Biomedical Sciences Institute Abel Salazar, Portugal and Hospital of Braga, Sete Fontes - São Victor, 4710-243 Braga, Portugal
b
Hospital of Braga, Sete Fontes - São Victor, 4710-243 Braga, Portugal
c
Higher Health Institute, Rua Castelo de Almourol, nº 13, 4720-155 Amares, Portugal
d
ISAVE – Instituto Superior de Saúde, CICS – Centro Interdisciplinar em Ciências da Saúde, Portugal
e
Health and Integrated Care division, Catalan Health Service, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Objective: The aim of this study is to evaluate the effect of perineal massage and warm compresses
Received 20 June 2022 technique on the perineum integrity during second stage of labor.
Revised 12 June 2023
Design and setting: A single-center, prospective, randomized controlled trial was conducted between
Accepted 14 June 2023
March 1st, 2019, and December 31st, 2020, at Hospital of Braga. Participants: Women with 18 years or
older, between 37 weeks and 41 weeks pregnant, in whom a vaginal birth of a fetus in the cephalic pre-
Keywords: sentation was planned were recruited. Eight hundred forty-eight women were randomly assigned (Per-
Perineal massage ineal massage and warm compresses group, n = 424 and control group, n = 424), and 800 women,
Warm compresses both perineal massage and warm compresses group (n = 400) and control group (n = 400) were in-
Perineal trauma
cluded in the strict per protocol analysis. Intervention: In the perineal massage and warm compresses
Second stage
group, women received perineal massage and warm compresses and in the control group, women re-
ceived hands-on technique.
Results: The incidence of intact perineum was significantly higher in the perineal massage and warm
compresses group [perineal massage and warm compresses group: 47% vs control group: 26.3%; OR 2.53,
95% CI 1.86–3.45, p<0.001], whereas second-degree tears and episiotomy rate were significantly lower in
this group [perineal massage and warm compresses group: 7.2% vs control group: 12.3%; OR 1.96, 95% CI
1.17–3.29, p = 0.010 and perineal massage and warm compresses group: 9.5% vs control group: 28.5%; OR
3.478, 95% CI 2.236–5.409, p<0.001, respectively]. Also, obstetric anal sphincter injury with and without
episiotomy and second-degree tears with episiotomy were significantly lower in the perineal massage and
warm compresses group [perineal massage and warm compresses group: 0.5% vs control group: 2.3%; OR
5.404, 95% CI 1.077–27.126, p = 0.040 and perineal massage and warm compresses group: 0.3% vs control
group: 1.8%; OR 9.253, 95% CI 1.083–79.015, p = 0.042, respectively].
Conclusions: The perineal massage and warm compresses technique increased the incidence of intact per-
ineum and reduced the incidence of second-degree tear, episiotomy and obstetric anal sphincter injury.
Implications for practice: Perineal massage and warm compresses technique is feasible, inexpensive and
reproductible. Therefore, this technique should be taught and trained to midwives students and midwives
team. Thus, women should have this information and have the option to decide whether they want to
receive the perineal massage and warm compresses technique in the second stage of labor.
© 2023 The Author(s). Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

∗
Corresponding author at: Avenida D. João II, nº48 3ºdireito 4715-304-Braga, Portugal.
E-mail addresses: [email protected] (S. Rodrigues), [email protected] (P. Silva), [email protected] (F. Rocha),
[email protected] (L. Monterroso), [email protected] (J.N. Silva), [email protected] (N.Q. de Sousa), [email protected] (R. Escuriet).

https://doi.org/10.1016/j.midw.2023.103763
0266-6138/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

