Repair Techniques for Obstetric Anal
Sphincter Injuries
A Randomized Controlled Trial
Ruwan J. Fernando, MD, MRCOG, Abdul H. Sultan, MD, FRCOG, Christine Kettle,
Simon Radley, MD, FRCS, Peter Jones, PhD, and P. M. S. O’Brien, MD, FRCOG
OBJECTIVE: To compare one-year outcomes of primary
overlap versus end-to-end repair of the external anal
sphincter after acute obstetric anal sphincter injury.
METHODS: Women who sustained third-degree (3b ⴝ
greater than 50% external anal sphincter thickness, 3c ⴝ
internal sphincter injury) or fourth-degree (including
anorectal epithelium) perineal tears were randomly allo-
cated to either immediate primary overlap or end-to-end
repair. They were prospectively followed up for 12
months postrepair with serial questionnaires. The pri-
mary outcome was fecal incontinence at 12 months.
Secondary outcomes were fecal urgency, flatus inconti-
nence, perineal pain, dyspareunia, quality of life, and
improvement of anal incontinence symptoms.
RESULTS: Thirty-two women were randomized to each
group. At 12 months, 24% (6/25) in the end-to-end and
none in the overlap group reported fecal incontinence (P
ⴝ .009, relative risk [RR] 0.07, 95% confidence interval
[CI] 0.00 –1.21, number needed to treat 4.2). Fecal ur-
gency at 12 months was reported by 32% (8/25) in the
end-to-end and 3.7% (1/27) in the overlap group (P ⴝ .02,
RR 0.12, 95% CI 0.02– 0.86, number needed to treat 3.6).
There were no significant differences in dyspareunia and
From the Academic Unit of Obstetrics and Gynecology, University Hospital of
North Staffordshire, Staffordshire, United Kingdom; Mayday University Hos-
pital, Croydon, Surrey, United Kingdom; Queen Elizabeth Hospital, Birming-
ham, United Kingdom; and Keele University, Staffordshire, United Kingdom.
Funding support provided by Rehabilitation and Medical Research Trust
(REMEDI), Bath, UK.
We acknowledge the contribution of the late Professor Richard Johanson, who
conceived and designed this study but died while the study was underway.
Findings of this study were presented at the Joint Meeting of the International
Continence Society and the International Urogynecological Association, Paris,
France, August 23–27, 2004.
Corresponding author: Ruwan J Fernando, MD, MRCOG, Subspecialty Trainee
in Urogynecology, Mayday University Hospital, Croydon, Surrey, CR7 7YE,
United Kingdom; e-mail: ruwanfernando@hotmail.com.
© 2006 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins.
ISSN: 0029-7844/06
VOL. 107, NO. 6, JUNE 2006
PhD,
quality of life between the groups. At 12 months, 20%
(5/25) reported perineal pain in the end-to-end and none
in the overlap group (P ⴝ .04, RR 0.08, 95% CI 0.00 –1.45,
number needed to treat 5). During 12 months, 16% (4/25)
in the end-to-end and none in the overlap group re-
ported deterioration of defecatory symptoms (P ⴝ .01).
CONCLUSION: Primary overlap repair of the external
anal sphincter is associated with a significantly lower
incidence of fecal incontinence, urgency, and perineal
pain. When symptoms do develop, they appear to remain
unchanged or deteriorate in the end-to-end group but
improve in the overlap group.
