Episiotomy, operative vaginal delivery, and significant perineal

trauma in nulliparous women

Julian N. Robinson, MD, Errol R. Norwitz, MD, Amy P. Cohen, BA, Thomas F. McElrath, MD, and
Ellice S. Lieberman, MD
Boston, Massachusetts

OBJECTIVES: The aim of this study was to determine whether choice of obstetric instrument at operative
vaginal delivery is associated with any differences in the rate of significant perineal trauma and whether this
rate is modified by the use of episiotomy.
STUDY DESIGN: The occurrence of significant perineal trauma among 323 consecutive operative vaginal
deliveries was evaluated according to type of instrument used and performance of episiotomy. These findings
were compared with spontaneous vaginal deliveries during the same period.
RESULTS: Among forceps deliveries the use of episiotomy was not associated with a difference in the oc-
currence of significant perineal trauma (55% vs 46%; relative risk, 1.2; 95% confidence interval, 0.8-1.9).
Among vacuum extraction deliveries an increased rate of such trauma was noted when episiotomy was used
(34.9% vs 9.4%; relative risk, 3.7; 95% confidence interval, 1.2-11.2). There was no difference in the rate of
significant perineal trauma according to type of forceps used. In a logistic regression analysis forceps deliv-
ery with or without episiotomy was associated with an increase of >10-fold in the rate of significant perineal
trauma with respect to vacuum extraction deliveries without episiotomy.
CONCLUSIONS: Our data suggest that in forceps delivery neither the type of forceps nor episiotomy influ-
ences the risk of significant perineal trauma. When vacuum extraction delivery is performed, the use of epi-
siotomy is associated with a higher risk of significant perineal trauma. (Am J Obstet Gynecol 1999;181:1180-4.)

Key words: Episiotomy, operative vaginal delivery, perineal trauma

When operative vaginal delivery is indicated in the sec-
ond stage of labor, both the choice of obstetric instru-
ment and the decision whether to perform an episiotomy
may affect the potential for maternal perineal injury.
Short-term complications of perineal injury include pain,
infection, and hemorrhage. Long-term effects include
dyspareunia, cosmetic deformity, and incontinence of fla-
tus or feces. These sequelae, particularly the last, may
have a profound impact on women’s health and self-
image.
Perineal trauma at childbirth involving the anal
sphincter (third- and fourth-degree lacerations) is associ-
ated with a much greater likelihood of incontinence than
is delivery with an intact perineum.1 Even when tears to
the anal sphincter are recognized and repaired immedi-
ately after delivery, almost half the affected women con-
tinue to have symptoms at 7 weeks post partum.1 Indeed
From the Department of Obstetrics and Gynecology, Brigham &
Women’s Hospital.
Presented at the Nineteenth Annual Meeting of the Society for Maternal-
Fetal Medicine, San Francisco, California, January 18-23, 1999.
Reprint requests: Julian N. Robinson, MD, Division of Maternal-Fetal
Medicine, Department of Obstetrics and Gynecology, Brigham &
Women’s Hospital, 75 Francis St, Boston, MA 02115.
Copyright © 1999 by Mosby, Inc.
0002-9378/99 $8.00 + 0 6/6/102313
women with intrapartum rupture of the anal sphincter
may still have related problems 2 decades later.2 In re-
cent years the issue of perineal trauma and its conse-
quences in later life has received increased attention
among both parturients and providers of maternity ser-
vices. A recent survey in London3 reported that 17%
(33/206) of female obstetricians would prefer elective
cesarean delivery to vaginal delivery for an uncompli-
cated singleton pregnancy, with 88% of these physicians
citing fear of perineal damage as the main reason for
their preference. This study was carried out to determine
whether choice of obstetric instrument at operative vagi-
nal delivery is associated with a difference in the rate of
severe perineal trauma and for each type of instrument
whether this rate is modified by the use of episiotomy.
Material and methods
We reviewed the medical records of all 5469 women
who were delivered at Brigham and Women’s Hospital
from December 1, 1994, through July 31, 1995. The pa-
tient population at this hospital consists of women who
receive their obstetric care through private practices, a
large health maintenance organization, and a hospital-
based residency practice (including a maternal-fetal
medicine practice). These analyses were limited to non-
diabetic nulliparous women at ≥36 weeks’ gestation who

1180
Volume 181, Number 5, Part 1 Robinson et al 1181
Am J Obstet Gynecol

Table I. Clinical characteristics according to severe perineal trauma

Third- or fourth-degree No third- or fourth-degree
laceration (n = 276) laceration (n = 1666)