Trial registration http://www.ClinicalTrials.gov:NCT05854888 to perineal massage alone technique (Hong et al., 2022). Although,
the traditional routine management of perineum during the sec-
ond stage of labor has been hands-on technique (Trochez et al.,
2011). Consequently, more research is needed to evaluate if per-
Problem: Perineal trauma are associated to short and long-
ineal massage and warm compresses technique during the sec-
morbidity. Women with episiotomy, second-degree tear and
OASIS reported worse birth experience. Specifically women ond stage of labor compared to hands-on technique could promote
with OASIS presented a less positive attitude toward future better perineal outcomes to women. The randomised controlled
pregnancies. trial is considered the best design to evaluate interventions, as it
What is Already Known: The perineal massage and warm distributes both measured and unmeasured confounding variables
compresses technique decreased the need of perineal sutur- equally between study groups, assuming the sample is of adequate
ing and major perineal injury compared to hands-off. How- size (Abramson and Abramson, 2011). This randomised controlled
ever, when compared to perineal massage alone, did not de- trial entitled Perineal Massage and Warm Compresses (PeMWaC)
crease the need of requiring suturing for perineal trauma, was part of a bigger study and was conducted in a Tertiary Hos-
OASIS and episiotomy.
pital maternity, in which high risk pregnancies and deliveries may
What this Paper Adds: The perineal massage and warm com-
be attended. This is the first study to compare the perineal mas-
presses technique increased the intact perineum and reduced
the second-degree tear, episiotomy and OASIS. sage and warm compresses during the second stage of labor with
hands-on technique. These provide justification for this trial and
assisted in the design of a randomised controlled trial for perineal
protection intervention.
Introduction
Aim
Approximately 85% of vaginal births are associated with some
degree of perineal trauma, which can cause a significant morbid- The aim of this study is to evaluate the effect of Perineal Mas-
ity in women (Begley et al., 2019). Perineal tears may cause short- sage and Warm Compresses intervention on the perineum integrity
term complications such as pain, infection and severe bleeding when applied in women on second stage of labor.
(Jiang et al., 2017) but when the anal sphincter is included, these
injuries can be associated with long-morbidity including inconti-
nence urinary, anal, pelvic pain and sexual dysfunction (Pierce- Methods
Williams et al., 2019). Generally, the degree of perineal pain is di-
rectly related to the degree of the perineal trauma. In the study Design and trial size
developed by Manresa et al. (2020) the odds of pain at 10 days
and dyspareunia at 6 months postpartum were four- and five- The present study was a single-center, prospective, randomised
fold greater, respectively, in women with second-degree trauma controlled trial and parallel design. It was conducted between
or episiotomy, than if the perineum had remained intact or suf- March 1st, 2019 and December 31st, 2020 at one Hospital, which
fered a first-degree perineal trauma. Also, women with an epi- is a tertiary hospital with an obstetric unit with more than 30 0 0
siotomy or obstetric anal sphincter injury (OASIS) mentioned a births annually. The calculation of sample size was performed with
worse birth experience than those with first- and second-degree https://clincalc.com/Stats/SampleSize.aspx. Briefly, the values of de-
perineal tears (Molyneux et al., 2021). Specifically women with sired statistical power (80%), significance level (5%) and effect size
OASIS presented a less positive attitude toward future pregnan- (10% difference between groups, considering an intact perineum
cies (Rodaki et al., 2022). Nevertheless, the more recent evidence prevalence rate of 25% in the control group and 35% in the ex-
does not support routine episiotomy, since it is associated with perimental group, according to previous works by Dahlen et al.,
an increased risk of OASIS and no clear difference in perineal 2007a and Rodrigues et al., 2019) were inserted in the software as
infection rates, chronic pain, long-term dyspareunia (at least six input parameters, concerning a two-tailed Student’s t test for in-
months postpartum) and long-term urinary incontinence (at least dependent samples. The automated calculation performed by soft-
six months postpartum) when compared with selective episiotomy ware suggests a minimum sample size of 656 participants, equally
(Jiang et al., 2017).Otherwise, the selective episiotomy increased distributed by the control (328 participants) and experimental (328
the incidence of intact perineum without increased the incidence participants) groups. Assuming an estimated exclusion rate of 15%,
of anal sphincter tears (Franchi et al., 2020) Hence, the challenge as suggested by Lavesson et al., 2014, the minimum sample size
is not only to reduce the incidence of OASIS, but also to reduce the was increased to 772 participants (= 656/0.85), considering the ex-
incidence of minor lacerations (first- and second-degree tears) and clusion effect previously mentioned. The effective sample size of
routine episiotomies. the study was 800 participants, 400 in the control group and 400
In addition, several perineal management techniques used dur- in the experimental group. The trial was approved by the Hospi-
ing labor have been studied and their effect on episiotomy, minor tal Ethics Committee (reg. no. 2_2019 approved February 2019).
tears and OASIS have been published. A Cochrane systematic re- It was registered in http://www.ClinicalTrials.gov:NCT05854888.
view published in 2017, concluded that the application of warm CONSORT recommendations for reporting randomised, controlled,
compresses, and the perineal massage during the second stage of clinical trials involving nonpharmacologic treatment were followed
labor, individually, may reduce the occurrence of OASIS but evi- (Boutron et al., 2017).
dence on the benefits of these techniques on intact perineum are
unclear (Aasheim et al., 2017). An overview of systematic reviews Inclusion criteria
about this field developed by Zang et al. (2022) concluded that
perineal massage and warm compresses, individually, in the sec- Eligible participants were women with 18 years or older, be-
ond stage of labor could be the better choice for preventing per- tween 37 and 41 weeks pregnant, in whom a vaginal birth of a
ineal laceration. Two years ago, a study was published comparing fetus in the cephalic presentation was planned and had to be able
the perineal massage and warm compresses technique to hands-off to provide informed written consent. Exclusion criteria were mul-
technique (Goh et al., 2021). Last year, another study was published tiple pregnancy, meconial amniotic fluid, fetal distress, suspicion of
comparing the perineal massage and warm compresses technique fetal growth restriction and gestational hypertensive disorders.

2
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

Recruitment of study participants sification (Roper et al., 2020). Secondary outcomes were newborn
Apgar score at five minutes, newborn admission to neonatal inten-
Recruitment started March 1st, 2019, and the trial was planned sive care unit, maternal satisfaction with intervention (satisfied or
to proceed for 24 months but the necessary sample was reached not satisfied) and recommend an intervention allocated to a friend
on 31st December and then the recruitment period lasted less than (yes or no).
24 months. Study participants were women in labor, who antic-
ipated a spontaneous vaginal birth in labor ward. On the labor Statistical analyses
ward, when the woman was in active labor, the midwife confirmed
the eligibility and asked for informed consent to participate in Continuous variables were assessed for normal distribution
the trial. All participants signed an informed consent form and re- with the Kolmogorov–Smirnov test. When normally distributed,
ceived a copy wherein the study was described in detail. Women they were analysed with an independent t-test, and when not,
could withdraw from the study at any point without prejudice to with a Mann–Whitney U test. Categorical variables including the
their care or their relationship with their health practitioners. primary outcome were analysed with the Pearson chi-squared test
or Fisher’s exact test as appropriate, and Adjusted Odds Ratio
Randomization and information (OR’s)s of the perineal subtypes were calculated using logistic re-
gression controlling for baseline characteristics with significant dif-
Block randomization was computer-generated, with blocks size ferences between the two groups (presented as a1 OR) and known
of two, four and six, and used randomly mixed block sizes for the predisposing factors (presented as a2 OR) including maternal age,
allocation to perineal massage and warm compresses technique or parity, BMI, type of previous birth, analgesia, duration of oxytocin
control group. The randomization was made by a third party not augmentation in second stage of labor, length of the active sec-
otherwise involved in the trial. Identical, opaque, sealed, sequen- ond stage of labor, birth position, height perineum, newborn birth-
tially numbered envelopes containing the randomization assign- weight, occiput posterior presentation and fundal pressure. The
ment were placed in the central office on the labor ward. The ran- odds ratios were presented with 95% confidence intervals (Cis). A
domization envelope was opened by the midwife when the partic- two-sided P value <0.05 was considered as significant. Data analy-
ipant entered the second stage of labor and was destroyed there- ses was performed using IBM SPSS Statistics 27 version (SPSS Inc).
after. Participants were random allocated to one of the two groups,
perineal massage and warm compresses on the second stage of la- Pilot study
bor or control group (hands-on). The allocation was only shown to
the midwife and if necessary, the obstetrician. The principal investigator trained five keypersons to be coordi-
nators and secure uniform use of the perineal massage and warm
Data collection and analysis compresses technique. All midwives were trained for the perineal
massage and warm compresses technique by the principal investi-
After the placenta delivery, a blinded midwife not otherwise gator and the coordinators to secure uniform use of the technique.
involved in the birth but trained, assessed the perineum (graded Training sessions included an introductory video of the combined
the perineal tears) and registered the primary outcomes. Recorded technique and practical training in the deliver ward during the pi-
third- and fourth-degree tears were validated through manual as- lot study. The midwives attended the three educational sessions
sessment of patient records and were incorporated into the fi- (including theoretical training in anatomy, assessment, and clas-
nal dataset. Midwives who assess the primary outcomes and the sification of perineal tears and suturing techniques) developed in
principal investigator were blinded to the randomization. The data the Hospital to ensure uniform training. All midwives involved in
about the variables included in the study (Clinical Characteristics the trial were trained in the technique and had regular updates
and Labor Process) and the secondary outcomes (Apgar at five sessions in the pilot study stage. The perineal massage and warm
minutes and admission to neonatal intensive care unit) were regis- compresses had not been used by most of the midwives partici-
tered by the midwife responsible for the birth. Twenty four hours pating in the trial previously, so education and training occurred
after birth, a questionnaire was sent to the participants to assess in the pilot study stage (during one month before the start of the
their satisfaction with the intervention and whether they would trial). Perineal massage and warm compresses technique was stan-
recommend the intervention they were allocated to a friend. Data dardized in a pilot study with 30 women in labor. Clear protocols
were recorded on clinical registration forms and inserted in a SPSS on the application of perineal massage and warm compresses were
database (IBM ®SPSS® software SPSS Inc, Chicago, IL, USA) by the available in the delivery ward. There were coordinators to super-
principal investigator. vise the procedure according to the protocol of the study. Regu-
To ensure that perineal massage and warm compresses tech- lar checking of the procedure by the principal investigator ensured
nique did not cause markedly more perineal trauma, interim anal- protocol compliance.
yses were performed by a data monitor not otherwise involved in
the trial after the first year and consecutively thereafter 6 months. Procedure