(Obstet Gynecol 2006;107:1261–8)
LEVEL OF EVIDENCE: I
O bstetric anal sphincter injury is the most com-
mon cause of anal incontinence among women
of childbearing age. The reported incidence of such
injury varies between 0.5% and 5%1 of vaginal deliv-
eries in centers where mediolateral episiotomy is
practiced. In a retrospective study of 2,858 vaginal
deliveries in the state of Michigan where midline
episiotomy was practiced, 17% of women sustained
anal sphincter injury.2 The reported incidence of anal
incontinence after obstetric anal sphincter injury is as
high as 67%.3 In the United Kingdom, anal inconti-
nence is believed to affect nearly 40,000 mothers
annually in the first year after birth and millions
worldwide.4
Anal incontinence incorporates a range of symp-
toms, including flatus incontinence, passive soiling,
and incontinence of solid or liquid stool that is a social
or hygienic problem.5 In addition, obstetric anal
sphincter injury can be associated with fecal urgen-
cy,1,6 rectovaginal fistula,7 perineal pain, and dyspa-
reunia.8 Because anal incontinence is a source of
embarrassment and a social taboo, many women do
not volunteer these symptoms but sadly “suffer in
OBSTETRICS & GYNECOLOGY 1261
silence.”9 Furthermore, anal incontinence caused by
sphincter injury has been reported to be associated
with very high cumulative costs for health services.10
Over the last decade, there has been a notable
increase in litigation related to obstetric anal sphincter
injury and its consequences.11
Anal sphincter injury sustained during childbirth
has been traditionally repaired by obstetricians in the
immediate postpartum period. However, it is con-
cerning that 25–59% of women have persistent symp-
toms despite primary repair.12 Poor understanding of
perineal anatomy and inadequate training in repair
techniques are possible reasons for the high incidence
of persistent symptoms.4 Furthermore, there is limited
and inconsistent information in the literature relating
to repair techniques, suture materials, antibiotics,
laxatives, and the management of subsequent preg-
nancies after anal sphincter injury.4,12
Traditionally obstetric anal sphincter injuries are
repaired by obstetricians as soon as possible after
birth by using the end-to-end technique to reapproxi-
mate the torn ends of the external anal sphincter with
either interrupted or figure-of-eight sutures.4 How-
ever, in cases of delayed or secondary anal sphincter
repair when women present with fecal incontinence,
colorectal surgeons prefer to reapproximate the dis-
rupted ends of the external anal sphincter by using
the overlap technique.4 Having observed colorectal
surgeons carrying out this procedure, Sultan et al13
demonstrated that it was feasible to use the overlap
technique for primary repair of the external anal
sphincter. Furthermore, Sultan et al13 reported that
the overlap technique, when compared with historical
controls (repaired with the end-to-end technique),
reduced anal incontinence from 40% to 8% and
persistent anal sphincter defects from 85% to 15%. We
carried out a comprehensive literature search on
MEDLINE (January 1966 to November 31, 1998),
EMBASE (January 1974 to November 31, 1998), and
SciSearch (January 1974 to November 31, 1998)
databases. The search terms used were “perin*,” “anal
sphincter AND tear*,” “rupture*,” “trauma,” “dam-
age,” “injur* AND labor,” “labour,” “birth,” “child-
birth,” “delivery,” “obstetric* AND tear*,” “rupture*,”
“injur*,” “damage,” “trauma.” In addition, we
searched conference proceedings of associations of
obstetrics and gynecology, surgery, and coloproctol-
ogy. No randomized controlled studies comparing
these 2 techniques were available at the commence-
ment of this study. The aim of our study was to
undertake a randomized controlled trial to compare
the overlap with the end-to-end method of primary
1262 Fernando et al Repair of Obstetric Anal Sphincter Injury
external anal sphincter repair performed immediately
after obstetric anal sphincter injury.
MATERIALS AND METHODS
This randomized controlled study took place between
December 1998 and November 2000 in a university
hospital in Staffordshire, United Kingdom. We clas-
sified obstetric anal sphincter injury according to the
recommendation made in the guidelines produced by
the Royal College of Obstetricians & Gynecologists14
and the International Consultation on Incontinence5
(Fig. 1). Only women who sustained grades 3b, 3c, or
4th-degree perineal tears were eligible to participate.
Women who sustained a 3a tear (⬍ 50% thickness of
the external anal sphincter torn) or a previous 3rd- or
4th-degree perineal tear were excluded from this
study. Ethical approval was granted by the North
Staffordshire Local Research Ethics Committee.
The study was designed as a parallel group
randomized controlled study with minimization for
parity, gestation, and mode of delivery using a cus-
tomized computer package. It was programmed to
minimize the possibility of unequal distribution of
confounding factors between the 2 groups, which
otherwise would have affected the outcome. Use of
“minimization” rather than random permuted blocks
for treatment allocation ensured that the 2 groups
were similar and that confounding factors were
evenly distributed. As stated by Pocock,15 the purpose
of minimization “is to balance the marginal treatment
totals for each level of patient factor.” The customized
computer randomization package, which was de-
signed by the Birmingham Clinical Trials Unit (Bir-
mingham, UK), was password-protected to ensure
concealment of treatment allocation. Participants
were randomly allocated to overlap or end-to-end
repair of the external anal sphincter immediately after
delivery, and they were blinded to the method of
suturing.