Characteristic No. % No. % Statistical significance

Maternal age*
≤21 y 19 6.9 288 17.3
22-34 y 216 78.3 1159 69.7
>34 y 41 14.9 217 13.0 P = .001
Race
Black 12 4.3 212 12.7
White 205 74.3 1119 67.1
Other 59 21.4 335 20.1 P = .001
Welfare parturient 34 12.3 272 16.3 P = .9
Birth weight ≥4 kg† 46 16.7 118 7.1 P = .001
Oxytocin
For induction 67 24.3 405 24.3
For augmentation 134 48.6 644 38.7
None 75 27.2 617 37.0 P = .002

*Data not available for 2 mothers without laceration.

†Data not available for 2 neonates without maternal laceration.

had a singleton pregnancy with a cephalic presentation
and underwent a successful vaginal delivery. Women who
underwent a medically indicated elective operative vagi-
nal delivery (such as parturients with cardiac disease)
were excluded because such conditions could affect the
management of the second stage of labor. The project
was approved by the human research committee at the
hospital. Data related to labor and delivery were ob-
tained from the medical records by trained abstractors.
All cases of operative vaginal delivery or births in which a
third- or fourth-degree laceration had been recorded
were reviewed by a panel of 2 obstetricians to verify the
accuracy of the data.
This study examined the association of mode of vagi-
nal delivery with occurrence of significant perineal
trauma. Perineal trauma was considered to be significant
if it involved the maternal rectal sphincter; that is, it was a
third- or fourth-degree obstetric laceration. For the pur-
poses of this study, third-degree lacerations were defined
as perineal lacerations that involved the external anal
sphincter and fourth-degree lacerations were defined as
lacerations that extended through the rectal mucosa to
expose the lumen of the rectum.
In our institution episiotomy is not performed rou-
tinely. The type of episiotomy performed is almost exclu-
sively midline. We reviewed all the charts in which opera-
tive vaginal delivery was listed as being performed and all
charts in which a third- or fourth-degree laceration was
noted. The proportion of episiotomies that were midline
was 97% (451/464). The results we report therefore
largely reflect the consequences of midline episiotomy.
The occurrence of third- or fourth-degree laceration
among operative vaginal deliveries of nulliparous par-
turients was also evaluated according to type of instru-
ment (Simpson forceps, Tucker-McLane forceps, or vac-
uum extractor) and the use of midline episiotomy.
Operative vaginal delivery was at the discretion of the in-
dividual obstetrician.
Statistical significance for comparison of categoric vari-
ables was evaluated by means of the χ2 statistic. P < .05
was taken as statistically significant. A logistic regression
model was performed among women who underwent op-
erative vaginal delivery to evaluate the relative effects of
different techniques of operative vaginal delivery. The lo-
gistic regression controlled for the potentially confound-
ing effects of indication for operative vaginal delivery, sta-
tion at application of forceps or vacuum (modeled as an
ordinal variable), fetal position, fetal weight ≥4 kg, occur-
rence of shoulder dystocia, oxytocin use, race, physician
training status, and maternal age. Method of delivery was
modeled as 3 indicator variables—forceps with epi-
siotomy, forceps without episiotomy, and vacuum extrac-
tion with episiotomy. The reference group was women
who underwent vacuum extraction delivery without epi-
siotomy. Indication for operative vaginal delivery was
modeled as 3 indicator variables—fetal distress, maternal
exhaustion, and occipitoposterior position. Failure to
progress was the reference group.
Results
There were 1942 nulliparous women with uncompli-
cated term pregnancy, singleton fetus in cephalic presen-
tation, and a successful vaginal delivery. Among these
women, 472 (24.3%) had labor induced with oxytocin
and 778 (40.1%) had labor augmented with oxytocin.
There were 1619 (83.4%) normal spontaneous vaginal
deliveries, 161 (8.3%) vacuum-assisted vaginal deliveries,
and 162 (8.3%) forceps deliveries. In all, 924 women
(47.6%) underwent episiotomy.
The overall rate of third-degree laceration was 10.8%
1182 Robinson et al November 1999
Am J Obstet Gynecol

Table II. Rate of significant perineal trauma according to type of delivery and use of episiotomy
Relative risk
Episiotomy (n = 924) No episiotomy (n = 1018)
Calculated 95% Confidence
Type of delivery No. % No. % value interval

Spontaneous vaginal delivery 99/657 15.1 42/962 4.4 3.4 2.4-4.9

Vacuum extraction delivery 45/129 34.9 3/32 9.4 3.7 1.2-11.2
Forceps delivery 76/138 55.1 11/24 45.8 1.2 0.8-1.9

