Research www. AJOG.

org

OBSTETRICS
Selective vs routine midline episiotomy for the prevention
of third- or fourth-degree lacerations in nulliparous women
Alexander Rodriguez, MD; Edgar A. Arenas, MD; Alba L. Osorio, MD; Omar Mendez, MD; John J. Zuleta, MD, MSc

OBJECTIVE: The purpose of this study was to determine whether se-
lective midline episiotomy contributes to the prevention of third- or
fourth-degree perineal lacerations.
STUDY DESIGN: A randomized controlled clinical trial was performed
with 446 nulliparous women with deliveries after 28 weeks of preg-
nancy. Patients were randomized to undergo either routine episiotomy
or selective episiotomy. In the selective episiotomy group, episioto-
mies were performed only in cases of imminent lacerations, fetal dis-
tress, or forceps delivery.
RESULTS: In the group of 223 patients who underwent routine epi-
siotomy, 32 (14.3%) had third- or fourth-degree perineal lacera-
tions, as compared to 15 (6.8%) in the group of 222 patients un-
dergoing selective episiotomy (relative risk, 2.12; 95% confidence
interval, 1.18-3.81). Only reduction in third-degree lacerations was
significant when analyzed separately. Moreover, periurethral, labia
minora, and superficial vaginal lacerations were significantly more
frequent in the selective episiotomy group.
CONCLUSION: The policy of performing selective midline episiotomy
in nulliparous patients results in a reduction in the risk of third-degree
perineal lacerations.
Key words: episiotomy, perineal laceration

Cite this article as: Rodriguez A, Arenas EA, Osorio AL, et al. Selective vs routine midline episiotomy for the prevention of third- or fourth-degree lacerations in
nulliparous women. Am J Obstet Gynecol 2008;198:285.e1-285.e4.

H istorically, the liberal use of epi-

siotomy was common when caring
for nulliparous women in order to pre-
vent complications for the mother (deep
perineal lacerations, relaxation of the
pelvic floor) as well as for the fetus (fetal
hypoxia, shoulder retention).1,2 With
the publication of several descriptive and
analytical papers assessing the effective-
ness of episiotomy in preventing deep
perineal lacerations (third and fourth
degree), it was determined that, para-
doxically, instead of playing a protective
role for those events, episiotomy actually
favored their occurrence.3 This creates
concerns about the associated morbid-
ity, including pain, bleeding, infection,
From the Department of Obstetrics and
Gynecology, San Vicente de Paul Hospital,
Antioquia University Medical School,
Medellín, Colombia.
Received Nov. 16, 2006; revised Oct. 23,
2007; accepted Nov. 5, 2007.
Reprints: Alexander Rodriguez, Calle 64 #40,
51D, San Vicente de Paul Hospital, Medellin,
Antioquia, Colombia. jorgear12@gmail.com.
0002-9378/$34.00
© 2008 Mosby, Inc. All rights reserved.
doi: 10.1016/j.ajog.2007.11.007
rectovaginal fistulas, fecal or flatus in-
continence, and intercourse disrup-
tion.3-10 Most of these studies assessed
mediolateral episiotomies10 and several
schools, including ours, have tradition-
ally performed and taught midline
episiotomy.2
There is a shortage of world literature
regarding midline episiotomy. In a re-
cent metaanalysis of the Cochrane col-
laboration, of the 6 studies included,
only 1 of them described the use of a
midline technique.11 The remaining
studies used the mediolateral technique,
including a collaborative Argentine
study with the largest number of pa-
tients.10 The systematic review in 2005
included basically the same studies
found in the Cochrane analysis.12
We performed a prospective, random-
ized clinical trial to determine whether
selective midline episiotomy is associ-
ated with the prevention of third- or
fourth-degree perineal lacerations in
nulliparous women.
M ATERIALS AND M ETHODS
The patients were nulliparous women
with pregnancies more than 28 weeks of
gestation who had vaginal deliveries.
Women with multiple pregnancies were
excluded, as were patients with breech
presentations and those who did not sign
the informed consent or refused to par-
ticipate in the study.
The randomized study was performed
at San Vicente de Paul teaching hospital,
a high complexity care level institution
in Medellín-Antioquia with an average
of 1200 deliveries per year; and the Hos-
pital del Sur, a mid-complexity care level
institution caring for an average of 800
deliveries per year in Itagui-Antioquia.
Ralloc software (Boston College Depart-
ment of Economics, Boston, MA) was
used to create a random sequence of num-
bers in blocks with 2, 4, and 6 size permu-
tations. Informed consent was obtained
when it was determined that a patient ful-
filled the requirements to enter the study;
the consent form to enroll in the study was
processed during the first stage of labor. All
the patients in this study were informed of
the benefits and risk of the episiotomy. De-
spite the evidence available in the litera-
ture, in our country, midline episiotomy is
still a medical procedure commonly per-
formed in nulliparous patients. For this
reason we still considered the study to be
ethical. An increase in the possibility of