The primary outcome measure was fecal inconti-
nence at 12 months. Secondary outcome measures
were fecal incontinence at six weeks, three and six
months, fecal urgency, flatus incontinence, perineal
pain, dyspareunia at six weeks, three, six and 12
months, and improvement of anal incontinence
symptoms over the 12 month period following the
procedure.
Two clinicians who were trained in both tech-
niques carried out all of the repairs in the operating
theater under regional or general anesthesia and in
the lithotomy position as described by Sultan et al.13
All repairs were carried out within 3 hours of detect-
ing the anal sphincter injury.
OBSTETRICS & GYNECOLOGY

In the end-to-end technique, the torn anal epithe-
lium was repaired using interrupted 3-0 standard
polyglactin (Vicryl; Ethicon, Edinburgh, UK) sutures
with knots tied within the anal canal (Fig. 2). If the
internal anal sphincter was torn, it was repaired
separately with interrupted 3-0 polydioxanone (PDS;
Ethicon) sutures. The torn ends of the external anal
sphincter were approximated and repaired with 2-3
mattress sutures using 3-0 PDS sutures. The vaginal
mucosa and perineal muscles were repaired with
continuous nonlocking 2-0 rapidly absorbed polygla-
ctin (Vicryl Rapide; Ethicon) sutures, and the perineal
skin was closed with subcuticular 2-0 rapidly ab-
sorbed polyglactin (Vicryl Rapide) sutures.16
In the overlap technique, the torn anal epithelium
and internal anal sphincter were repaired as described
above using interrupted 3-0 standard polyglactin (Vi-
cryl) and interrupted 3-0 polydioxanone (PDS) su-
tures, respectively (Fig. 3). After identification of the
torn ends of the external anal sphincter, the outer
surface of the sphincter was mobilized from the
VOL. 107, NO. 6, JUNE 2006 Fernando et al
Fig. 1. Classification of obstetric
anal sphincter injury. 3a: Less than
50% thickness of external anal
sphincter damage. 3b: More than
50% thickness of external anal
sphincter damage with intact in-
ternal anal sphincter. 3c: Both ex-
ternal anal sphincter and internal
anal sphincter damage with intact
anal mucosa. 4: External anal
sphincter, internal anal sphincter
and anal mucosa damage. Repro-
duced with permission from Sul-
tan AH. Primary repair of obstetric
anal sphincter injury. In: Cardozo
L, Staskin D, editors. Textbook of
Female Urology and Urogynecol-
ogy. London, UK: Informa Health-
care; 2005.
Fernando. Repair of Obstetric Anal
Sphincter Injury. Obstet Gynecol
2006.
surrounding tissue if necessary. If the external anal
sphincter was incompletely torn (⬎ 50%), the remain-
ing fibers were divided so that the torn ends could be
fully overlapped. The first row of sutures was inserted
about 1.5 cm from one side of the torn edge of
external anal sphincter (open arrow, Fig. 3) and
carried through to within 0.5 cm of the other edge of
the torn external anal sphincter. A second row of
sutures (small arrows, Fig. 3) was inserted to attach the
loose end of the overlapped muscle.13 The vaginal
mucosa and perineal skin were closed as described in
the end-to-end method.
All women received intra-operative intravenous
antibiotics and post operative oral antibiotics for
seven days and a bulking agent (Ispaghula husk) and
a stool softener (Lactulose) for 14 days post-opera-
tively. There were no dietary restrictions during the
post-operative period.
The women were followed with self-administered
questionnaires at 6 weeks, and 3, 6, and 12 months
after the repair. These questionnaires included the
Fig. 2. End-to-end repair tech-
nique of the external anal sphinc-
ter. Reproduced with permission
from Sultan AH. Primary repair of
obstetric anal sphincter injury. In:
Cardozo L, Staskin D, editors.
Textbook of Female Urology and
Urogynecology. London, UK: In-
forma Healthcare; 2005.
Fernando. Repair of Obstetric Anal
Sphincter Injury. Obstet Gynecol
2006.
Repair of Obstetric Anal Sphincter Injury 1263

modified Wexner anal incontinence scoring system,17
in which the highest score of 24 refers to complete
incontinence and 0 refers to complete continence. In
addition, the questionnaires contained the Fecal Incon-
tinence Quality of life Scale,18 which has 4 separate
scales: Life Style, Coping/Behavior, Depression/Self-
Perception, and Embarrassment. A second set of ques-
tionnaires was sent to those women who did not re-
spond, and they were also reminded by a telephone call.