Table III. Logistic regression analysis comparing severe perineal trauma associated with forceps with that associated with
vacuum extraction delivery
Variable Odds ratio 95% Confidence interval

Method of delivery*
Forceps with episiotomy 15.8 3.4-73.6
Forceps without episiotomy 11.0 1.9-62.7
Vacuum extraction with episiotomy 6.8 1.4-31.8
Indication for operative vaginal delivery†
Fetal distress 0.6 0.3-1.0
Maternal exhaustion 0.6 0.3-1.3
Occipitoposterior position 1.3 0.6-2.6
Station at delivery 0.9 0.6-1.4
Weight >4 kg 1.7 0.8-3.5
Shoulder dystocia 0.7 0.2-2.0
Use of oxytocin 0.9 0.5-1.7
Maternal age <21 y 1.5 0.4-5.1
Race‡
Black 0.6 0.2-1.9
Other race 1.2 0.6-2.5
Resident physician 1.5 0.7-3.0

*Reference group was vacuum extraction with no episiotomy.

†Reference group was failure to progress.
‡Reference group was white.

(n = 210) and that of fourth-degree laceration was 3.2%
(n = 63). The rates of significant perineal laceration were
8.7% (141/1619) with spontaneous vaginal delivery,
29.8% (48/161) with vacuum extraction delivery, and
53.7% (87/162) with forceps delivery. The rates of signif-
icant perineal laceration were 23.8% (220/924) when an
episiotomy was performed and 5.5% (56/1018) when it
was not.
The characteristics of the women with and without
third- or fourth-degree laceration are compared in Table
I. The rate of third- or fourth-degree laceration was lower
among women <21 years of age, among black women,
and among women who received no oxytocin during
labor. The rate of significant perineal trauma was higher
among women whose babies weighed ≥4 kg.
We then evaluated the association of episiotomy with
significant perineal laceration separately for sponta-
neous vaginal deliveries, forceps deliveries, and vacuum
extraction deliveries (Table II). The rate of episiotomy
was 40.6% (657/1619) among women with spontaneous
vaginal delivery, 80.1% (129/161) among women with
vacuum extraction delivery, and 85.2% (138/162)
among women with forceps delivery. There was a sub-
stantially higher rate of third- or fourth-degree laceration
when an episiotomy was performed at spontaneous vagi-
nal delivery (15.1% vs 4.4%; P = .001) and at vacuum ex-
traction delivery (34.9% compared with 9.4%; P = .005).
In contrast, when forceps were used, there was no signifi-
cant difference in the rate of third- or fourth-degree lac-
erations when episiotomy was performed (55.1% vs
45.8%; P = .4). The rate of significant perineal trauma
was similar between Simpson forceps (56.9%, 29/51)
and Tucker-McLane forceps (53.4%, 55/103, P = .3).
There was no difference in type of forceps used accord-
ing to station.
A multiple logistic regression was performed on data
from the 323 women who underwent operative vaginal
deliveries to examine the effects of method of operative
vaginal delivery and episiotomy while potential con-
founding factors were controlled for (Table III). Because
our data indicated that the association of episiotomy with
significant perineal trauma differed according to
whether a forceps or vacuum extraction delivery was per-
formed, we modeled method of delivery with 3 indicator
variables—forceps delivery with episiotomy, forceps deliv-
ery without episiotomy, and vacuum extraction with epi-
Volume 181, Number 5, Part 1
Am J Obstet Gynecol
siotomy. The reference group was vacuum extraction de-
livery without episiotomy, the group with the lowest rate
of significant perineal trauma. In that model it was found
that forceps delivery with episiotomy (odds ratio, 15.8;
95% confidence interval, 3.4-73.6) and forceps delivery
without episiotomy (odds ratio, 11.0; 95% confidence in-
terval, 1.9-62.7) were associated with a >10-fold increase
in the rate of significant perineal trauma with respect to
vacuum extraction delivery without episiotomy. The use
of vacuum extraction with episiotomy was associated with
a 7-fold increase in significant perineal trauma with re-
spect to vacuum extraction alone (odds ratio, 6.8; 95%
confidence interval, 1.4-31.8). In that model the associa-
tions with significant perineal trauma of maternal age,
race, use of oxytocin, and birth weight were not statisti-
cally significant.
Comment
Vaginal delivery is the single biggest determinant of
damage to the maternal anal sphincter. Although this
damage can be occult,4 it is most apparent as a third- or
fourth-degree obstetric laceration. Obstetric anal sphinc-
ter rupture can have long-term effects in the form of
both urinary and fecal incontinence.5, 6 In a follow-up