Outcome measurements A soft perineal massage/warm compresses in the second stage

of labor program: In the second stage of labor, the midwife per-
Primary outcomes were intact perineum (absence of tissue sep- formed a soft perineal massage between 3 o’clock and 9 o’clock
aration at any site), first-degree tear (involves skin of the per- positions (U-shaped reciprocating motion) wearing sterile gloves
ineum and vaginal mucosa); second-degree tear (involves deeper and lubricated their hand with sterile lubricant. The massage
layers of perineal muscle); third-degree tear (involves the anus); lasted 10 min and the degree of downward pressure by the thumb
fourth-degree tear (involves the anus and rectal mucosa); OA- was determined according to mothers’ response (Mei-dan et al.,
SIS (third and fourth-degree tears) and the incidence of epi- 2008). Perineal massage was established on the II Hodge Plan, be-
siotomy. The Sultan Classification of Perineal trauma was used, tween maternal contractions and regardless of maternal position
which is internationally validated, allowing data reproducibility (Demirel et al., 2015). After the perineal massage, the women could
(Sultan and Kettle, 2007). Of the thirteen guidelines published be- change and adopted the birth position that she preferred. The ap-
tween 2013 and 2019, twelve guidelines included the Sultan clas- plication of warm compresses was performed by the midwife be-

3
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

Table 1
Clinical characteristics, Portugal, 2019–2020.

PeMWaC
group Control group
n = 400 n = 400 p

Clinical characteristics
Maternal age (years), mean (SD) 31.8 (5.1) 31.4 (4.8) 0.276
Body mass index (kg/m2), mean (SD) 28.7 (4.2) 28.9 (4.6) 0.460
Parity, n (%) Nullipara (without vaginal birth) 206 (51.5) 224 (56) 0.114
Multipara 194 (48.5) 176 (44)
Prenatal childbirth preparation course, n (%) Yes 154 (38.5) 181 (45.3) 0.025
No 246 (61.5) 216 (54)
Gestational age (weeks), mean (SD) 39.3 (1.1) 39.3 (1.1) 0.717
Type Previous Birth, n (%) Vaginal birth 182 (45.5) 167 (41.8) 0.536
Cesarean section 24 (6) 33 (8.3)
Vaginal birth + Cesarean section 10 (2.5) 11 (2.8)
Newborn birthweight (g), mean (SD) 3345 (385) 3328 (400) 0.545

Data are expressed as mean (SD) or count (%).

