ARTICLE IN PRESS

Original Investigation

Comparative Effectiveness of
Imaging Modalities for the Diagnosis
of Intestinal Obstruction in Neonates
and Infants:
A Critically Appraised Topic
A. G. Carroll, MB, BCh, BAO, LRCPI, FFRRCSI, R. G. Kavanagh, MB, BSc, MCh, MRCSI,
C. Ni Leidhin, MB, BCh, BAO, MRCSI, N. M. Cullinan, MB, BCh, BAO, LRCSI, MRCPI,
L. P. Lavelle, MB, BCh, MRCPI, FFRRCSI, D. E. Malone, MD, FRCPI, FFRRCSI, FRCR, FRCPC

Abbreviations and Rationale and Objectives: The purpose of this study was to critically appraise and compare the di-
Acronyms agnostic performance of imaging modalities that are used for the diagnosis of upper and lower
gastrointestinal (GI) tract obstruction in neonates and infants.
MR
magnetic resonance Methods: A focused clinical question was constructed and the literature was searched using the patient,
intervention, comparison, outcome method comparing radiography, upper GI contrast study, and ul-
PICO trasound in the detection of upper GI tract obstruction such as duodenal atresia and stenosis, jejunal
patient, intervention, and ileal atresia, and malrotation and volvulus. The same methods were used to compare radiogra-
comparison, outcome phy and contrast enema in the detection of lower GI tract obstruction such as meconium plug syndrome,
MI meconium ileus, Hirschsprung disease, and imperforate anus. Retrieved articles were appraised and
meconium ileus assigned a level of evidence based on the Oxford University Centre for Evidence-Based Medicine hi-
US erarchy of validity for diagnostic studies.
ultrasound Results: There were no sensitivities/specificities available for the imaging diagnosis of duodenal atresia
or stenosis, jejunal or ileal atresias, meconium plug, and meconium ileus or for the use of cross-table
lateral radiography for the diagnosis of rectal pouch distance from skin in imperforate anus. The re-
trieved sensitivity for the detection of malrotation on upper GI contrast study is 96%, and the sensitivity
for the diagnosis of midgut volvulus on upper GI contrast study is 79%. The retrieved sensitivity and
specificity for the detection of malrotation with volvulus on ultrasound were 89% and 92%, respec-
tively. The retrieved sensitivity and specificity for the detection of Hirschsprung disease on contrast
enema were 70% and 83%, respectively. The retrieved sensitivity of invertogram for the diagnosis of
rectal pouch distance from skin in imperforate anus is 27%. The retrieved sensitivities of perineal ul-
trasound and colostography for the diagnosis of rectal pouch distance from skin in imperforate anus
were 86% and 100%, respectively.
Conclusions: There is limited evidence for the imaging diagnosis of duodenal atresia and stenosis,
jejunal and ileal atresias, meconium plug, meconium ileus, and imperforate anus, with recommended
practice based mainly on low-quality evidence or expert opinion. The available evidence supports the
use of upper GI contrast study for the diagnosis of malrotation and volvulus, with ultrasound as an
adjunct to diagnosis. Contrast enema is useful in the investigation of suspected Hirschsprung disease,
but a negative study does not outrule the condition. Colostography is the investigation of choice for
the work-up of infants with complex anorectal malformations before definitive surgical repair.
Key Words: Evidence-based medicine; upper gastrointestinal tract; lower gastrointestinal tract; small
bowel obstruction; large bowel obstruction; diagnosis; sensitivity and specificity; confidence interval;
duodenal atresia; duodenal stenosis; jejunal atresia; ileal atresia; malrotation; volvulus; meconium plug
syndrome; meconium ileus; Hirschsprung disease; imperforate anus.
© 2015 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.