When we were designing this study, there were
no pre-existing published randomized studies com-
paring primary overlap with end-to-end external anal
sphincter repair techniques. Therefore, the sample
size was based on a feasibility study conducted by
Sultan et al,13 in which 32 women had overlap repairs
of the external anal sphincter. The study reported that
8% of the participants experienced anal incontinence
in comparison with 47% in an historical group from a
previous study (n ⫽ 34) who underwent end-to-end
repair of the external anal sphincter.1 Based on the
above figures, the sample size was calculated by using
NCSS-PASS 6.0 (J. L. Hintze, Kaysville, UT) soft-
ware. A total sample size of 48 women (24 in each
arm) would allow detection of a change in the pri-
mary outcome of fecal incontinence from 47% to 8%,
with greater than 90% power and 2-sided 5% signifi-
cance. To accommodate participants lost to follow-up
at 12 months, we planned to recruit at least 60 women
before the end of the recruitment period (November
30, 2000).
The primary statistical analyses were carried out
on an “intention-to-treat” basis. All data were initially
entered in a customized Microsoft 1997 Excel data-
base and then imported into NCSS (NCSS 2001) and
StatXact 4 (StatXact 4 Cytel, Cambridge, MA) for
statistical analysis. StatXact 4 is a special statistical
program designed to calculate exact P values when
the frequencies are small and the ␹2 significance test
1264 Fernando et al Repair of Obstetric Anal Sphincter Injury
Fig. 3. Overlap repair technique of
the external anal sphincter. Repro-
duced with permission from Sul-
tan AH. Primary repair of obstetric
anal sphincter injury. In: Cardozo
L, Staskin D, editors. Textbook of
Female Urology and Urogynecol-
ogy. London, UK: Informa Health-
care; 2005.
Fernando. Repair of Obstetric Anal
Sphincter Injury. Obstet Gynecol
2006.
was carried out on binary and nominal data (exact P
values are quoted to allow for small frequencies). The
Mann-Whitney test was used for ordered response
categories. Relative risks were calculated with Review
Manager 4.2.7 software (Cochrane Collaboration,
Oxford, UK).
RESULTS
Of the 75 eligible women, 64 were randomized into
either the overlap or end-to-end technique of repair
(Fig. 4). Eight women declined to participate in the
trial. Three women were not offered participation in
the trial because the anal sphincter injury was diag-
nosed after undergoing manual removal of the pla-
centa in the operating theater. We considered it
unethical to recruit these women into the trial while
they were under regional anesthesia and undergoing
another procedure. None suffered from diabetes mel-
litus or irritable bowel symptoms, and none had a
prior history of anal incontinence. Fifty-two women
(81%) returned the 12-month questionnaire, 2 women
had left the area and could not be contacted, and 10
declined completing further questionnaires.
Twenty five women (78%) in the overlap group
and 27 women (81%) in the end-to-end group were
primiparous.
The number of women recruited with third-
degree (grades 3b and 3c) and fourth-degree perineal
tears were comparable in both groups (Table 1). In 4
of 25 women in the overlap group, some fibers of the
torn external anal sphincter were divided to achieve
complete overlap. Maternal age, period of gestation,
mode of delivery, birth weight, and head circumfer-
ences between the 2 intervention groups are shown in
Table 1.
The median operating time in the overlap group
was 38 minutes (range 15–70) compared with 28
minutes (range 15–55) in the end-to-end group (P ⫽
OBSTETRICS & GYNECOLOGY

Fig. 4. Flow of participants through the trial.
Fernando. Repair of Obstetric Anal Sphincter Injury. Obstet
Gynecol 2006.
.003, Mann-Whitney test). The median estimated
blood loss in the overlap group was 260 mL (range
100 – 600) compared with 100 mL (range 100 – 450) in
the end-to-end group (P ⫽ .05, Mann-Whitney test).
None of the 64 women required blood transfusion or
developed wound dehiscence, fistula formation, or
suture migration.
Compared with the overlap group, there was a
statistically significant proportion of women in the
end-to-end group who reported fecal incontinence at
12 months (Table 2). A statistically significant propor-
tion of women in the end-to-end group reported fecal
incontinence at 3 and 6 months (Table 3). In the
end-to-end group, there was an increase in fecal
incontinence symptoms from 6 weeks to 3 months,
with a decrease afterward. In Table 3 the decrease of
fecal incontinence from 3 to 6 months by 3 patients
was not attributed to the 3 women who did not return
the questionnaire.
