Ultrasound Obstet Gynecol 2008; 31: 72–77

Published online 6 December 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.5200

Deficient lower-segment Cesarean section scars: prevalence

and risk factors
D. OFILI-YEBOVI*, J. BEN-NAGI*, E. SAWYER*, J. YAZBEK*, C. LEE*, J. GONZALEZ† and
D. JURKOVIC*
*Early Pregnancy and Gynaecology Assessment Unit and †Clinical Research Statistics Unit, Weston Education Centre, Denmark Hill
Campus, King’s College Hospital, London, UK

K E Y W O R D S: Cesarean section; deficiency; dehiscence; ultrasound; uterine scar

ABSTRACT
Objective To examine the sonographic features of
transverse lower-segment uterine Cesarean section scars
in non-pregnant, premenopausal women and to identify
factors associated with scar deficiency.
Methods Non-pregnant, premenopausal women with
histories of previous transverse lower-segment Cesarean
sections, who were referred for an ultrasound scan for
a variety of gynecological indications, were included
in this study. An attempt was made to identify the
uterine scars on transvaginal ultrasound scan and to
describe their locations and morphological features.
Various demographic, clinical and ultrasound data were
examined in order to identify factors associated with
deficient scars. Deficient scars were defined as detectable
myometrial thinning at the site of the Cesarean section
scar.
Results Lower-segment uterine scars were detected in
321/324 (99.1%; 95% CI, 98.0–100) women with a
history of previous Cesarean section. Sixty-three (19.4%;
95% CI, 15.1–23.8) women had evidence of deficient
Cesarean scars. Using multivariate analysis, a history of
multiple Cesarean sections, uterine retroflexion and the
inability to visualize all Cesarean scars in women with
previous multiple Cesarean sections were all shown to be
significantly associated with deficient scars.
Conclusion Deficient uterine scars are a frequent finding
in women with a history of previous Cesarean section.
The risk of scar deficiency is increased in women with
a retroflexed uterus and in those who have undergone
multiple Cesarean sections. Copyright  2007 ISUOG.
Published by John Wiley & Sons, Ltd.
INTRODUCTION
The number of deliveries by Cesarean section has
been increasing steadily worldwide in recent decades.
Although it is often assumed that Cesarean section
improves neonatal outcomes, there is no hard scientific
evidence to support this view1 . The safety of Cesarean
section, however, has increased owing to improvements
in surgical and anesthetic techniques, increased safety
of blood transfusion and routine use of antibiotics
and thromboprophylaxis2,3 . Cesarean section is also
associated with long-term risks such as postoperative
pelvic adhesions, uterine scar rupture, and placental
complications such as placenta previa and accreta4 – 6 .
The latter two complications are likely to be associated
with the poor uterine scar healing following Cesarean
sections.
Uterine scar dehiscence may present as an acute
event in the antenatal or intrapartum period, leading
to significant fetal and maternal morbidity7,8 . The
frequency of uterine rupture is estimated at 0.2–3.8%
and that of uterine dehiscence is between 0.6 and
3.8%9 – 11 .
Gilliam et al. identified an increased risk of placenta
previa with a history of Cesarean section5 . The
consequences of abnormally adherent placenta are
particularly severe, and they are responsible for 41–64%
of all obstetric hysterectomies; 65% of these cases have a
history of previous Cesarean section7,8 . 80% of maternal
deaths associated with placenta previa in the UK over
the last 12 years occurred in women with a history
of previous Cesarean section and abnormally adherent
placenta.
In recent years the first-trimester diagnosis of early
pregnancy implantation into a deficient Cesarean section
scar has been described9,10 . The term ‘Cesarean scar

Correspondence to: Mr D. Jurkovic, Early Pregnancy and Gynaecology Assessment Unit, 3rd Floor Golden Jubilee Wing, King’s College
Hospital, Denmark Hill, London SE5 8RX, UK (e-mail: davor.jurkovic@kcl.ac.uk)
Accepted: 10 July 2007

Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd. ORIGINAL PAPER
Cesarean section scars in non-pregnant women 73

pregnancy’ has been used to describe this condition,
which often leads to serious maternal morbidity due to
severe hemorrhage. There is also evidence that viable
Cesarean scar pregnancies have the potential to develop
into placenta previa or accreta at term9,10 . Routine
surveillance of Cesarean section scars by ultrasonography
during pregnancy has been proposed by some authors,
in an attempt to identify ‘silent’ or asymptomatic scar
dehiscence11,12 . Several studies have attempted to assess
scar integrity during pregnancy, but the sonographic
detection of uterine scars is easiest in the non-pregnant
state13 . Scar integrity has also been assessed by saline
contrast sonohysterography, in order to delineate scar
deficiency more accurately14 . However this method of
assessment is not without risks and therefore is limited in
its practical application.
The purpose of this study was to describe morphologi-
cal features of transverse lower-segment uterine Cesarean
section scars on non-enhanced B-mode transvaginal ultra-
sound scans in a large group of non-pregnant women.
The morphological appearances of the scars were com-
pared to the demographic data and obstetric history in
an attempt to identify factors associated with deficient
Cesarean section scars.
all the women, the distance between the uterine Cesarean
section scar and the top of the uterine cavity (a) was
measured in the longitudinal plane (Figure 1). The ‘height
ratio’ of the uterine Cesarean scar was defined as the
ratio of this distance (a) to the whole length of the uterine
cavity measured from the internal os to the top (b).
A value of 1 corresponds to the scar being located at
the level of the internal os, while a value < 1 indicates
that the scar is above the level of the internal os. In
cases of multiple Cesarean sections, the scar closest to
the fundus was taken as representative for the particular
case.
Scars were described as deficient if there was detectable
myometrial thinning at the site (Figure 2). The degree of
thinning was expressed as the ratio of the myometrial
thickness at the depth of the scar (c) to the thickness
of the adjacent normal myometrium measured in the
longitudinal section (d) (‘deficiency ratio’). The loss of
more than 50% of myometrial mantle at the scar level
was classified as severe deficiency.
The study was approved by the local ethics committee
and all patients gave consent to take part in it. All

METHODS
Non-pregnant, premenopausal women with a history of
previous transverse lower-segment Cesarean section, who
were referred to our early pregnancy and gynecology
assessment unit for a variety of gynecological indications,
were invited to join this study. A full medical history
was taken in each case including a detailed obstetric and
gynecological history, the number of previous Cesarean
sections, the gestational age at the time of the operation
and the indication for the operation. Women were
included only if the Cesarean section had been performed Figure 1 A longitudinal view of the uterus illustrating the
more than 3 months prior to the assessment. Women who assessment of Cesarean section scar position (S) in relation to the
underwent classical Cesarean sections, open myomectomy internal os (O). Distances from the scar to the uterine fundus
and those who underwent a hysterectomy following a (a) and from the internal os to the fundus (b) are measured as
shown, and the height ratio is expressed as a/b.
Cesarean section were excluded from the study. All
the women underwent a transvaginal ultrasound scan,
which was performed by gynecologists with expertise in
transvaginal scanning using high-frequency transducers
of 5–7.5 MHz (Aloka SSD-5000, Aloka Co., Tokyo,
Japan).
On ultrasound scan, the uterus was examined in the
longitudinal plane and the internal os was identified as
the point of junction between the endometrial cavity and
the cervical canal. The uterine flexion was determined by
assessing the angle between the longitudinal axis of the
uterus and the longitudinal axis of the cervix. Uterine
anteflexion was diagnosed when the long axis of the
uterine body was deviating anteriorly in relation to the
long axis of the cervix, while posterior deviation was
classified as retroflexion. An attempt was then made to Figure 2 A longitudinal view of a uterus with a deficient Cesarean
section scar. The severity of scar deficiency was assessed by
ascertain the location of a Cesarean section scar within measuring the myometrial thickness at the depth of the scar (c) and
the anterior uterine wall. In cases of multiple previous the thickness of the adjacent unaffected myometruium (d). The
Cesareans, the number of all visible scars was recorded. In degree of deficiency is expressed as the ratio c/d.

Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2008; 31: 72–77.
74 Ofili-Yebovi et al.

clinical findings were stored in a clinical database that
facilitated data entry and retrieval (PIA-Fetal Database,
Viewpoint Bildverabeitung GmbH, Wessling, Germany).
Statistical evaluation was by univariate and multivariate
analysis. Univariate analysis involved comparing different
variables in the two main groups, i.e. those women
with deficient and those with non-deficient scars. The
continuous variables were compared using two-sample
t-test or Mann–Whitney test where indicated, while the
categorical variables were compared using the Pearson
chi-square or Fisher’s exact test as appropriate. A multiple
logistic regression model was used to see the concurrent
effects of all the potential prognostic factors, paying
special attention to interactions. The variable Cesarean
section was fitted in two ways, one according to the total
number of scars visualized, the other by differentiating
full-term and preterm Cesareans. Term gestations were
defined as pregnancies that were completed at 37 weeks’
gestation or more. In addition, logistic function was
used in a discriminant analysis to provide a risk
classification for these patients. Multivariate analysis
involved constructing multiple logistic regression models
for all variables to identify which retained statistical
significance, and a value of P < 0.05 was considered
to be statistically significant.
RESULTS
In a period of 18 months a total of 354 women with
histories of previous transverse lower-segment Cesarean
sections were examined by transvaginal ultrasound.
Thirty (8.5%) data sets were incomplete and these women
were excluded from further analysis. The remaining 324
(91.5%) women, who were included in the data analysis,
had undergone a total of 471 Cesarean sections. Two
hundred and eleven (65.1%, 95% CI 61–69) women had
had one Cesarean section, 84 (25.9%, 95% CI 22–30)
had had two Cesarean sections, 24 (7.4%, 95% CI 5–10)
had had three, and five (1.5%, 95% CI 0–2.6) had had
four Cesarean sections.
There was no statistically significant difference in the
mean gestational age at the time of surgery or in the type of
Cesarean section (emergency or elective) between women
who underwent single or multiple Cesareans (Table 1).
However, women who had had three or more Cesarean
sections were younger at the time of their first Cesarean
compared to those who had had fewer Cesarean sections
(Table 1).
Cesarean section scars were visible on ultrasound
scan in 321/324 (99.1%, 95% CI 98.0–100) women.
However, it was not always possible to identify individual
scars in women with histories of multiple Cesarean
sections. The total number of visible individual scars
decreased from 210/211 (99.5%, 95% CI 97.4–99.9)
in women with a single previous section to 123/168
(73.2%, 95% CI 66.0–79.3) in women with two previous
operations and 40/92 (43.5%, 95% CI 39.7–53.7)
in those with three or more operations (P < 0.0001)
(Table 1).
The majority of scars were located close to the
internal os but 14/321 (4.4%, 95% CI 2.2–6.6)
women had evidence of corporal scars (Table 2). A
total of 63/324 (19.4%, 95% CI 15.1–23.8) women
had evidence of myometrial thinning at the Cesarean
section site as previously defined. In 32/324 (9.9%,
95% CI 7.1–13.4) women the defects were severe,
involving ≥ 50% of the myometrium (Table 2). The
distribution of myometrial thickness measurement is
shown in Figure 3. On univariate analysis there were
significant differences between women with deficient and
Table 2 Distribution of height and myometrial deficiency ratios
(n = 321)
Ratio range
Parameter 1.0–0.76 0.75–0.51 0.5–0.26 0.25–0.01
Height ratio 307 (95.6) 11 (3.4) 3 (0.9) —
(a/b) (n (%))
Deficiency ratio 271 (84.4) 18 (5.6) 28 (8.7) 4 (1.2)
(c/d) (n (%))
a, distance between the uterine Cesarean section scar and the top of
the uterine cavity in the longitudinal plane; b, whole length of the
uterine cavity measured from the internal os to the top;
c, myometrial thickness at the depth of the scar; d, thickness of the
adjacent normal myometrium measured in the longitudinal section.

