Differentiation of Simple and Strangulated Small-Bowel Obstructions ...

Hyun K. Ha, MD #{149} Jin S. Kim, MD #{149} Moo S. Lee, MD #{149} Hong J. Lee, MD #{149} Yoong K. Jeong, MD
Pyo N. Kim, MD #{149} Moon C. Lee, MD #{149} Kee W. Kim, MD #{149} Mi Y. Kim, MD #{149} Yong Ho Auh, MD
Differentiation of Simple and Strangulated
Small-Bowel Obstructions: Usefulness
of Known CT Criteria’
PURPOSE: To evaluate the usefulness
of known computed tomographic
(CT) criteria for the differentiation of
simple and strangulated small-bowel
obstructions.
MATERIALS AND METHODS: CT
scans of 84 patients with simple (n
43) and strangulated (n = 41) smallbowel
obstructions caused by adhesions,
hernia, and volvulus were reviewed
retrospectively. Diagnoses
were made with surgery (n 55) and
during clinical follow-up (n 29).
CT criteria evaluated were configuralion
of obstructed bowel ioop, target
sign, bowel wall thickening and enhancement,
changes in mesentery
and mesenteric vasculature, and
amount and attenuation of ascites.
RESULTS: CT findings that enabled
the detection of strangulated obstructions
were poor or no enhancement
of bowel wall (sensitivity, 34%;
specificity, 100%) and a serrated beak
(sensitivity, 32%; specificity 100%).
When these two findings were excluded
from analysis, a large amount
of ascites, an unusual course of mesenteric
vasculature, and diffuse engorgement
of mesenteric vasculature
were shown to be useful CT findings
for performing multivariate regression
analysis. Application of these
five CT findings enabled identification
of 35 (85%) of 41 patients with
strangulated obstructions.
CONCLUSION: Detecting a combination
of selected, known CT criteria
increases the diagnostic accuracy of
CT to enable differentiation of simple
and strangulated small-bowel obstructions.
Abdominal and Gastrointestinal Radiology
S IMPLE and strangulated smallbowel
obstructions are distinguished
by the respective absence or
presence of impaired vascular circulalion
to the obstructed intestine. Their
differentiation is one of the principal
diagnostic problems for physicians
who treat patients with small-bowel
obstructions. Various findings on abdominal
CT images aid the diagnosis
of advanced bowel strangulation, yet
its early detection is often problematic.
Many reports of studies stress
the importance of noting mesenteric
changes, configuration of obstructed
bowel loop, or contrast enhancement
patterns of the bowel wall on CT images
(1-8). However, the lack of control
groups in these studies potentially
renders these CT findings less specific
and reliable (9). These findings may
also appear on CT scans of patients
with inflammatory bowel diseases,
gastrointestinal neoplasms, or paralytic
ileus; they may even appear on
CT images of healthy subjects. Furthermore,
some of the CT findings
that were considered to be highly suggestive
of strangulated small-bowel
obstruction have been seen on images
of simple obstruction (4). To ascertain
the diagnostic value of known CT criteria
of small-bowel strangulation, a
study that compares findings on CT
scans of a large number of patients
with simple and strangulated smallbowel
obstructions is necessary.
The purpose of this study was to
evaluate the usefulness of known CT
Index terms: Intestines, C’I 748.12112 #{149} Intestines, stenosis or obstruction, 748.723, 748.724
Radiology 1997; 204:507-512
criteria for identifying small-bowel
strangulation in a large series of 84
patients with simple or strangulated
small-bowel obstructions caused by
adhesions, hernia, or volvulus.
MATERIALS AND METHODS
A computerized search was conducted
at two institutions for cases from January
1991 to March 1996 of small-bowel obstructions
caused by adhesions, hernia, or volvulus
in patients who had undergone
surgery before developing small-bowel
obstruction. Of 450 patients identified, 173
had undergone CT for evaluation of their
small-bowel obstructions. However, 89
of these patients were excluded from the
study because of the following: CT scans
were not available (n = 15), small-bowel
obstructions were caused by malignant
(n = 35) or inflammatory (n = 23) lesions,
contrast material was not given intravenously
during the CT examination (n = 6),
or findings on CT scans reviewed did not
meet criteria for those of small-bowel obstruction
(n = 10) (2,10). The study population,
therefore, consisted of 84 consecutively
evaluated patients with simple (n
43) or strangulated (n = 41) small-bowel
obstructions whose Cr were reviewed
retrospectively. Patient ages ranged from
19 to 83 years (mean, 49 years). There
were 40 women and 44 men.