MARCH 2008 American Journal of Obstetrics & Gynecology 285.e1

Research Obstetrics www.AJOG.org

erations. In order to achieve an 80%

TABLE 1 power for finding such differences with
General characteristics of the study population an alpha error of 5%, it was estimated
Routine episiotomy Selective episiotomy that 203 women were required for each
n ⴝ 223 n ⴝ 222 group. To compensate for the possibility
Age (y) 19.7 ⫾ 4.0 19.8 ⫾ 4.1 of patients leaving the study, we then
..............................................................................................................................................................................................................................................
Gestational age (wk) 38.0 ⫾ 2.1 38.0 ⫾ 2.1 projected a 10% increase in the sample
..............................................................................................................................................................................................................................................
Birth weight (g) 2936 ⫾ 470 2899 ⫾ 481
size to finally include 223 patients in each
.............................................................................................................................................................................................................................................. group.
Mean head circumference (cm) 33.3 ⫾ 1.5 33.1 ⫾ 1.5 Patients assigned to the selective epi-
..............................................................................................................................................................................................................................................
Oxytocin use 142 (63.7) 149 (67.1) siotomy group underwent the procedure
..............................................................................................................................................................................................................................................
Forceps use 4 (1.8) 3 (1.4) only in cases of forceps delivery, fetal dis-
..............................................................................................................................................................................................................................................
Epidural use 40 (17.9) 37 (16.7)
tress, or shoulder dystocia or when the
.............................................................................................................................................................................................................................................. operator considered that a severe lacera-
Prolonged delivery 5 (2.2) 13 (5.9) tion was impending and could only be
..............................................................................................................................................................................................................................................
Data are presented as mean ⫾ standard deviation for continuous variables and as n (percent) for categorical variables. avoided by performing an episiotomy.
Rodriguez. Selective vs routine midline episiotomy. Am J Obstet Gynecol 2008. This decision was made by the treating
physician. All the patients in the routine
third- and fourth-degree lacerations and midline of the perineum, from the in- episiotomy group underwent the proce-
the lack of evidence regarding benefits for troitus to the rectum, including skin, dure at the time the fetal head was dis-
the pelvic floor after the making of the epi- subcutaneous cellular tissue, superficial tending the introitus. The deliveries were
siotomy was also discussed. The patients fascia, and perineal muscle, followed by a assisted by obstetrics and gynecology
were taken to the delivery room where they 4-6 cm incision of the vaginal mucosa, residents or by the obstetrician on call. In
received routine care. Upon admission under local anesthesia with lidocaine our city, the vacuum is not employed.
during the second stage of labor, patients without epinephrine. After the delivery, the perineum was
were assigned either to the routine episiot- The primary outcome of severe lacer- examined to determine the presence of
omy or the selective episiotomy group, de- ation to perineal tissues was defined as a lateral, anterior, superficial or deep vag-
pending of the basis of the randomization third-degree laceration when the extent inal lacerations, and posterior lacera-
sequence kept at the institution. of the lesion included the external anal tions classified as first-degree to fourth-
The ethics board of the San Vicente de sphincter totally or partially, and fourth- degree according to the currently
Paul University Hospital, which works as degree laceration when the rectal mu- accepted classification.1
the institutional review board (IRB), cosa was involved.
granted its approval. The objective was to find a significant
Statistical analysis
The episiotomy consisted of an inci- reduction from 15% down to 6% in the
The distribution of continuous variables
sion of approximately 4 cm over the incidence of third- or fourth-degree lac-
was assessed using the Kolmogorov–
Smirnov test. Continuous variables with
TABLE 2 a normal distribution were summarized
Perineal injury outcomes of routine vs selective episiotomy using the mean plus standard deviation.
Routine Selective The Student t test was used for compar-
episiotomy episiotomy ing population characteristics for the
n ⴝ 223 n ⴝ 222 RR (95% CI) P value normally distributed continuous vari-
First-degree laceration n/a 30 (13.5) n/a n/a ables. Categorical variables were ex-
..............................................................................................................................................................................................................................................
Second-degree laceration n/a 33 (14.9) n/a n/a pressed as percentages. The ␹2 test was
..............................................................................................................................................................................................................................................
Third- or fourth-degree laceration 32 (14.4) 15 (6.8) 2.12 (1.18-3.81) .007 used for comparing categorical variables.
.............................................................................................................................................................................................................................................. Results are presented in the form of rel-
Third-degree laceration 22 (9.9) 10 (4.5) 2.19 (1.06-4.52) .03
.............................................................................................................................................................................................................................................. ative risk (RR) with their respective 95%
Fourth-degree laceration 10 (4.5) 5 (2.3) 1.99 (0.69-5.7) .192 confidence interval (95% CI). The anal-
..............................................................................................................................................................................................................................................
Periurethral laceration 6 (2.7) 25 (11.3) 0.24 (0.10-0.57) ⬍.001 ysis was performed using the SPSS 11.5
..............................................................................................................................................................................................................................................
Labia minora laceration 6 (2.7) 21 (9.5) 0.28 (0.12-0.69) .03 statistical software package (SPSS, Chi-
.............................................................................................................................................................................................................................................. cago, IL).
Superficial vaginal laceration 5 (2.2) 23 (10.4) 0.22 (0.08-0.56) .001
..............................................................................................................................................................................................................................................
Deep vaginal laceration 6 (2.7) 7 (3.2) 0.85 (0.29-2.5) .77
.............................................................................................................................................................................................................................................. R ESULTS
n/a, not applicable.
The numbers within parentheses represent percentages and to the left the numerical value for the number of cases.
Between February 2002 and November
Rodriguez. Selective vs routine midline episiotomy. Am J Obstet Gynecol 2008.
2004, 446 patients were randomly as-
signed to 2 groups of 223 patients each.