tween Hodge plans III and IV, [the height of the fetal presentation 95% CI 1.861–3.447, p<0.001); second degree tear (a2 OR 1.966,
was assessed by the plans of Hodge (World Health Organization, CI 1.174- 3.293, p = 0.010) and episiotomy (a2 OR 3.478, 95% CI
2018)] during pushes and regardless of maternal position. A metal 2.236–5.409, p<0.001) did not change.
jug filled with warm water (between 45° and 59 °C) was used to The incidence of OASIS total (with and without episiotomy) was
soak the compresses, which were squeezed out before being gen- 2/400 (0.5%) [PeMWaC] vs 9/400 2.3% [control group] and did not
tly placed on the perineum during contractions. The compresses differ between the groups. After adjustment for clinical character-
were rewarmed between pushes. Water was changed each fifteen istics and labor process variables with significant differences be-
minutes and the compresses were changed as needed to main- tween groups, the result did not change. However, after adjustment
tain warmth and cleanliness at the end of the second stage of la- for predefined risk factors, significantly differences between groups
bor. Women assigned to the standard care group did not have the were found (a2 OR 5.404, 95% CI 1.077–27.126, p = 0.040). The in-
combination of perineal massage and warm compresses applied to cidence for second-degree tears with episiotomy was (1/400 0.3%
their perineum on the second stage of labor (Dahlen et al., 2007a). [PeMWaC] vs 7/400 (1.8%) [control group] and did not differ be-
Considering the technique hands-on, the midwife placed the tween the groups. After adjustment for clinical characteristics, la-
index, middle ring and little fingers of the non-dominant hand bor process variables with significant differences between groups
close together on the infant’s occiput, with the palm turned to- and for predefined risk factors (a2 OR 9.253, 95% CI 1.083–79.015,
ward the anterior region of the perineum, when the infant’s head p = 0.042), significantly differences between groups were found.
was crowning. In this manner, expulsion was controlled, by main- In the secondary outcomes, the maternal satisfaction with inter-
taining the flexion of the head. Simultaneously, the dominant hand vention and recommend an intervention allocated to a friend was
was flattened out and placed on the posterior perineum, with the significantly higher for women in the PeMWaC group compared to
index finger and the thumb, forming a "U" shape, exerting pressure control group. There was no difference in APGAR scores and new-
on the posterior region of the perineum during the crowning pro- born admission to neonatal intensive care unit between the groups.
cess. During the birth of the shoulders and the remainder of the
body, the dominant hand was kept in place, protecting the pos-
terior region of the perineum, while the non-dominant hand sup- Discussion
ported the infant’s head, allowing external rotation and the birth
of the shoulders spontaneously. After both shoulders had been de- Primary outcomes
livered, the midwife removed the dominant hand from the pos-
terior perineum and supported the infant’s neck with one hand, Several systematic reviews have shown the increased of in-
while supporting the remainder of the body with the other hand tact perineum and reduction in episiotomy and OASIS rates af-
(De Costa et al., 2006). ter individually these techniques has been performed on sec-
ond stage of labor (Magoga et al., 2019; Aquino et al., 2020;
Results Li et al., 2023). Which is consistent with our results. Since 2016,
first the warm compresses technique and then the perineal mas-
From March 2019 to December 2020, 848 women were ran- sage technique have been recommended (ACOG, 2016; WHO, 2018;
domised, 826 women received the protocol but were 800 women Gimovsky et al., 2022). However, recent systematic review about
that received the strict protocol. The flow of participants is illus- perineal massage during the second stage of labor indicated a de-
trated in Fig. 1. Clinical characteristics are shown in Table 1, Labor creased in episiotomy rate but did not find a reduction in in-
process variables are show in Table 2. Most characteristics were cidence and severity of perineal tears (Marcos-Rodríguez et al.,
similar between groups. Table 3 shows the primary and secondary 2023). On the other hand, another systematic review showed that
outcomes for 400 women randomised to PeMWaC and for 400 perineal massage during labor decreased OASIS, but not episiotomy
women in control group. The incidence of intact perineum was (Venugopal et al., 2022). Contrary, a systematic review concluded
188/400 (47%) [PeMWaC] vs 104/400 (26.3%) [control group]; sec- that intervention during labor was not effective for the prevention
ond degree tear was 29/400 (7.2%) [PeMWaC] vs 49/400 (12.3%) of perineal trauma when compared to no intervention. However,
[control group], and episiotomy was 38/400 (9.5%) [PeMWaC] vs prenatal perineal massage was associated with a lower risk of per-
114/400 (28.5%) [control group], which significantly differed be- ineal tear (da Silva et al., 2023). On the other hand, another study
tween groups. After adjustment for clinical characteristics, labor suggested that women who did not performed prenatal perineal
process variables with significant group differences and for pre- massage can benefit from receiving perineal massage during labor
defined risk factors, the results of intact perineum (a2 OR 2.533, (Kiremitli et al., 2022).

4
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

Fig. 1. Recruitment flow chart of a randomised trial of perineal massage and warm compresses techinque compared with control group (hands of techinque) in the second
stage of labor.

The results from the present study are in line with a ran- sage and warm compresses during second stage of labor did not
domised controlled trial which involved 180 nulliparas and demon- decrease the likelihood of nulliparous women requiring sutur-
strated that combined perineal massage and warm compresses ing for perineal trauma, episiotomy and OASIS rates compared
during active second stage pushing decreased the need of per- to perineal massage alone group (Hong et al., 2022). The evi-
ineal suturing, episiotomy and major perineal injury (second- dence from recent studies, at first site seems to be controversial,
degree tears or worse) compared to control group ‘‘hands off” but we know the control group was different between studies. In
technique (Goh et al., 2021). Our study added that PeMWaC tech- Hong et al. (2022) the control group was alone perineal massage
nique decreased the incidence of episiotomy plus second-degree and in Goh et al. (2021) the control group was hands-off technique,
tear compared to control group (hands-on). Another trial involved while in the present study was hands-on technique which is the
282 term nulliparas reported that the combined perineal mas- current practice in the hospital. After a critical analysis, it allowed

5
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

Table 2
Labor Process Variables, Portugal, 2019–2020.

PeMWaC
group Control group
n = 400 n = 400 p

Labor Process Variables

Onset of labor, n (%) Spontaneous 267 (66.8) 258 (64.5) 0.291
Induction 131 (32.8) 139 (34.8)
Rupture of membranes, n (%) Before labor 93 (23.3) 111 (27.8) 0.078
During labor 307 (76.8) 287 (71.8)
Analgesia, n (%) Epidural 386 (96.5) 371 (92.8) 0.135
Raquianestesy 4 (1) 12 (3)
No analgesy 10 (2.5) 15 (3.8)
First stage of labor, n (%) No 102 (25.5) 115 (28.8) 0.824
Birth ball 1 (0.25) 0
Hidroterapy 1 (0.25) 2 (0.5)
Walking 96 (24) 87 (21.8)
Birth ball +walking 45 (11.3) 47 (11.8)
Hidroterapy +walking 39 (9.8) 39 (9.8)
Maternal birth position, n (%) Lithotomy 204 (51) 224 (56) 0.083
Semiseated 134 (33.5) 124 (31)
Lateral 58 (14.5) 50 (12.5)
All fours 4 (1) 2 (0.5)
Length of the first stage of labor (h), mean (SD) 5.37 (3.6) 4.85 (2.9) 0.025
Length of the active second stage of labor (min), mean (SD) 40.80 (30.5) 37.53 (30.1) 0.128
Duration of oxytocin augmentation in second stage of labor (min), mean (SD) 16.5 (24.9) 15.3 (23.1) 0.467
Type of cephalic presentation, n (%) Occiput anterior 366 (91.5) 363 (90.8) 0.402
Occiput posterior 34 (8.5) 37 (8.2)
Nuchal cord, n (%) Yes 244 (61) 276 (69) 0.006
No 156 (39) 120 (30)
Height Perineum (cm), mean (SD) 3.30 (0.58) 3.35 (0.54) 0.268
Perineal edema, n (%) Yes 37 (8.2) 34 (8.5) 0.402
No 363 (90.8) 366 (91.5)
Fundal Pressure, n (%) Yes 66 (16.5) 69 (17.3) 0.425
No 334 (83.5) 331 (82.7)
Valsalva
Maneuver, n (%) Yes 60 (15) 53 (13.3) 0.267
No 340 (85) 347 (86.7)
Method of pushing, n (%) Directed pushing coaching 28 (7) 28(7) 0.263
Spontaneous pushing 239 (59.8) 237 (59.3)
Both 133 (33.2) 135 (33.7)