Table 1 Demographic data and scar morphology in women with single and multiple Cesarean sections (CS)

Three or more
Single CS Two CS CS
Parameter (n = 211) (n = 84) (n = 29) P

Maternal age at time of first CS (years, mean (range)) 29.4 (14–44) 30.8 (20–41) 25.7 (18–31) 0.01
Gestational age at CS (weeks, mean (range)) 38.0 (24–42) 38.6 (26–43) 38.4 (28–42) NS
First CS by emergency (n (%)) 154/211 (73.0) 78/168 (46.4) 34/92 (37.0) NS
Women with deficient scars (n (%)) 32/211 (15.2) 20/84 (23.8) 11/29 (37.9) 0.01
Visualized scars (n (%))
0 1 (0.5) 1 (1.2) 1 (3.4) 0.0001
1 210 (99.5) 43 (51.2) 19 (65.5)
2 — 40 (47.6) 6 (20.7)
3 — — 3 (10.3)

NS, not significant.

Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2008; 31: 72–77.
Cesarean section scars in non-pregnant women 75

Table 3 Univariate analysis of women with deficient and non-deficient Cesarean section (CS) scars

Non-deficient Deficient
Parameter (n = 261) (n = 63) P

Age (years, mean (95% CI)) 41 (39.5–41.6) 40 (38–43) NS

Parity (median (range)) 2 (1–12) 2 (1–4) NS
Number of CSs (median (range)) 1 (1–4) 1 (1–3) 0.01
Indication for CS* (n (%))
Elective 67 (27) 19 (33) NS
Emergency 179 (73) 38 (67)
Proportion of preterm CS (n (%))
Term 222 (85.1) 54 (85.7) NS
Preterm 39 (14.9) 9 (14.3)
Number of vaginal births in addition to CS (median (range)) 0 (0–10) 0 (0–3) NS
Number of visible scars (median (range)) 1 (0–3) 1 (1–2) NS
Height ratio of Cesarean scar above internal os (a/b) (median (range)) 1 (0.25–1) 1 (0.46–1) NS
Uterine position (n (%))
Anteflexed 226 (87) 45 (71.4) 0.005
Retroflexed 35 (13) 18 (28.6)

*Women who had both emergency and elective Cesarean section are excluded. a, Distance between the uterine Cesarean section scar and the
top of the uterine cavity in the longitudinal plane; b, whole length of the uterine cavity measured from the internal os to the top; NS, not
significant.

25 for each additional Cesarean section (OR = 1.9, 95% CI

1.3–2.9).
20
DISCUSSION
15 Our study showed that lower-segment uterine Cesarean
Number

section scars can be identified on standard B-mode

10 transvaginal ultrasound scan in almost all women with a
history of previous Cesarean sections. In women with
a history of multiple Cesarean sections it was often
5
impossible to identify all previous scars individually,
and the number of detectable individual scars was
0 lower with increasing number of previous Cesarean
0− 2.0− 4.0− 6.0− 8.0− 10.0− 12.0−
1.9 3.9 5.9 7.9 9.9 11.9 13.9 sections. This may be explained by the fact that
Myometrial thickness (mm) some uterine incisions are made at exactly the same
site as the previous scar. In addition there was a
Figure 3 Distribution of myometrial thickness at the Cesarean tendency for uterine scars to become deficient in cases
section site in women with evidence of scar deficiency (n = 63).
of multiple sections, which also interferes with the
ability to visualize individual scars on the ultrasound
Table 4 Multiple logistic regression analysis of deficient versus scan.
non-deficient Cesarean section scars
We found that nearly 20% of women had detectable
myometrial thinning at the site of a previous Cesarean
Coeffi- Odds ratio
section. In half of these women the defects were
Variable cient (95% CI) P
large, involving more than 50% of the myometrial
Position of uterus 0.89 2.4 (1.3–4.8) 0.01
thickness. There were only three variables that were
Number of scars visualized −1.15 0.31 (0.13–0.75) 0.01 associated with deficient scars: history of multiple
Number of Cesarean sections 0.69 1.9 (1.3–2.9) 0.001 Cesarean sections; uterine retroflexion; and the inability
Constant −2.23 to visualize all uterine scars in cases of multiple
sections.
A relationship between multiple previous Cesarean
non-deficient scars in the number of previous Cesarean sections and scar deficiency has also been reported
sections and uterine flexion (Table 3). Using multivariate by Regnard et al.15 . Our study has confirmed this
analysis, the history of multiple Cesarean sections, uterine observation, and we showed that the odds of a
retroflexion and the inability to visualize all Cesarean scars scar becoming deficient increase with the number of
in women with previous multiple Cesarean sections were previous sections. An analogy to this observation is
all significantly associated with deficient scars (Table 4). where repeated trauma to a wound can disrupt the
The odds ratios (OR) for deficient scars nearly doubled normal healing process16 . Furthermore, this can be

Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2008; 31: 72–77.
76
extrapolated from the final stage of wound healing in
the skin, whereby the highly vascular granulation tissue
is replaced by avascular scar tissue17 . Thus, further
injury to the scar tissue, an area with poor vascular
perfusion, will compromise the pathways involved in
healing18 .
There was also an association between scar defi-
ciency and inability to identify all the scars in women
with multiple Cesarean sections. However, this find-
ing is more likely to be a consequence, rather than
the cause, of scar deficiency. Deficient scars tend to
occupy larger areas of the lower uterine segment,
which hampers visualization of individual scars. As
expected, the ability to see individual scars decreased
with increasing number of previous Cesarean sec-
tions.
Uterine retroflexion was another variable that was
associated with deficient scars. The chance of a woman
with a retroflexed uterus having a deficient scar was more
than twice that of a woman with an anteflexed uterus.
The flexion point of the uterus is at the level of the
internal os. In a retroflexed uterus the lower segment of
the uterus is therefore under a degree of tension, which
may compromise healing of a Cesarean section scar. This
could be the result of mechanical traction to the scar
or reduced vascular perfusion caused by stretching of the
lower uterine segment. Impaired tissue perfusion results in
reduced wound oxygen tension, which has been reported
to delay wound healing by slowing the production of
collagen16 .
The clinical significance of these findings is uncertain.
Uterine rupture in women with previous Cesarean section
is a rare event that occurs in 12 per 10 000 with
elective repeat sections and up to 35 per 10 000 with
vaginal births after Cesarean section19 . Therefore a
high level of observed scar deficiency, as seen in our
study, is unlikely to be helpful in identifying women
at risk of scar rupture. Even with evidence of severe
deficiency of more than 50% of myometrial thickness the
risk of scar rupture would still be less than 5%. The
results of the study may also be important for patients
undergoing endometrial ablation. There have been cases
of thermal injury to the bowel following microwave
endometrial ablation in patients who had undergone
Cesarean deliveries as a result of the microwave energy
passing through the thin myometrial layer and affecting
the bowel20 .
Detection of severe scar deficiency may be helpful
in identifying women at risk of Cesarean scar ectopic
pregnancy in the future. This is also a rare complication
of Cesarean sections that occurs in approximately 1
in 1800 pregnancies9 . However, scar implantation is a
severe complication, which is associated with significant
maternal morbidity. There is some evidence from the
literature that scar implantations tend to occur only in
women with severe deficient scars and that the risk of
scar implantation may be proportional to the size of the
myometrial defect. The treatment of scar pregnancies
is simple and highly effective when the diagnosis is
Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd.
Ofili-Yebovi et al.
made in the first trimester, but the risk of severe
complications increases with increasing gestational age.
In view of the observed association between multiple
Cesarean sections and scar deficiency it may be feasible
to offer all women with two or more previous sections
an early scan to rule out scar implantation. In our
hospital, with approximately 5000 deliveries/year and
a Cesarean section rate of 22%, the percentage of women
booking for antenatal care with a history of two or
more previous Cesarean sections was 1.5–1.8%. If all
these women attended for ultrasound screening of scar
pregnancy an additional 70–90 scans/year would need
to be performed in order to detect 50% of all scar