Before developing obstruction, all patients
had undergone surgery for the following:
inflammatory bowel disease (n
20), gastric or colon cancer (ii 19), gynecologic
disease (n = 20), hepatobiliary disease
(n = 13), and other conditions (n = 12).
Of these 84 patients, 55 underwent surgery
again because they exhibited clinical
I From the Departments of Radiology (H.KH.,J.S.K., H.J.L, Y.K.J., P.N.K., M.G.L, Y.H.A.) and Pre-
ventive Medicine (MS.L), Man Medical Centei University of Ulsan Medical College, 388-1 Poongnap-
Dong, Songpa-Ku, Seoul 138-040, Korea; Department of Radiology, Yongdong Severance Hospital,
Yonsei University Seoul, Korea (K.W.K.); and Department of Radiology, Inha University Hospital,
Kyunggi-do, Korea (M.Y.K.). From the 1996 RSNA scientific assembly. Received November 5, 1996;
revision requested December 20; revision received March 6, 1997; accepted March 17. Address reprint
requests to H.K.H.
C RSNA, 1997
507

signs and symptoms of small-bowel strangulation
or did not respond favorably to
conservative treatment; they underwent
simple adhesiotomy (n = 21) or segmental
bowel resection (n = 34) to relieve the obstruction.
Obstruction was confirmed during
surgery in all 55 patients; 41 had strangulated
obstructions and 14 had simple
obstructions. Obstructions were caused by
adhesions or bands (n = 43), hernia (n =
9), or volvulus (ri = 3). The time between
surgery and CT ranged from a few hours
to 13 days (mean, 2 days); for 31 patients,
it was less than or equal to 1 day.
In the other 29 patients, obstructions improved
with fluid and electrolyte replacement
and/or after the insertion of a nasogastric
tube for 2-11 days (mean, 5 days)
for decompression; in seven of these 29
patients, long decompression tubes were
used. Although not confirmed with surgeny,
the obstructions of 14 of these 29 patients
were diagnosed with small-bowel
follow-through examinations: complete
obstruction, two patients; high-grade partial
obstruction, eight patients; and lowgrade
partial obstruction, four patients.
With use of CT criteria proposed by
Fukuya et al (10) and Gazelle et al (11),
findings consistent with small-bowel obstruction,
and not paralytic ileus, were
seen on the CT scans of 10 of the other 15
patients. Diffuse small-bowel dilatation
(luminal diameter, 3.5 cm) and collapsed
distal small bowel, colon, on both were
seen on the CT scans of these 10 patients.
On the CT scans of the other five patients,
small-bowel loops with luminal diameters
less than 3.5 cm but greater than 3.0 cm
were seen. The plain abdominal radiographs
of these five patients, obtained
within 1 week before CT, revealed findings
typical of small-bowel obstruction (distended
fluid-filled proximal loops, multiple
step-ladder patterns of air-fluid 1evels,
and collapsed distal small-bowel,
colon, or both).
We considered these 29 patients to have
simple small-bowel obstruction. To support
the belief that these patients did not
have strangulated obstruction, the number
of the following clinical features in these
29 patients was compared with that in the
41 patients with strangulated obstruction:
abdominal tenderness, tachycardia (heart
rate, > 100 beats per minute), fever (ternperature,
>38.5#{176}C),and leukocytosis (white
blood cell count, >10,000/mr& [10.0 X
109/LJ). Of the 41 patients with strangulated
obstruction, 12 patients had all four
features, 10 had three, 16 had two, and
three had one. Of the 29 patients with
simple obstruction, four patients had two
of the features, 15 had one, and 10 had
none.
CT was performed with a GE 9800
Quick System, a HiLight Advantage (GE
Medical Systems, Milwaukee, Wis), or a
Somatom Plus-S (Siemens, Erlangen, Germany)
scanner. Sections of 8- or 10-mm
thickness were imaged from the diaphragm
to the pubis at intervals of 8 or 10 mm.