Acad Radiol 2016; ■:■■–■■
From the Department of Radiology, St. Vincent’s University Hospital, Elm Park,
Merrion Rd, Dublin 4, Ireland (AG.C., RG.K., C.N.L., LP.L., DE.M.); Department
of Pediatrics, Our Lady’s Children’s Hospital, Crumlin, Dublin 12, Ireland (NM.C.).
Received December 9, 2015; revised December 9, 2015; accepted December
27, 2015. Address correspondence to: A.G.C. e-mail: annegcarroll@gmail.com
© 2015 The Association of University Radiologists. Published by Elsevier Inc.
All rights reserved.
http://dx.doi.org/10.1016/j.acra.2015.12.014

1
CARROLL ET AL
INTRODUCTION
Duodenal atresia occurs in 1 in 5000–10,000 live births. In
one-third of cases, it is an isolated anomaly. In one-third of
cases, it is associated with Down syndrome. It can also be as-
sociated with other congenital abnormalities, especially cardiac
and esophageal anomalies. It results from failure of the du-
odenum to recanalize during weeks 8–10 of gestation (1).
Neonates typically present with bilious or nonbilious vom-
iting in the first 24 hours after birth.
Infants with less severe duodenal stenosis may have delayed
presentation with frequent vomiting and failure to thrive. Du-
odenal stenosis is often due to the presence of a duodenal web
or may be secondary to extrinsic compression from an annular
pancreas, malrotation, or a preduodenal portal vein (2).
Similarly, the even rarer jejunal and ileal atresias (3) also
present with bilious vomiting and abdominal distension, usually
in the first 24 hours after birth. They are thought to result
from a vascular incident causing intestinal necrosis in utero
and are less commonly diagnosed antenatally than duodenal
atresia.
Malrotation is a rotational anomaly of the embryonic gut
that leads to abnormal positioning of the duodenal-jejunal (D-
J) junction. The D-J junction is not in its normal position
to the left of the left-sided vertebral pedicles at the level of
the duodenal bulb on the frontal view and posterior (retro-
peritoneal) on the lateral view of an upper gastrointestinal (GI)
contrast study (4). Often, the small bowel is located in the
right side of the abdomen and the colon including the cecum
is on the left side of the abdomen. The position of the cecum
can be variable and may be subhepatic or in the midline (5).
Small bowel volvulus is a complication of malrotation, where
the abnormally narrow base or pedicle of the small bowel mes-
entery becomes twisted around itself, leading to obstruction
at the level of the duodenum, followed by intestinal isch-
emia if prompt surgical treatment is not provided. It should
be suspected in any infant with bilious vomiting. Physical ex-
amination may demonstrate diffuse abdominal tenderness with
or without signs of peritonitis, abdominal distension, and blood
in the stool on rectal examination (a late sign suggestive of
bowel necrosis) (6). Delayed treatment is associated with high
mortality rates or severe morbidity secondary to short gut syn-
drome if the child survives (5).
Malrotation occurs in approximately 1 in 500 live births
(7). Although a previously undiagnosed patient with malro-
tation may present at any age with an acute volvulus,
approximately 80% of patients with malrotation will present
in the first month of life (7,8). Of those presenting in the first
month of life, most will do so in the first week (9). Some
series report that 10–20% present after 1 year of age (4,10–12).
Meconium plug syndrome is a relatively benign cause of
large bowel obstruction in neonates who have delayed passage
of meconium in the first 24–72 hours of life. Although it can
resolve spontaneously, a contrast enema is often required. It
is associated with small left colon syndrome, which also usually
resolves after contrast enema. Most of the time, there are no
2
Academic Radiology, Vol ■, No ■, ■■ 2016
further problems. An association with Hirschsprung disease
has been reported in 13–38% of cases (13,14).
Meconium ileus (MI) is a common presentation of cystic
fibrosis (CF) in the neonatal period. MI is the presenting feature
in approximately 10% of patients with CF, and 80–90% of
patients who have MI are subsequently diagnosed with CF
(15). Infants present with bowel obstruction and failure to pass