implantations.
In conclusion deficient Cesarean scars are a frequent
finding in a population of women who had had
previous Cesarean sections. A policy of routine scar
assessment by ultrasonography in order to prevent
uterine rupture is unlikely to be practical or helpful.
However, a small number of women with multiple
previous sections may benefit from an early scan at
6–7 weeks’ gestation to identify and treat scar ectopic
pregnancies.
REFERENCES
1. National evidence-based clinical guidelines: Caesarean section
April 2004. National collaborating centre for women’s and
children’s health commissioned by the National Institute
of Clinical Excellence. RCOG Press: London, 2004; 3 &
48.
2. Jolly J, Walker J, Bhabra K. Subsequent obstetric performance
related to primary mode of delivery. Br J Obstet Gynaecol 1999;
106: 227–232.
3. LaSala A, Berkeley A. Primary Cesarean section and subsequent
fertility. Am J Obstet Gynecol 1987; 157: 379–383.
4. Clark S, Koonings P, Phelan J. Placenta previa/accreta and prior
Cesarean section. Obstet Gynecol 1985; 66: 89–92.
5. Gilliam M, Rosenberg D, Davis F. The likelihood of placenta
previa with greater number of Cesarean deliveries and higher
parity. Obstet Gynecol 2002; 99: 976–980.
6. Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for
placenta previa- placenta accreta. Am J Obstet Gynecol 1997;
177: 210–214.
7. Zelop C, Harlow B, Frigoletto F, Safon L, Saltzman D. Emer-
gency peripartum hysterectomy. Am J Obstet Gynecol 1993;
168: 1443–1448.
8. Castenada S, Karrison T, Ciblis L. Peripartum hysterectomy. J
Perinat Med 2000; 28: 472–481.
9. Jurkovic D, Hillaby K, Woelfer B, Lawrence A, Salim R,
Elson C. First-trimester diagnosis and management of pregnan-
cies implanted into the lower uterine segment Cesarean section
scar. Ultrasound Obstet Gynecol 2003; 21: 220–227.
10. Ben Nagi J, Ofili-Yebovi D, Marsh M, Jurkovic D. First-
trimester Cesarean scar pregnancy evolving into placenta
previa/accreta at term. J Ultrasound Med 2005; 24:
1569–1573.
11. Armstrong V, Hansen W, Van Voorhis B, Syrop C. Detection
of Cesarean scars by transvaginal ultrasound. Obstet Gynecol
2003; 101: 61–65.
12. Zilberman A, Sharf M, Polishuk W. Evaluation of Cesarean
section scar by hysterography. Obstet Gynecol 1968; 32:
153–157.
13. Bromley B, Pitcher B, Klapholz H, Lichter E, Benacerraf B.
Sonographic appearance of uterine scar dehiscence. Int J
Gynecol Obstet 1995; 51: 53–56.
Ultrasound Obstet Gynecol 2008; 31: 72–77.
Cesarean section scars in non-pregnant women
14. Monteagudo A, Carreno C, Timor-Tritsch I. Saline infusion
sonohysterography in nonpregnant women with previous
Cesarean delivery: the ‘‘niche’’ in the scar. J Ultrasound Med
2001; 20: 1105–1115.
15. Regnard C, Nosbusch M, Fellemans C, Benalli N, Van Rys-
selberghe M, Barlow P, Rozenberg S. Cesarean section scar
evaluation by saline contrast sonohysterography. Ultrasound
Obstet Gynecol 2004; 23: 289–292.
16. Wound healing, chronic wounds http://www.emedicine.com/
plastic/topic477.htm [Accessed 5 January 2007].
17. Alison M. Repair and regenerative processes. In Oxford
textbook of pathology. Vol. 1 Principles of Pathology, O’D
Copyright  2007 ISUOG. Published by John Wiley & Sons, Ltd.
77
McGee J, Isaacson PG, Wright NA, Dick HM, (eds). Oxford
University Press: New York, 1992; 377–378.
18. Whaley K, Burt A. Factors influencing wound healing. In Muir’s
textbook of pathology, MacSween R, Whaley K (eds). Edward
Arnold: London, 1992; 160–161.
19. National evidence-based clinical guidelines: Caesarean section
April 2004. National collaborating centre for women’s and
children’s health commissioned by the National Institute of
Clinical Excellence. RCOG Press: London, 2004; 18.
20. Itzkowic D, Beale M. Uterine perforation associated with
endometrial ablation. Aust N Z J Obstet Gynaecol 1992; 32:
359–361.
Ultrasound Obstet Gynecol 2008; 31: 72–77.