VOL. 107, NO. 6, JUNE 2006
Compared with the overlap group, there was a
statistically significant proportion of women in the
end-to-end group who reported fecal urgency at 6 and
12 months. There was no difference in flatus inconti-
nence between the 2 groups at 3, 6, and 12 months
(Tables 2 and 3). There was no difference in the
median incontinence scores at 3 and 6 months. The
median incontinence score at 12 months in the end-
to-end group was 1 (range 0 –9) compared with 0
(range 0 –5) in the overlap group (P ⫽ .05, Mann-
Whitney test).
There was no significant difference in perineal
pain from 6 weeks up to 6 months (Table 3). How-
ever, at 12 months a significant proportion of women
in the end-to-end group reported perineal pain (Table
2). There was no significant difference in dyspareunia
between the 2 groups up to 12 months (Tables 2 and
3). There were no significant differences between the
2 techniques in terms of the mean Life Style, Coping/
Behavior, Depression/Self-Perception, and Embar-
rassment scales of the Fecal Incontinence Quality of
Life Scale.18
Compared with the end-to-end group, a signifi-
cant proportion of women (P ⫽ .01) in the overlap
group reported an improvement in symptoms of anal
incontinence from 6 weeks to 12 months (Table 4).
None of the patients in either group complained of
difficulty in bowel evacuation or dyschezia. There
was no correlation between anal incontinence and the
mode of delivery.
DISCUSSION
This randomized controlled study demonstrates that
primary overlap external anal sphincter repair, com-
pared with end-to-end repair, is associated with a
significantly lower incidence of fecal incontinence,
fecal urgency, and perineal pain at 12 months after
the procedure. Furthermore, the study also shows that
the overlap technique, compared with the end-to-end
method, is associated with a significant improvement
in fecal incontinence and fecal urgency at 12 months.
There are 2 randomized controlled studies com-
paring the primary overlap and end-to-end repair
techniques19,20. Fitzpatrick et al19 compared the same 2
techniques but followed participants to only 3 months
postrepair. Fitzpatrick and colleagues reported that
49% of the overlap group, compared with 58% of
women in the end-to-end group, had alteration in
fecal continence. Power calculation of this study was
based on the identification of a 30% “symptomatic”
difference between the 2 methods of repair, with a
90% probability. In addition there were methodolog-
ical differences between the study of Fitzpatrick et al19
Fernando et al Repair of Obstetric Anal Sphincter Injury 1265
Table 1. Obstetric Characteristics and Degree of Obstetric Anal Sphincter Injuries
Overlap End to End
(n ⴝ 32) (n ⴝ 32)
Median maternal age [y (range)] 30 (22–45) 30 (18–39)
Median period of gestation [wk (range)] 40.1 (36–40.5) 40.7 (32.4–42.2)
Mode of delivery [n (%)]
Normal 21 (65.6) 23 (71.9)
Ventouse 5 (15.6) 4 (12.5)
Forceps 6 (18.7) 5 (15.6)
Birth weight (g)
Mean (standard deviation) 3,515 (549.9) 3,727 (561.4)
Range 2,520–5,240 1,852–5,060
Head circumference (cm)
Mean (standard deviation) 34.7 (1.53) 35.0 (1.58)
Range 31–37 32–38
Degree of anal sphincter injury [n (%)]
3b degree 25 (78.1) 24 (75.0)
3c degree 2 (6.3) 2 (6.3)
4th degree 5 (15.6) 6 (18.7)

and our study, in that they included partial (grade 3a)
tears of the external anal sphincter in their random-
ization and did not identify and repair the internal
anal sphincter. The women were also prescribed
codeine-based constipating agents for 3 days, fol-
lowed by a laxative regimen for 5 days or until
defecation had occurred, which was different from
our postoperative management. Fitzpatrick and col-
leagues19 found no statistical difference in alteration in
fecal continence symptoms at 3 months between the
groups and therefore recommended the end-to-end
technique of external anal sphincter repair because of
its simplicity.
Garcia et al20 also performed a randomized trial
of the 2 techniques and took great care to include only
complete ruptures of the external anal sphincter (full
thickness 3b, 3c, and fourth-degree tears). There were
23 women in the end-to-end group and 18 in the
overlap group. Unfortunately, only 15 and 11 women,
respectively, returned for follow-up, which occurred
at only 3 months. No significant difference was found
between the groups in terms of symptoms of fecal
incontinence or transperineal ultrasound findings.