285.e2 American Journal of Obstetrics & Gynecology MARCH 2008

www.AJOG.org
One patient was excluded from the selec-
tive episiotomy group because she failed
to fulfill the inclusion criteria (gesta-
tional age ⬍28 weeks). All the 223 pa-
tients in the routine episiotomy group
underwent midline episiotomy, as well
as 54 patients in the selective episiotomy
group. During the same study period,
279 deliveries of nulliparous women
were cared for at the Southern Hospital
and 167 at the St. Vincent Paul Hospital.
There are no statistically significant dif-
ferences in any of the patient character-
istics between the 2 arms (Table 1).
In the routine episiotomy group 22
(9.9%) developed third-degree lacera-
tions compared to 10 (4.5%) of patients
in the selective episiotomy group (RR,
2.19; 95% CI, 1.06-4.52). There was no
significant difference between the fre-
quency of fourth-degree lacerations with
4.5% and 2.3% in the routine vs selective
groups (Table 2).
Of the third- and fourth-degree lacer-
ations seen in the selective episiotomy
group, 86.6% (13/15) occurred in
women in whom episiotomy was per-
formed. Of the 168 women who did not
undergo the procedure in the selective
episiotomy group, only 2 (1.19%) pa-
tients had third- or fourth-degree
lacerations.
Fifty-four patients (24.3%) in the se-
lective episiotomy group underwent the
procedure. The main reason to perform
the procedure was: impending laceration
46 patients (85.2%), fetal distress 4 pa-
tients (7.4%), forceps delivery 3 patients
(5.5%), other reasons 1 patient (1.7%).
Periurethral, labial minor, and super-
ficial vaginal lacerations were all signifi-
cantly more frequent in the selective
compared to routine episiotomy group
(Table 2).
C OMMENT
The use of routine episiotomy in nullip-
arous women is still commonly seen in
our Colombian hospitals, with a fre-
quency ranging between 51-61%. This
high frequency may be explained by the
fact that many patients and their treating
doctors are still convinced of the benefits
derived from the performing of such
technique.
The purpose of this study was to de-
velop the hypothesis that selective epi-
siotomy has a decreasing effect on the oc-
currence of third- or fourth-degree
perineal lacerations. We found that rou-
tine episiotomy was associated with
twice as many severe perineal lacerations
compared to the selective episiotomy.
This difference cannot be attributed to
any of the variables involved such as fetal
weight, gestational age, or head circum-
ference given the similarity between the 2
study groups. Supporting this risk of epi-
siotomy, the majority of the third- and
fourth-degree tears in the selective group
were in those women who received an
episiotomy. Consistent with the litera-
ture, the presence of other minor peri-
neal lacerations such as labia major and
minor, superficial vaginal, and periure-
thral lacerations was significantly higher
in the selective episiotomy group.1-4,7
The main recommendation based on re-
cent reviews is to not perform routine
episiotomy because it favors the occur-
rence of severe posterior perineal lacera-
tions and their associated morbidity.11,12
However, the results cannot be extrapo-
lated to a setting where the midline epi-
siotomy prevails because the majority of
reviewed trials used mediolateral epi-
siotomy. In retrospective studies midline
episiotomy is associated with a risk of
deep perineal damage, an additional ar-
gument to emphasize the relevance of
our results.13-17
Even though the process of patient al-
location within different groups was
done by computer, blinding was not per-
formed. However, different attending
physicians with no connection to the re-
search group were in charge of the clini-