Contrast material (500-800 mL; E-Z-CAT,
2% iodinated, water soluble; E-Z-Em,
Westbury, NY) was given orally 30 mm-
Table 1
Findings on CT Scansof 84 Patients with Strangulated or Simple Obstruction
CT Findings
Strangulated
(n = 41)
Simple
(n = 43)
Bowel wall changes
Smootltserrated beak 5:13 26:0
Target sign 12 3
Wall thidmess (mean ± SD, mm) 5.1 ± 2.1 3.5 ± 1.0

1. #{163};1
Table 2
Diagnostic Value of CT Findings for the Detection of Small-Bowel Strangulation in
84 Patients
CT Findings Sensitivity SpeCifiCity P Value
Poor or no contrast enhancement 34 100

er of these findings on the CT scans
of the 43 patients with simple obstruction
were the following: no findings,
38 patients (88%) and one finding, five
patients (12%). The CT findings of
these five patients consisted of an unusual
course of mesentenic vasculature
(n = 2), diffuse mesenteric changes (n =
2), and a large amount of ascites (n = 1).
Interobserver agreement was excellent
for the identification of the target
sign (K = 0.76), poor on no contrast
enhancement of the bowel wall (K =
0.88), a small amount of ascites (K =
0.79), and a large amount of ascites
(K 0.83). It was good for the identification
of a smooth beak (K = 0.61), a
serrated beak (K 0.53), an unusual
course of mesenteric vasculatune (K =
0.56), focal engongement of mesentenic
vasculature (K 0.63), focal haziness
of the mesentery (K = 0.64), diffuse
engorgement of the mesenteric vasculature
(K = 0.69), diffuse haziness of
the mesentery (K 0.67), mesentenic
vascular thrombosis (K = 0.55), and
high attenuation of ascites (K = 0.63).
DISCUSSION
An understanding of the basic
pathophysiologic processes of smallbowel
strangulation is important because
CT findings are closely associated
with secondary changes in the
bowel wall, mesentery, and penitoneal
cavity. Three phenomena usually occur
together in small-bowel strangulation:
mechanical obstruction that is
proximal to the involved bowel segment,
closed-loop obstruction of the
involved bowel segment, and venous
congestion of the involved bowel loop
(14). Initially, venous return of blood
from the involved bowel segment is
compromised while influx of arterial
blood continues; this leads to engorgement
and distension of vessels, frank
hemorrhage into the bowel lumen or
wall, and exudation and transudation
of fluid across the senosa into the pentoneal
cavity (15). Although the CT
criteria (1-4) we used largely reflect
these pathophysiologic processes, the
different frequencies of each CT finding
seem to result from the different
primary causes and seventies of the
obstructions. There are also no data
that describe precisely the sequential
changes of the CT findings of strangulation
or how these findings are useful
for enabling the diagnosis of strangulation.
According to several researchers
(1-4,16,17), CT findings of slight thickening
of the bowel wall, the target
sign, engorgement of the mesentenic
vasculatune, and mesentenic edema
4.-.
3. 4.
Figures 3, 4. (3) Strangulated small-bowel obstruction caused by adhesions in a 45-year-old
man. Contrast-enhanced CT scan of the abdomen shows a serrated beak (arrows) in the bowel
loop at the site of obstruction. There is diffuse haziness and vascular engorgement in the regional
mesentery (arrowheads). (4) Strangulated small-bowel obstruction caused by adhesions
in a 70-year-old woman. Contrast-enhanced CT scan of the pelvis shows a fluid collection (F)
in the pelvic cavity that is hyperattenuating compared with fluid in the bowel loop.
‘%/#
5. 6.
Figures 5, 6. (5) Strangulated small-bowel obstruction caused by volvulus in a 28-year-old
man. Contrast-enhanced CT scan of the abdomen shows mesenteric vessels converging into
the site of obstruction (arrow). A fluid collection is present in the peritoneal cavity. (6) Strangulated
small-bowel obstruction caused by adhesive bands in a 46-year-old man. Contrastenhanced
CT scan of the abdomen shows diffuse segmental haziness of the mesentery (*).
are early or reversible signs of smallbowel
strangulation, whereas bowel
infarction on gangrene is indicated
with CT findings of high attenuation
of the bowel wall, pneumatosis, hemorrhagic
changes in the mesentery, gas
in the portal vein, and poor or no enhancement
of the bowel wall.
In a recent report, Frager et al (6)
stated that on CT scans there was no
completely specific preoperative indicaton
of strangulation in patients with
small-bowel obstruction caused by
hernias or adhesions. This suboptimal
specificity of CT was attributed mainly
to the subjective interpretation of various
ischemic signs on CT scans (without
the use of CT scans of a control group
with other diseases) and problems of
inspecting affected bowel during sungery.
However, as our study demonstrated,
some of the CT findings-if
one is attempting to differentiate simple
and strangulated small-bowel
obstructions-had very high specificities.