Data are expressed as mean (SD) or count (%).

us to conclude that the groups that included perineal massage The trial was underpowered to assess the OASIS, however, the
showed a reduction in perineal trauma and this finding is consis- difference in the rates was sufficient to warrant further investiga-
tent with our findings. In Hong et al. (2022) study, in the massage tion. In this study, OASIS incidence was 11/800 (1.4%), higher than
alone group, the participants received perineal massage continu- reported in the literature (1%) from 2019 to 2020 data in sponta-
ously during and between pushes whereas in the combined group neous vaginal birth in Portugal (OECD, 2021). However, equal to
massage was done only during contractions in the second stage. In the average for OECD countries between 2019 and 2020 in sponta-
Goh et al. (2021) study, the perineal massage was restricted for the neous vaginal birth (1.4%) (OECD, 2021). In Portugal, the OASIS rate
duration of the pushing coordinated with contractions until near increased from 0.4% between 2012 and 2017 (OECD, 2019) to 1%
crowning. In the other hand, in this study was different to other between 2019 and 2020 in spontaneous vaginal birth (OECD, 2021).
studies because the perineal massage was only performed between On the other hand, Blondel et al. (2016) reported 66.8% episiotomy
contractions. In systematic review and meta-analysis developed by rate in spontaneous vaginal birth in Portugal from 2010 data and
Aquino et al. (2020), from 9 randomised controlled trials (RCT’S), was among the European countries with the highest rate of epi-
in one trial, the perineal massage was performed between con- siotomies. However, the episiotomy rate in spontaneous vaginal
tractions which is line to our study. In the other 8 trials, in three, birth has been gradually decreasing in Portugal, from 63% in 2013
the perineal massage was performed during and between pushing to 54.0% in 2015 (Teixeira et al., 2022). In a recent study, the data
time, in four trials was only performed during the pushing time from surveys answered by 21,027 women of 12 countries in Europe
and in one was not reported. Despite, these differences between (1683 women in Portugal) demonstrated that Portugal presented
RCT’S, Aquino et al. (2020) in the systematic review and meta- 40.1% of episiotomy rate in spontaneous vaginal birth between
analysis, concluded that perineal massage during labor was asso- 2020 and 2021 (Lazzerini et al., 2022). According to data from
ciated with significant higher intact perineum rate and lower risk public hospitals in Portugal between 20 0 0 and 2015, OASIS rate
of OASIS and episiotomy which converges with our results. Thus, decreased in women with spontaneous vaginal birth without epi-
the different timings of perineal massage between studies, allowed siotomy but increased in women with episiotomy (Teixeira et al.,
us to assume that regardless the timing to perform the perineal 2022). Therefore, the rationale for conducting routine episiotomies
massage, this perineal protection technique showed a reduction in to prevent OASIS was not justified by a recent literature (Jing et al.,
perineal trauma. However, in this study, the perineal massage was 2017; Franchi et al., 2020; Teixeira et al., 2022) and is consistent
performed only between contractions and pushes, with the pur- with our findings. After a critical analysis, we found significantly
pose to not cause more discomfort during contraction or when fe- fewer OASIS (0.5%) in PeMWaC group compared to control group
tal presentation progressing in the birth canal. (2.3%) and fewer episiotomies (11.1%) in PeMWaC group compared

6
 S. Rodrigues, P. Silva, F. Rocha et al.
Table 3
Primary and secondary outcomes, Portugal, 2019–2020.

PeMWaC
group Control group
n = 400 n = 400

Primary Outcomes n (%) n (%) OR (95% CI) p a1ORa (95% CI) p a2ORb (95% CI) p

Intact Perineum 188 104 2.502 p<0.001 2.552 p<0.001 2.533 p<0.001
(47) (26.3) (1.858–3.369) (1.877–3.470) (1.861–3.447)
First degree tear 138 105 0.733 p = 0.052 0.710 p = 0.031 0.748 p = 0.070
(34.5) (26) (0.536–1.003) (0.520–0.970) (0.546–1.024)
Second degree tear 29 49 1.826 p = 0.022 1.914 p = 0.013 1.966 p = 0.010
(7.2) (12.3) (1.091–3.054) (1.147–3.195) (1.174- 3.293)
Episiotomy 38 114 3.692 p<0.001 3.690 p<0.001 3.478 p<0.001
(without spontaneous tears) (9.5) (28.5) (2.456–5.551) (2.379–5.722) (2.236–5.409)
Episiotomy Total 44 (11.1) 135 3.010 p<0.001 3.284 p<0.001 3.135 p<0.001
(33.8) (2.125–4.251) (2.456–4.465) (2.275–4.358)
Episiotomy + first degree tear 4 12 2.867 p = 0.074 3.275 p = 0.047 3.206 p = 0.055
(1) (3) (0.904–9.089) (1.017–10.548) (0.976–10.528)
Episiotomy + second degree tear 1 7 7.920 p = 0.053 9.507 p = 0.037 9.253 p = 0.042
(0.3) (1.8) (0.971–64.618) (1.147–78.824) (1.083–79.015)
7