However, the authors acknowledged that the major
limitations of their study were that randomization was
inaccurate and that their study was underpowered.
Nevertheless, the findings of both of these studies
concur with our findings in that the continence scores
were not significantly different at 3 and 6 months, but
reached statistical significance at 12 months.
In terms of anal incontinence symptoms, we
found that a significant proportion of women in the
end-to-end group reported fecal incontinence at 3, 6,
and 12 months. Moreover, we found a significant
improvement of anal incontinence symptoms during
the 12-month period in the overlap group, with no
women showing deterioration (Table 4). In the end-
to-end group, 32% reported no change in anal incon-
tinence symptoms, whereas 16% reported symptom
deterioration between 6 weeks and 12 months. This
highlights the need for longer term follow-up. How-
ever, it remains to be established why the overlap
technique is associated with superior results. It could
be postulated that, to perform an overlap repair, the

Table 2. Defecatory Symptoms, Perineal Pain, and Dyspareunia at 12 Months

Overlap End to End
(n ⴝ 27) (n ⴝ 25)
[n (%)] [n (%)] RR (95% CI) P* NNT
Fecal incontinence 0 6 (24.0) 0.07 (0.00–1.21) .009 4.2
Fecal urgency 1 (3.7) 8 (32.0) 0.12 (0.02–0.86) .02 3.6
Flatus incontinence 4 (14.9) 4 (16.0) 0.93 (0.26–3.31) 1.0 100
Perineal pain 0 5 (20) 0.08 (0.00–1.45) .04 5
Dyspareunia 2 (7.4) 3 (12.0) 0.62 (0.11–3.39) 1.0 20
RR, relative risk; CI, confidence interval; NNT, number needed to treat.
* Chi-square test.

1266 Fernando et al Repair of Obstetric Anal Sphincter Injury OBSTETRICS & GYNECOLOGY
Table 3. Defecatory Symptoms, Perineal Pain, and Dyspareunia From 6 Weeks to 6 Months
Overlap End to End
[n/n (%)] [n/n (%)] RR (95% CI) P*
Fecal incontinence
6 weeks 4/32 (9.4) 6/31 (19.4) 0.65 (0.20–2.07) .30
3 months 2/29 (6.9) 9/31 (29.0) 0.24 (0.06–1.01) .04
6 months 0/28 6/28 (21.4) 0.08 (0.00–1.30) .02
Fecal urgency
6 weeks 9/32 (28.1) 8/31 (25.8) 1.09 (0.48–2.46) .16
3 months 9/29 (31.0) 14/31 (45.2) 0.69 (0.35–1.34) .26
6 months 2/28 (7.1) 9/28 (32.1) 0.22 (0.05–0.94) .04
Flatus incontinence
6 weeks 3/32 (9.4) 6/31 (19.4) 0.48 (0.13–1.77) .30
3 months 7/29 (24.1) 5/31 (16.2) 1.50 (0.53–4.19) .53
6 months 7/28 (25.0) 4/28 (14.3) 1.75 (0.58–5.32) .50
Perineal pain
6 weeks 8/32 (25) 8/31 (25.8) 0.97 (0.42–2.26) 1.0
3 months 2/29 (6.8) 5/31 (16.1) 0.43 (0.09–2.03) .43
6 months 1/28 (3.6) 4/28 (14.3) 0.25 (0.03–2.10) .35
Dyspareunia
3 months 12/29 (41.4) 9/31 (29.0) 1.43 (0.71–2.87) .45
6 months 6/28 (21.4) 7/28 (25.0) 0.86 (0.33–2.23) .78
RR, relative risk; CI, confidence interval.
* Chi-square test.

Table 4. Improvement of Anal Incontinence
Symptoms From 6 Weeks to 12 Months
Overlap
(n ⴝ 27)
[n (%)]
No anal incontinence 7 (25.9)
Symptoms improved 17 (62.9)
Symptoms remained the same 3 (11.1)
Symptoms became worse — 4 (16)
P ⫽ .01, Mann-Whitney test.
full length of the external anal sphincter has to be
identified, whereas the end-to-end method can be
performed without the full length being identified,
which may result in a deficient repair.13 Secondly,
because the anal sphincter is normally under tonic
contraction, the end-to-end technique may be more
vulnerable to ischemia (particularly with figure-of-
eight sutures) due to retraction of the apposed mus-
cles. Conversely, the overlap technique allows for
some retraction while still maintaining apposition.