study of 72 women who had obstetric anal sphincter dam-
age, 30 (42%) had fecal incontinence 2 to 4 years post
partum, 23 (32%) had urinary incontinence, and 13
(18%) had both.5 Incontinence of flatus has been re-
ported in 31% of women with a history of anal sphincter
damage at childbirth.6 Avoidance of severe perineal lac-
erations should therefore be an important objective of all
obstetric clinicians.
Our study found that significant perineal trauma was
in fact dependent on mode of delivery. In our study we
found rates of significant perineal laceration of 8.7%
with spontaneous vaginal delivery, 29.8% with vacuum
extraction delivery, and 53.7% with forceps delivery. It is
more interesting that this study suggests that the associa-
tion of episiotomy with significant perineal trauma dif-
fers between vacuum extraction and forceps deliveries.
Episiotomy was associated with an increased risk of signif-
icant laceration in vacuum extraction but not forceps de-
liveries. Because only a small percentage of women (<3%
of women with operative vaginal delivery or significant
perineal trauma) had mediolateral episiotomy per-
formed, these results largely reflect the effect of midline
episiotomy.
The overall rates of third-degree lacerations (10.8%)
and fourth-degree lacerations (3.2%) that we found are
similar to those reported for other institutions.7, 8 Legino
et al7 reported a rate of 10.7% for third-degree lacera-
tions and 6.4% for fourth-degree lacerations in a mixed
population of nulliparous and multiparous women.
Ecker et al8 in a California study found the rate for third-
Robinson et al 1183
degree lacerations to be 26.5% and that for fourth-de-
gree lacerations to be 5.4%. Although we found that the
rate of third- or fourth-degree lacerations among women
undergoing forceps delivery to be relatively high
(53.7%), similar rates are reported in the literature
where an analysis includes forceps delivery alone as a sub-
group.1, 9 Sultan et al1 reported a rate of laceration in-
volving the anal sphincter of 50% when forceps were
used. Coombs et al9 reported a somewhat lower rate of
31% for this clinical situation. In both these series the
populations included both nulliparous and multiparous
patients, whereas our population consisted of nulli-
parous women only. Coombs et al9 suggested that the
rates of third- and fourth-degree lacerations are higher
among nulliparous women, but they did not provide the
rate of this complication among nulliparous women un-
dergoing forceps delivery.
In our study lower rates of third- or fourth-degree lac-
eration were observed with vacuum extraction deliveries
(29.8%) than were observed with forceps deliveries
(57.3%). This finding is consistent with previous reports
in the obstetric literature.10, 11 We did not find the associ-
ation between station at operative vaginal delivery and
third- or fourth-degree laceration that has previously
been reported.9 However, our population included only
a small number of patients (n = 19) with outlet operative
vaginal delivery. We also report that overall midline epi-
siotomy increased the rate of third- or fourth-degree lac-
eration.12-14 We found that the effects of midline epi-
siotomy on significant perineal trauma may differ,
however, depending on whether it is used in combina-
tion with spontaneous vaginal delivery, vacuum extrac-
tion delivery, or forceps delivery. In our study episiotomy
at spontaneous vaginal delivery was associated with a
higher rate of third- or fourth-degree laceration (15.1%
vs 4.4%; P = .001). Similarly, with vacuum extraction de-
livery significant perineal trauma rates were higher when
episiotomy was performed (34.9% vs 9.4%; P = .005). In
contrast, when forceps delivery was performed, the use of
episiotomy was not associated with a substantial increase
in the rate of significant perineal trauma (55.1% vs
45.8%; P = .4). This may be because the greater rate of
perineal injury seen with forceps delivery overrides any
potential contribution to perineal damage from the mid-
line episiotomy. It is important to interpret our results
with some caution, however, because of the small num-
ber of women who underwent forceps deliveries (n = 24)
and vacuum extraction deliveries (n = 32) and did not
have an episiotomy. As in any observational study, we can-
not rule out the possibility of residual confounding vari-
ables; for example, our database did not include the
length of the perineal body.
The finding of procedure-dependent third- and