These CT findings were poor or
no enhancement of the bowel wall, a
serrated beak, diffuse engorgement of
mesenteric vasculature or mesentenic
haziness, an unusual course of mesentenic
vasculature, and a large amount
of ascites.
When compared to results of prior
studies (1-8), the results of our blinded
reinterpretation of the CT scans of a
large number of patients show that
most of the known CT criteria, when
used alone, are not sensitive indicatons
of simple or strangulated smallbowel
obstruction (Tables 1-3). The
CT scans of 19 (46%) of the 41 patients
with surgically proved strangulation
showed bowel-wall thickening of less
than 5 mm, scans of 27 patients (66%)
showed normal bowel-wall enhancement,
and scans of 29 patients (71%)
showed no target sign. Even three of
the five highly specific CT findings
510 #{149} Radiology August 1997

7. 8. 9.
Figures 7-9. (7) Strangulated small-bowel obstruction owing to incisional hernia in a 38-year-old man. Contrast-enhanced CT scan of the abdomen
shows prolonged enhancement of thickened bowel wall in the herniated, strangulated segment (Ii). There are fluid collections in the hernia
sac and right paracolonic gutter. (8) Simple obstruction caused by adhesive bands in a 38-year-old man. Contrast-enhanced CT scan of the
abdomen shows diffuse dilatation of a small-bowel loop with the target sign. The obstructed site (arrow) shows a smooth beak configuration.
(9) Simple obstruction caused by adhesions in a 60-year-old man. Contrast-enhanced CT scan of the abdomen shows multiple dilated smallbowel
loops and focal engorgement of the mesentenic vasculature (arrows) next to the obstructed site.
Table 3
Diagnostic Value of CT Findings for the Detection of Small-Bowel Strangulation in
61 Patients
CT Findings Sensitivity SpecifiCity P Value
Large amount of ascites 22 98 .024
Diffuse mesenteric haziness 39 98 .005
Diffuse engorgement of MV* 50 95

greatly aids the diagnosis of smallbowel
strangulation (15). The high
attenuation of ascites on CT images
of strangulated obstruction is probably
caused by increased protein content
in the fluid. However, in our study,
the attenuation of ascites was cornpared
with that of bile in the gallbladder
or cysts in the liver or kidney, and
images of both simple and strangulated
obstructions revealed hyperattentuating
asdtes. Since a large amount
of ascites may be caused by many
conditions, confirmation of other CT
findings in the mesentery or bowel is
required so that the diagnosis of strangulation
is not made incorrectly.
We defined the mesenteric changes
(haziness or vascular engorgement) as
focal or diffuse because these changes
may be observed on CT images of patients
with simple small-bowel obstruction.
As expected, focal mesenteric
changes were seen on CT images
of both simple and strangulated obstructions
(Fig 9). Diffuse mesenteric
changes, however, had high specificities
(>95%) for the detection of strangulation.
An unusual course of mesenteric
vasculature (indicated with a
reversed position of superior mesenteric
artery and vein, the whirl sign,
or convergence of vessels) has been
seen also in cases of closed-loop obstruction
(1,2,8) or malrotation (21).
Although further study is necessary,
some of these findings may be seen
on CT scans of asymptomatic patients.
Furthermore, strangulation does not
always occur in patients with closedloop
obstruction. Therefore, this finding-an
unusual course of the mesenteric
vasculature-alone on CT scans
of patients with small-bowel obstruction
does not always indicate bowel
strangulation.
Some drawbacks of our study could
have reduced the reliability of the results.
There were some differences in
the scanning and contrast-infusion
techniques at the two institutions where
patients were evaluated; these might
have affected the true frequency of CT
findings that were analyzed. In addition,
because strangulation can develop
in only a few hours, the relatively long
time (mean, 2 days) between CT examination
and the surgery for obstruction
in this study might have affected
the severity of the obstructions exammed.
Nevertheless, our study demonstrated
the lack of sensitivity of known
CT criteria when used alone and their
relative diagnostic values. Furthermore,
the use of a combination of
highly specific CT findings enabled
differentiation of simple and strangulated
obstruction in 85% of the patients
in this retrospective study. Prospective
application, however, of these
CT criteria may not yield results as
accurate as ours, as implied by Frager
etal(6).