cal care of the patients and the group in
which the patient was allocated was
known only during the second stage of
labor. However, the attending physicians
doing the classification of the degree of
perineal laceration had no connection
with the research group, as stated before;
the implemented strategy to decrease the
potential bias on this matter was to in-
clude the evaluation of the perineal lac-
eration by the resident who was then re-
quired to be in agreement with the
attending physician.
MARCH 2008 American Journal of Obstetrics & Gynecology
Obstetrics Research
The study included patients with pre-
term delivery because in our city the epi-
siotomy is recommended also in those
cases, with the purpose to decrease the
chance of fetal trauma. Our protocol ac-
cepted pregnant patients with more than
28 weeks of gestation; however, we did
not have cases presenting with less than
30 weeks. Our study is the first random-
ized clinical trial in the world comparing
the routine use of midline episiotomy vs
the use of selective midline episiotomy in
nulliparous patients in the prevention of
third- and fourth-degree lacerations. We
recommend performing selective epi-
siotomy in all birth centers when caring
for nulliparous women. f
ACKNOWLEDGMENTS
We thank the Hospital San Vicente de Paul of
Medellín.
REFERENCES
1. Cunningham F. Conduct of normal labor and
delivery, In: Cunningham FG, Mac Donald PC,
Leveno KJ. Williams obstetrics, 20th ed. Nor-
walk: Appleton & Lange; 1997. p. 315-9.
2. Woolley RJ. Benefits and risk of episiotomy: a
review of the English-language literature since
1980. Part I. Obstet Gynecol Survey 1995;
50:806-20.
3. Woolley RJ. Benefits and risk of episiotomy: a
review of the English-language literature since
1980. Part II. Obstet Gynecol Survey 1995;
50:820-35.
4. Thacker S, Banta D. Benefits and risks of
episiotomy: an interpretative review of the En-
glish language literature, 1960-1980. Obstet
Gynecol Survey 1983;38:322-34.
5. Nygaard I. Anal incontinence after anal
sphincter disruption: a 30 year retrospective co-
hort study. Obstet Gynecol 1997;89:896-901.
6. Trorp JM. Selected use of midline episiot-
omy: effect on perineal trauma. Obstet Gynecol
1987;70:260-2.
7. Ecker JL. Is there a benefit to episiotomy at
operative vaginal delivery? Observations over
ten years in a stable population. Am J Obstet
Gynecol 1997;176:411-4.
8. Crawford LA. Incontinence following rupture
of the anal sphincter during delivery. Obstet Gy-
necol 1993;82:527-31.
9. Walker J. Epidural anesthesia, episiotomy
and obstetric laceration. Obstet Gynecol
1991;77:668-71.
10. Argentine episiotomy trial collaborative
group. Routine vs selective episiotomy: a ran-
domized controlled trial. Lancet 1993;
342:1517-8.
11. Carroli G, Belizan J. Episiotomy for vaginal
birth. Cochrane Database Syst Rev 2005;
CD000081.
285.e3
Research Obstetrics www.AJOG.org

12. Hartmann K, Viswanathan M, Palmieri R,
Gartlehner G, Thorp J, Lohr K. Outcomes of
routine episiotomy: a systematic Review. JAMA
2005;293:2141-8.
13. Signorello LB. Midline episiotomy and anal
incontinence: retrospective cohort study. BMJ
2000;320:86-90.
14. Shiono P, Klebanoff MA, Carey JC. Midline
episiotomies: more harm than good? Obstet
Gynecol 1990;75:765-70.
15. Green JR, Soohoo SL. Factors asso-
ciated with rectal injury in spontane-
ous deliveries. Obstet Gynecol 1989;73:
732-8.
16. Gass MS, Dunn C, Stys SJ. Effect of
episiotomy on the frequency of vaginal out-
let laceracions. J Reprod Med 1986;31:
240-4.
17. Coats P, Chan K, Wilkins M, Beard R. A
comparison between midline and mediolateral
episiotomies. BJOG 1980;87:408-12.

285.e4 American Journal of Obstetrics & Gynecology MARCH 2008