fourth-degree extension rates may be useful to the obste-
1184 Robinson et al
trician at the time of operative vaginal delivery. In an ef-
fort to reduce the rate of third- and fourth-degree lacera-
tion the accoucheur may elect not to cut an episiotomy
when performing a vacuum-assisted delivery. If, however,
it is the judgment of the obstetrician that an episiotomy
would be of benefit (such as in the setting of fetal dis-
tress) during a forceps delivery, our study suggests that a
midline episiotomy can be cut without substantially in-
creasing the rate of significant perineal trauma.
Many different designs of obstetric forceps are in clini-
cal use. These numerous designs can be separated into 2
broad classifications, classical and rotational. We did not
evaluate the effects of rotational forceps in this study be-
cause of the small number (n = 8) of cases in which they
were used. In general, there are 2 types of classical in-
struments, those in which the shank overlaps (Tucker-
McLane type) and those in which it does not (Simpson
type).15 We hypothesized that the separated shank of the
Simpson forceps would cause a greater rate of severe per-
ineal trauma than would the Tucker-McLane forceps be-
cause of the greater width of the shanks. However, we
found similar rates of significant perineal trauma for
Simpson forceps (29/51, 56.9%) and Tucker-McLane
forceps (55/103, 53.4%).
Forceps delivery was associated with a higher rate of
third- or fourth-degree obstetric laceration than was vac-
uum extraction delivery. If vacuum extraction delivery is
being carried out, electing to cut a midline episiotomy
would increase the likelihood of significant perineal
trauma. If obstetric indications necessitate forceps delivery
rather than vacuum extraction delivery, our data suggest
that performance of a midline episiotomy would not in-
crease the risk of third- or fourth-degree laceration. Finally,
the type of obstetric forceps used does not appear to affect
the rate of third- and fourth-degree obstetric laceration.
November 1999
Am J Obstet Gynecol
REFERENCES
1. Sultan AH, Kamm MA, Hudson CN, Bartram CI. Third degree
obstetric anal sphincter tears: risk factors and outcome of pri-
mary repair. BMJ 1994;308:887-91.
2. Haadam K, Gudmundsson S. Can women with intrapartum rup-
ture of anal sphincter still suffer after-effects two decades later?
Acta Obstet Gynecol Scand 1997;76:601-3.
3. Al-Mufti R, McCarthy A, Fisk NM. Survey of obstetricians’ per-
sonal preference and discretionary practice. Eur J Obstet
Gynecol Reprod Biol 1997;73:1-4.
4. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI.
Anal-sphincter disruption during vaginal delivery. N Engl J Med
1993;329:1905-11.
5. Tetzchner T, Sorensen M, Lose G, Christiansen J. Anal and uri-
nary incontinence in women with obstetric anal sphincter rup-
ture. Br J Obstet Gynaecol 1996;103:1034-40.
6. Nygaard IE, Rao SS, Dawson JD. Anal incontinence after anal
sphincter disruption: a 30-year retrospective cohort study.
Obstet Gynecol 1997;89:896-901.
7. Legino LJ, Woods MP, Rayburn WF, McCoogan LS. Third- and
fourth-degree tears, 50 years experience at a university hospital.
J Reprod Med 1988;33:423-6.
8. Ecker JL, Tan WM, Bansal RK, Bishop JT, Kilpatrick SJ. Is there a
benefit to episiotomy at operative vaginal delivery? Am J Obstet
Gynecol 1997;176:411-4.
9. Coombs CA, Robertson PA, Laros RK. Risk factors for third-de-
gree and fourth-degree perineal lacerations in forceps and vac-
uum deliveries. Am J Obstet Gynecol 1990;163:100-4.
10. Johanson RB, Rice C, Doyle M, Arthur J, Anyanwu L, Ibrahim J,
et al. A randomized prospective study comparing the new vac-
uum extractor policy with forceps delivery. Br J Obstet Gynaecol
1993;100:524-30.
11. Broekhuizen FF, Washington JM, Johnson F, Hamilton PR.
Vacuum extraction versus forceps delivery: indications and com-
plications, 1979 to 1984. Obstet Gynecol 1987;69:338-42.
12. Helwig JT, Thorp JM, Bowes WA. Does midline episiotomy in-
crease the risk of third- and fourth-degree lacerations in opera-
tive vaginal deliveries? Obstet Gynecol 1993;82:276-9.
13. Shiono P, Klebanoff MA, Carey JC. Midline episiotomies: more
harm than good? Obstet Gynecol 1990;75:765-70.
14. Labrecque M, Baillargeon L, Dallaire M, Tremblay A, Pinault JJ,
Gingras S. Association between median episiotomy and severe
perineal lacerations in primiparous women. CMAJ
1997;156:797-802.
15. Dennen PC. Dennen’s forceps deliveries. 3rd ed. Philadelphia:
FA Davis; 1989.