Episiotomy + OASIS 1 2 2.466 p = 0.466 3.253 p = 0.377 2.719 p = 0.473

(Episiotomy+ third or fourth-degree tear) (0.3) (0.5) (0.218–27.897) (0.238–44.475) (0.177–41.836)
OASIS 1 7 7.107 p = 0.067 7.427 p = 0.063 7.060 p = 0.077
(Third or fourth-degree tear) (0.3) (1.8) (0.870–58.031) (0.900–61.280) (0.810–61.539)
OASIS Total (with and without episiotomy) 2 9 4.581 p = 0.053 4.697 p = 0.050 5.404 p = 0.040
(0.5) (2.3) (0.983–21.335) (0.999–22.073) (1.077–27.126)
Secondary Outcomes
Admission 1 0 p = 0.497
NICU (0.3)
Apgar 5 min 10 10 p = 0.347
(10–10) (10–10)
Maternal satisfaction with intervention, n (%) Satisfeid 386 (96.5) 371 (92.8) p = 0.019
Recommend an intervention allocated to a friend, Not satisfeid 14 (3.5) 29 (7.2) p = 0.019
n (%) Yes 386 (96.5) 371 (92.8)
No 14 (3.5) 29 (7.2)

Data are expressed as mean (SD), median (interquartile range), or count (%).
OR odds ratios. NICU, neonatal intensive care unit.
aa1OR: OR adjusted for significant characteristic differences between the two groups (Tables 1 and 2), length of the first stage of labor, nuchal cord and prenatal childbirth preparation course.
ba2OR: OR adjusted for predefined risk factors: maternal age, parity, BMI, type of previous birth, analgesia, duration of oxytocin augmentation in second stage of labor, length of the active second stage of labor, birth position,

Midwifery 124 (2023) 103763

height perineum, newborn birthweight, occiput posterior presentation and fundal pressure.
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

to control group (33.8%). This study found a positive relation be- This study also has some limitations, like the validity of this
tween episiotomies and OASIS. Thus, the reason for the increase trial that could be affected by the fact that several midwives per-
in the OASIS rate appears to be a better diagnosis of OASIS rather formed the intervention, and the perineal tears were evaluated by
than the decreased in episiotomy rate. Thus, it is considered safe several midwives, although numerous birth assistants is the real-
decreased the episiotomy rate in Portugal (Teixeira et al., 2022) and ity at most centres, which increases the external validity and re-
the accepted indications for episiotomy are to shorten the second productible of the results. Other limitation is that it was a single-
stage of labor when there is suspected fetal hypoxia; to prevent center trial, which might be interpreted as a limitation although
obstetric anal sphincter injury in vaginal operative births, or when it increases internal validity of the trial. The anal examination was
obstetric sphincter injury occurred in previous births (Laine et al., only performed in women with second-degree tears and OASIS. We
2022) tried to overcome this limitation, by validating all cases of OASIS
Studies with aggregated data suggested that obstetric practices included in the trial through the access of the clinical files and
as maternal birth position and perineal protection techniques on sending two questionnaires, one about the pelvic floor function
the second stage of labor could contribute to differences in rates and other about sexual function at three and six months after birth
of OASIS and episiotomies (Blondel et al., 2016; Teixeira et al., to all the women included in the trial.
2022). In this study, perineal massage and warm compresses tech-
nique contributed to differences in episiotomy rate and OASIS be- Conclusion
tween groups, with maternal birth positions being similar between
groups. Nevertheless, the solution for OASIS problem include edu- This study showed a protective effect of perineal massage and
cation, communication and teamwork (Persson et al., 2023). warm compresses technique during the second stage of labor com-
pared with control group (hands-on), particularly with an increase
in incidence of intact perineum and a decrease second-degree tear,
Secondary outcomes OASIS, and episiotomy rates. The morbidity generated by perineal
trauma during spontaneous vaginal birth justifies the need to find
Women’s preferences and experiences are an important part an intervention to minimize it, and this technique is feasible, in-
of the clinical decision-making process. In this study, participants expensive and reproductible. Therefore, this technique should be
expressed higher satisfaction with the intervention in PeMWaC taught and trained to midwives students and midwives team. In
group compared to hands-on group. A different satisfaction met- addition, the midwives should provide perineal massage and warm
ric was also used, that of recommending an intervention allocated compresses technique according to women’s preferences. Thus,
to a friend, and we found that more women allocated to PeMWaC women should have this information and have the option to de-
group would recommend this intervention to their friends. The cide whether they want to receive the perineal massage and warm
findings from this study were in line with Goh et al. (2021) that compresses technique in the second stage of labor.
reported significantly higher maternal satisfaction score with com- Ethical approval was granted by the Hospital of Braga, Portugal.
bined massage and warm compresses compared to ‘‘hands off”
technique. Contrary, in other trial, the maternal satisfaction was Funding sources
not significantly different between groups which combined per-
ineal massage and warm compresses compared to massage only None Declared.
(Hong et al., 2022). It allowed us to conclude that the groups
that included perineal massage showed more satisfaction with the Declaration of Competing Interest
technique. Also, according to Dahlen (2007b), 85.7% of women
preferred to use warm compresses again for the next birth and The authors declare that they have no known competing finan-
would also recommend them to friends. Despite, the evidence on cial interests or personal relationships that could have appeared to
women’s preferences being insufficient to make recommendations, influence the work reported in this paper.
our data demonstrated that PeMWaC technique was satisfactory for
women. CRediT authorship contribution statement
Otherwise, according to the feedback from the coordinators, the
entire team of midwives participated in the study. After completing Silvia Rodrigues: Conceptualization, Data curation, Formal anal-
participant recruitment, the team of midwives adopt the technique, ysis, Investigation, Methodology, Project administration, Resources,
as they continued to perform the perineal massage and warm com- Software, Supervision, Validation, Writing – original draft, Writ-
presses technique. This is a feasible, inexpensive and reproductible ing – review & editing. Paulo Silva: Conceptualization, Data cu-
technique. ration, Formal analysis, Investigation, Methodology, Project admin-
istration, Software, Validation. Fátima Rocha: Visualization. Ligia
Monterroso: Visualization. João Neves Silva: Methodology, Vali-
Strengths and limitations dation. Natacha Quintal de Sousa: Methodology, Project adminis-
tration, Resources, Software, Supervision, Validation, Visualization,
This study has several strengths, the first strength was the in- Writing – original draft, Writing – review & editing. Ramon Es-
clusion of confounders variables as fundal pressure, Valsalva ma- curiet: Methodology, Project administration, Software, Supervision,
neuver and perineal edema during the second stage of labor, that Validation, Writing – original draft, Writing – review & editing.
were not included in previous studies about this field. In this study
these confounding variables were similar between groups. The sec- Acknowledgments
ond strength was the sample size, in this study the sample size is
approximately three times larger than recent studies on the same We thank all the women who participated in the trial and their
topic (Goh et al., 2021; Hong et al., 2022). The third strength was care providers for their contribution to this trial.
the fact that was analysed the incidence of episiotomy plus an-
other degree of tear, was not investigated in previous studies about Supplementary materials
this topic. This study found that perineal massage and warm com-
presses reduced spontaneous second-degree tears in women with Supplementary material associated with this article can be
episiotomy. found, in the online version, at doi:10.1016/j.midw.2023.103763.