In our study, if the internal anal sphincter was
torn, it was repaired separately. The internal anal
sphincter is a smooth muscle and contributes to most
of the resting anal pressure. Internal anal sphincter
dysfunction is usually associated with symptoms of
incontinence of flatus and passive soiling. However,
because combined external and internal anal sphinc-
ter injuries can occur, mixed symptoms may develop.
In contrast, when colorectal surgeons perform a sec-
ondary repair for fecal incontinence, they find it
difficult to identify and repair a scarred internal anal
sphincter, although one study has shown good results.21
End to End
(n ⴝ 25) Fecal urgency and fecal urge incontinence can be
[n (%)] more distressing than passive fecal incontinence.22
4 (16) Women with fecal urgency are continually inhibited
9 (36) by the threat of fecal urge incontinence. It has been
8 (32) reported that many women with fecal urgency were
not willing to put their continence to the test.6 This
study has clearly demonstrated the advantage of the
primary overlap repair in terms of fecal urgency.
Compared with the end-to-end technique, the
primary overlap technique was associated with a
significantly longer median operating time (28 min-
utes versus 38 minutes, P ⫽ .03). However, this could
be attributed to more dissection associated with over-
lap technique compared with the end-to-end tech-
nique. The median estimated blood loss was margin-
ally higher (100 mL versus 260 mL, P ⫽ .05) with the
overlap technique, but there was no significant
change in the hematocrit, and none of the participants
required a blood transfusion.
Williams23 reported that suture migration with the
use of permanent sutures such as Prolene (Ethicon)
occurs in one third of women after primary repair of
obstetric anal injury. In our study no woman com-
plained of suture migration or required suture re-
moval, and none developed wound dehiscence. We
did, however, ensure that the PDS suture ends were
cut short and carefully buried by overlying perineal

VOL. 107, NO. 6, JUNE 2006 Fernando et al Repair of Obstetric Anal Sphincter Injury 1267
muscles. It is difficult to offer an explanation as to why
more women in the end-to-end group reported peri-
neal pain at 12 months. It is acknowledged that
denervation and reinnervation may play a role, but
we have not performed any invasive neurophysiolog-
ical tests.
We calculated the sample size based on 2 com-
parable, but separate, observational studies at the
commencement of the study1,13 and required 48
women. Although we took every possible step to
maximize compliance, only 52 (81%) of the original
64 women returned the questionnaire at 12 months.
In most clinical settings in the United Kingdom,
when obstetric anal sphincter injuries are diagnosed,
they are repaired by middle-grade obstetric trainees
with variable experience.4 However, in this study we
specifically aimed to compare the outcome of 2
techniques and therefore restricted the trained oper-
ators to two. Having established the outcomes, we
have now commenced a more pragmatic randomized
controlled study comparing the 2 techniques, with
stratification for operator experience. We may then
be in a position to establish whether training needs
reappraisal.
This randomized study has shown that primary
overlap repair of the external sphincter is associated
with a significantly lower incidence of fecal inconti-
nence, urgency, and perineal pain. Furthermore,
when symptoms do develop, they appear to remain
unchanged or deteriorate in the end-to-end group but
improve in the overlap group. Because an increasing
number of women are requesting elective cesarean
delivery out of fear of perineal trauma and its conse-
quences,24 it is important that we are able to reassure
them that, when obstetric anal sphincter injury is
identified, it will be repaired by a skilled clinician
using an evidence-based suture technique, thus mini-
mizing the associated morbidity.
REFERENCES
1. Sultan AH, Kamm MA, Hudson CN, Bartram CI. Third
degree obstetric anal sphincter tears: risk factors and outcome
of primary repair. BMJ 1994;308:887–91.
2. Fenner DE, Genberg B, Brahma P, Marek L, DeLancey JO.
Fecal and urinary incontinence after vaginal delivery with anal
sphincter disruption in an obstetric unit in the United States.
Am J Obstet Gynecol 2003;189:1543–50.
3. Nazir M, Carlsen E, Jacobsen AF, Nesheim BI. Is there any
correlation between objective anal testing, rupture grade and
bowel symptoms after primary repair of obstetric anal sphinc-
ter injury? Dis Colon Rectum 2002;45:1325–31.