In conclusion, the usefulness of
known CT criteria for aiding the diagnosis
of strangulated small-bowel obstruction
was examined with review
of CT scans of a large number of patients
with simple and strangulated
small-bowel obstructions. Statistical
analyses of various CT findings revealed
that poor or no contrast enhancement
of bowel wall, a serrated
beak, a large amount of ascites, an unusual
course of mesenteric vasculature,
diffuse engorgement of mesenteric
vasculature, and mesenteric
haziness were the most useful CT
findings for identifying strangulated
obstruction. Detection of a combination
of these CT findings increases the
diagnostic accuracy of CT to enable
differentiation of simple and strangulated
small-bowel obstructions. #{149}
Acknowledgment The authors thank Bonnie
Hami, MA, for editorial assistance in preparation
of the manuscript.
References
1. Balthazar EJ, Birnbaum BA, Megibow AJ,
Gordon RB, Whelan CA, Hulnick DH.
Closed-loop and strangulating intestinal
obstruction: CT signs. Radiology 1992; 185:
769-775.
2. Balthazar EJ. CT of small-bowel obstruction.
AJR 1994; 162:255-261.
3. Megibow AJ, Balthazar EJ, Cho KC, Medwid
SW, Birnbaum BA, Noz ME. Bowel
obstruction: evaluation with CT. Radiology
1991; 180:313-318.
4. Ha FIX, Park CH, Kim 5K, et al. CT analysis
of intestinal obstruction due to adhesions:
early detection of strangulation. J
Comput Assist Tomogr 1993; 17:386-389.
5. Ha HK. CT in the early detection of strangulation
in intestinal obstruction. Semin
Ultrasound CT MR 1995; 16:141-150.
6. Frager D, Baer JW Medwid SW, Rothpearl
A, Bossart P. Detection of intestinal ischemia
in patients with acute small-bowel
obstruction due to adhesions or hernia:
efficacy of CT. AJR 1996; 166:67-71.
7. Frager DH, Baer JW. Role of CT in evaluating
patients with small-bowel obstruction.
Semin Ultrasound CT MR 1995; 16:
127-140.
8. FisherJK. Computed tomographic diagnosis
of volvulus in intestinal mairotation.
Radiology 1981; 140:145-146.
9. WillsJS. Closed-loop and strangulating
obstruction of the small intestine: a new
twist. Radiology 1992; 185:635-636.
10. Fukuya T, Hawes DR. Lu CC, Chang PJ,
Barloon TJ. CT diagnosis of small-bowel
obstruction: efficacy in 60 patients. AJR
1992; 158:765-769.
11. Gazelle CS, Goldberg MA, WiftenbergJ,
Halpern EF, Pinkney L, Mueller PR Efficacy
of CT in distinguishing small-bowel
obstruction from other causes of smallbowel
dilatation. AJR 1994; 162:43-47.
12. Coblentz CL, Babcook CJ, Alton D, Riley BJ,
Norman C. Observer variation in detecting
the radiologic features associated with
bronchiolitis. Invest Radiol 1991; 26:115-
118.
13. Hosmer DWJr, Lemeshow S. Applied
logistic regression. New York, NY: Wiley,
1989; 1-23.
14. Laufman H, Nora PF. Physiological problems
underlying intestinal strangulation
obstruction. Surg Clin North Am 1966; 42:
219-229.
15. Leffall LDJr, QuanderJ, Syphax B. Strangulation
intestinal obstruction: a clinical
appraisal. Arch Surg 1965; 91:592-596.
16. Alpem MB, Glazer CM, FranCis IR. Ischemic
or infarcted bowel: CT findings. Radiology
1988; 166:149-152.
17. Federle M1 Chun C,Jeffrey RB, Rayor R
Computed tomographic findings in bowel
infarction. AJR 1984; 142:91-95.
18. Teefey SA, Roarke MC, BrinkJA, et al.
Bowel wall thickening: differentiation of
inflammation from ischemia with color
Doppler and duplex US. Radiology 1996;
198:547-551.
19. Zalcman M, Cansbeke V, Lalmand B,
Braude P, atj, StruyvenJ. Delayed
enhancement of the bowel wall: a new CT
sign of small-bowel strangulation. J Comput
Assist Tomogr 1996; 20:379-381.
20. SaItZSteIn EC, Marshall WJ, Freeark RJ.
Gangrenous intestinal obstruction Surg
CynecolObstet 1962; 114:694-699.
21. Gaines PA, Saunders AJS, Drake D. Midgut
malrotation diagnosed by ultrasound.
Clin Radiol 1987; 38:51-53.
512 #{149} Radiology August 1997