8
S. Rodrigues, P. Silva, F. Rocha et al. Midwifery 124 (2023) 103763

References Lavesson, T., Griph, I.D., Skarvad, A., et al., 2014. A perineal protection device de-
signed to protect the perineum during labor: a multicenter randomized con-
Aasheim, V., Nilsen, A.B.V., Reinar, L.M., et al., 2017. Perineal techniques during the trolled trial. Eur. J. Obstet. Gynecol. Reprod. Biol. 181, 10–14. doi:10.1016/j.ejogrb.
second stage of labor for reducing perineal trauma. Cochrane Database Syst. Rev. 2014.07.006.
6 (6), CD006672. doi:10.1002/14651858.CD006672.pub3. Lazzerini, M., Covi, B., Mariani, I., et al., 2022. Quality of facility-based maternal and
Abramson, J.H., Abramson, A.Z., 2011. Research methods in community medicine: newborn care around the time of childbirth during the COVID-19 pandemic:
surveys, epidemiological research, programme evaluation. Clin. Trials John Wiley online survey investigating maternal perspectives in 12 countries of the WHO
& Sons. European Region. Lancet Reg. Health Eur. 13. doi:10.1016/j.lanepe.2021.100268,
ACOG: The American College of Obstetricians and Gynecologists, 2016. Practice Bul- 100268Epub 2021 Dec 24. Erratum in: Lancet Reg Health Eur 19:100461.
letin No. 165 summary: prevention and management of obstetric lacerations at Li, Y., Wang, C., Lu, H., et al., 2023. Effects of perineal massage during childbirth
vaginal delivery: correction. Obstet. Gynecol. 128 (1), e1–e15. doi:10.1097/AOG. on maternal and neonatal outcomes in primiparous women: a systematic re-
0 0 0 0 0 0 0 0 0 0 0 01523. view and meta-analysis. Int. J. Nurs. Stud. 138, 104390. doi:10.1016/j.ijnurstu.
Aquino, C., Guida, M., Saccone, G., et al., 2020. Perineal massage during labor: a 2022.104390.
systematic review and meta-analysis of randomized controlled trials. J. Matern. Magoga, G., Saccone, G., Al-Kouatly, H., et al., 2019. Warm perineal compresses dur-
Fetal Neonatal Med. 33 (6), 1051–1063. doi:10.1080/14767058.2018.1512574. ing the second stage of labor for reducing perineal trauma: a meta-analysis. Eur.
Begley, C., Guilliland, K., Dixon, L., et al., 2019. A qualitative exploration of tech- J. Obstet. Gynecol. Reprod. Biol. 240 (2), 93–98. doi:10.1016/j.ejogrb.2019.06.011.
niques used by expert midwives to preserve the perineum intact. Women Birth Manresa, M., Pereda, A., Goberna-Tricas, J., et al., 2020. Postpartum perineal pain
32 (1), 87–97. doi:10.1016/j.wombi.2018.04.015. and dyspareunia related to each superficial perineal muscle injury: a cohort
Boutron, I., Altman, D.G., Moher, D., et al., 2017. CONSORT statement for randomized study. Int. Urogynecol. J. 31, 2367–2375. doi:10.10 07/s0 0192- 020- 04317- 1.
trials of nonpharmacologic treatments: a 2017 update and a CONSORT extension Marcos-Rodríguez, A., Leirós-Rodríguez, R., Hernandez-Lucas, P., et al., 2023. Efficacy
for nonpharmacologic trial abstracts. Ann. Intern. Med. 167 (1), 40–47. doi:10. of perineal massage during the second stage of labor for the prevention of per-
7326/M17-0046. ineal injury: a systematic review and meta-analysis. Int. J. Gynaecol. Obstet. 17.
Blondel, B., Alexander, S., Bjarnadóttir, R.I., et al., 2016. Variations in rates of severe doi:10.1002/ijgo.14723.
perineal tears and episiotomies in 20 European countries: a study based on rou- Mei-dan, E., Walfisch, A., Raz, I., et al., 2008. Perineal massage during pregnancy: a
tine national data in Euro-Peristat Project. Acta Obstet. Gynecol. Scand. 95 (7), prospective controlled trial. Israeli Med. Assoc. J. 10, 499–502.
746–754. doi:10.1111/aogs.12894. Molyneux, R., Fowler, G., Slade, P., et al., 2021. The effects of perineal trauma on
da Silva, M.L., de Sousa, T.A.B.P., Leite, L.W.C., et al., 2023. The effectiveness of in- immediate self-reported birth experience in first-time mothers. J. Psychosom.
terventions in the prevention of perineal trauma in parturients: a systematic Obstet. Gynaecol. doi:10.1080/0167482X.2021.1923689.
review with meta-analysis. Eur. J. Obstet. Gynecol. Reprod. Biol. 28, 100–111. OECD, 2019. Safe acute care – obstetric trauma. Health at a Glance 2019: OECD In-
doi:10.1016/j.ejogrb.2023.02.008. dicators. OECD Publishing, Paris.
Dahlen, H.G., Homer, C.S.E., Cooke, M., et al., 2007a. Perineal outcomes and mater- OECD, 2021. Health at a Glance 2021: OECD Indicators. OECD Publishing, Paris
nal comfort related to the application of perineal warm packs in the second doi:10.1787/ae3016b9-en.
stage of labor: a randomized controlled trial. Birth 34, 282–290. doi:10.1111/j. Pierce-Williams, R., Saccone, G., Berghella, V., et al., 2019. Hands-on versus hands-