4. Fernando RJ, Sultan AH, Radley S, Jones PW, Johanson RB.
Management of obstetric anal sphincter injury: a systematic
review and national practice survey. BMC Health Serv Res
2002;2:9.
1268 Fernando et al Repair of Obstetric Anal Sphincter Injury
5. Norton C, Christiansen J, Butler U, Harari D, Nelson RL,
Pemberton J, et al. Anal incontinence. In: Abrams P, Cardozo
L, Khoury, Wein A, editors. Incontinence. 2nd ed. Plymouth
(UK): Health Publication Ltd; 2002. p. 985–1044.
6. Haadem K, Dahlstrom JA, Ling L, Ohrlander S. Anal sphinc-
ter function after delivery rupture. Obstet Gynecol 1987;70:
53–6.
7. Giebel GD, Mennigen R, Chalabi KH. Secondary anal recon-
struction after obstetric injury. Coloproctology 1993;1:55–8.
8. Haadem K, Ohrlander S, Lingman G. Long term ailments due
to anal sphincter rupture caused by delivery: a hidden prob-
lem. Eur J Obstet Gynecol Reprod Biol 1988;27:27–32.
9. Leigh RJ, Turnberg LA. Faecal incontinence: the unvoiced
symptoms. Lancet 1982;1:1349–51.
10. Mellgren A, Jensen LL, Zetterstrom JP, Wong WD, Hofmeister
JH, Lowry AC. Long term cost of fecal incontinence secondary
to obstetric injuries. Dis Colon Rectum 1999;42:857–67.
11. Sultan AH. Obstetric perineal injury and anal incontinence.
Clin Risk 1999;5:193–6.
12. Sultan AH, Thakar R. Lower genital tract and anal sphincter
trauma. Best Pract Res Clin Obstet Gynecol 2002;16:99–115.
13. Sultan AH, Monga AK, Kumar D, Stanton SL. Primary repair
of obstetric anal sphincter rupture using the overlap technique.
Br J Obstet Gynecol 1999;106:318–23.
14. Adams EJ, Fernando RJ. Management of third and fourth
degree perineal tears following vaginal delivery. Guideline No
29. London: Royal College of Obstetricians & Gynecologists;
2001. Available at: http://www.rcog.org.uk/resources/Public/
pdf/Perineal_Tears_No29.pdf. Retrieved March 20, 2006.
15. Pocock SJ. Clinical trials: a practical approach. Chichester
(UK): John Wiley & Sons; 1984.
16. Kettle C, Hills RK, Jones P, Darby L, Gray R, Johanson R.
Continuous versus interrupted perineal repair with standard or
rapidly absorbed sutures after spontaneous vaginal birth: a
randomized controlled trial. Lancet 2000;359:2217–23.
17. Vaizey CJ, Carapeti E, Cahill JA, Kamm MA. Prospective
comparison of faecal incontinence grading system. Gut 1999;
44:77–80.
18. Rockwood TH, Church JM, Fleshman JW, Kane RL, Mavran-
tonis C, Thorson AG, et al. Fecal incontinence Quality of Life
Scale: quality of life instrument for patients with fecal inconti-
nence. Dis Colon Rectum 2000;43:9–16.
19. Fitzpatrick M, Behan M, O’Connell PR, O’Herlihy C. A
randomized clinical trial comparing primary overlap with
approximation repair of third degree obstetric tears. Am J
Obstet Gynecol 2000;183:1220–4.
20. Garcia V, Rogers RG, Kim SS, Hall RJ, Kammerer-Doak DN.
Primary repair of obstetric anal sphincter laceration: a random-
ized trial of two surgical techniques. Am J Obstet Gynecol
2005;192:1697–701.
21. Meyenberger C, Bertschinger P, Zala GF, Buchmann P.
Anal sphincter defects in fecal incontinence: correlation
between endosonography and surgery. Endoscopy 1996;28:
217–24.
22. Gee AS, Durdey P. Urge incontinence of faeces is a marker of
severe external anal sphincter dysfunction. Br J Surg 1995;82:
1179–82.
23. Williams A. Third degree perineal tears: risk factors and
outcome after primary repair. J Obstet Gynaecol 2003;23:
611–14.
24. Sultan AH, Stanton SL. Preserving the pelvic floor and peri-
neum during childbirth: elective caesarean section? Br J Obstet
Gynecol 1996;103:731–4.
OBSTETRICS & GYNECOLOGY