1523-536X.20 07.0 0186.x. off techniques for the prevention of perineal trauma during vaginal delivery: a
Dahlen, H., 2007b. Reduction of perineal trauma and improved perineal comfort systematic review and meta-analysis of randomized controlled trials. J. Matern.
during and after childbirth: the Perineal Warm Pack Trial. http://hdl.handle.net/ Fetal Neonatal Med. doi:10.1080/14767058.2019.1619686.
10453/49236 Rodaki, E., Diamanti, A., Sarantaki, A., et al., 2022. The effects of perineal tears dur-
De Souza Caroci da Costa, A., Gonzalez Riesco, M.L., 2006. A comparison of “hands ing childbirth on women’s sex life. Maedica (Bucur) 17 (2), 297–305. doi:10.
off’’ versus “hands on’’ techniques for decreasing perineal lacerations during 26574/maedica.2022.17.2.297.
birth. J. Midwifery Womens Health 51 (2), 106–111. doi:10.1016/j.jmwh.2005.10. Rodrigues, S., Silva, P., Agius, A., et al., 2019. Intact perineum: what are the pre-
017. dictive factors in spontaneous vaginal birth? Mater. Sociomed. 31 (1), 25–30.
Demirel, G., Golbasi, Z., 2015. Effect of perineal massage on the rate of episiotomy doi:10.5455/msm.2019.31.25-30.
and perineal tearing. Int. J. Gynecol. Obstet. 131, 183–186. doi:10.1016/j.ijgo. Roper, J.C., Amber, N., Wan, O.Y.K., et al., 2020. Review of available national guide-
2015.04.048. lines for obstetric anal sphincter injury. Int. Urogynecol. J. 31 (11), 2247–2259.
Franchi, M., Parissone, F., Lazzari, C., et al., 2020. Selective use of episiotomy: what doi:10.10 07/s0 0192- 020- 04464- 5.
is the impact on perineal trauma? Results from a retrospective cohort study. Persson, M., Lindberg, I., Öhman, A., 2023. Regional and clinical guidelines for pre-
Arch. Gynecol. Obstet. 301, 427–435. doi:10.10 07/s0 0404- 019- 05404- 5. vention and care of obstetric anal sphincter injuries-a critical frame analysis.
Gimovsky, A., Berghella, V., 2022. Evidence-based labor management: second stage Midwifery 119, 103608. doi:10.1016/j.midw.2023.103608.
of labor (part 4). Am. J. Obstet. Gynecol. MFM 4 (2), 100548. doi:10.1016/j. Sultan, A.H., Kettle, C., 2007. Diagnosis of perineal trauma. In: Sultan, A.H.,
ajogmf.2021.100548. Thakar, R., Fenner, D.E. (Eds.), in Perineal and Anal Sphincter Trauma Diagno-
Goh, Y.P., Tan, P.C., Hong, J.G.S., et al., 2021. Combined massage and warm compress sis and Clinical Management. Springer Science + Business Media, London, UK,
to the perineum during active second stage of labor in nulliparas: a randomized pp. 13–19 vol. 2.
trial. Int. J. Gynaecol. Obstet. 155 (3), 532–538. doi:10.1002/ijgo.13613. Teixeira, C., Lorthe, E., Barros, H., 2022. Time trends in episiotomy and severe
Hong, J., Abdullah, N., Rajaratnam, R., et al., 2022. Combined perineal massage and perineal tears in Portugal: a nationwide register-based study. BMC Pregnancy
warm compress compared to massage alone during active second stage of labor Childbirth 22, 976. doi:10.1186/s12884- 022- 05314- 6.
in nulliparas: a randomised trial. J. Obstet. Gynecol. Reprod. Biol. 270, 144–150. Trochez, R., Waterfield, M., Freeman, R.M., 2011. Hands on or hands off the per-
doi:10.1016/j.ejogrb.2022.01.011. ineum: a survey of care of the perineum in labour (HOOPS). Int. Urogynecol. J.
Jiang, H., Qian, X., Carroli, G., et al., 2017. Selective versus routine use of epi- 22, 1279–1285. doi:10.10 07/s0 0192-011-1454-8776.
siotomy for vaginal birth. Cochrane Database Syst. Rev. (2), CD0 0 0 081 doi:10. Venugopal, V., Deenadayalan, B., Maheshkumar, K., et al., 2022. Perineal massage
10 02/14651858.CD0 0 0 081.pub2. for prevention of perineal trauma and episiotomy during labor: a systematic
Kiremitli, S., Kiremitli, T., Uluğ, P., et al., 2022. Comparison of obstetric outcomes review and meta-analysis. J. Fam. Reprod. Health 16 (3), 162–169. doi:10.18502/
and morbidity in nulliparous pregnant women who received antenatal and in- jfrh.v16i3.10575.
trapartum perineal massage. Pelviperineology 41 (1), 17–22. doi:10.34057/PPj. World Health Organization (WHO) 2018. Intrapartum Care for a Positive
2022.41.01.2022- 2- 1. Childbirth Experience.https://apps.who.int/iris/bitstream/handle/10665/260178/
Laine, K., Yli, B.M., Cole, V., et al., 2022. European guidelines on perinatal care- Peri- 9789241550215-eng.pdf.
partum care Episiotomy. J. Maternal-Fetal Neonatal. Med. 35 (25), 8797–8802. Zang, Y., Hu, Y., Lu, H., 2022. Effects of different techniques during the second stage
doi:10.1080/14767058.2021.2005022. of labour on reducing perineal laceration: an overview of systematic reviews. J.
Clin. Nurs. 00, 1–18. doi:10.1111/jocn.16276.