meconium in the first 3 days of life.
Hirschsprung disease is a functional obstruction of the bowel
caused by the absence of intrinsic enteric ganglion cells. It
usually presents with failure to pass meconium or delayed
passage of meconium in the newborn period. Infants may also
present with abdominal distension and bilious vomiting. It may
involve variable lengths of colon extending back from the
rectum, and rarely can involve the entire colon.
Anorectal malformations/imperforate anus comprise a wide
spectrum of anomalies of the anus and rectum and often also
of the urinary tract and genitalia. They can vary from simple
malformations, which are easily corrected with good func-
tional outcomes, to complex anomalies requiring multiple
complex surgeries with poorer outcomes. Imperforate anus
is most commonly associated with the VACTERL associa-
tion (vertebral anomalies, anal atresia, cardiac defects,
tracheoesophageal fistula and/or esophageal atresia, renal and
radial anomalies, limb defects), and occurs with increased fre-
quency in trisomy 21 (16). The mainstay of diagnosis of
anorectal malformations is from physical examination of the
neonate. At 16–36 hours post birth, neonates are assessed for
the presence of a fistula to the perineum or to the urinary
or genital tracts and for associated anomalies. In the major-
ity of patients, a colostomy will be performed and further work-
up will be arranged after the colostomy has healed. However,
in certain cases such as rectoperineal fistula or rectovestibular
fistula in girls, colostomy can be avoided and anoplasty is all
that is needed. Ultrasound and urine analysis are used to outrule
obstructive uropathy, hydrocolpos, or rectovesical fistula.
ASK
We questioned whether radiography, upper GI contrast study,
or ultrasound perform better for the detection of upper GI
tract obstruction and for the diagnosis of duodenal atresia and
stenosis, jejunal and ileal atresia, and malrotation and volvu-
lus. Before undertaking this review using evidence-based
methods, we note the prevailing popular opinion (before using
explicit critical appraisal) that plain radiography is sufficient
for the diagnosis of duodenal, jejunal, and ileal atresias, whereas
upper GI contrast study is the investigation of choice for mal-
rotation and volvulus.
Similarly, we questioned whether radiography or contrast enema
has a better diagnostic performance for the detection of meco-
nium plug, MI, Hirschsprung disease, and imperforate anus (the
common causes of lower GI tract obstruction in neonates).
We constructed a focused standardized patient, investiga-
tion, comparison, and outcome question to search the available
literature as follows: (1) “In neonates and infants with
Academic Radiology, Vol ■, No ■, ■■ 2016
suspected upper GI tract obstruction, how does radiography
vs. upper GI contrast study vs. ultrasound compare to each
other for detection of upper GI tract obstruction and diag-
nosis?” and (2) “In neonates and infants with suspected lower
GI tract obstruction, how does radiography vs. contrast enema
compare to each other for detection and diagnosis?”
SEARCH
Secondary and primary evidence were searched according to
the evidence pyramid described by Haynes (17). For second-
ary literature (eg, information systems, synopses, and syntheses)
(18), we searched Up To Date (19), The American College
of Physicians Journal Club (20), Cochrane Collaboration Library
(21), Turning Research Into Practice (22), and PubMed (23).
For primary literature (original studies), we searched PubMed
(23) and EMBASE (24).
A comprehensive Medline search (United States National
Library of Medicine database) for original articles published
between May 2005 and May 2015 using the Ovid and PubMed
search engines was performed using a combination of the fol-
lowing key terms: ((“diagnosis”) AND (“small bowel
obstruction”) AND (radiography OR upper GI contrast study
OR ultrasound)) and ((“diagnosis”) AND (“large bowel ob-
struction”) AND (radiography OR contrast enema)) and
((“diagnosis”) AND (“duodenal atresia” OR “jejunal atresia”
OR “ileal atresia” OR “malrotation and volvulus” OR “me-
conium plug” OR “meconium ileus” OR “Hirschsprung”)
AND (“evidence based”)) and “imperforate anus” OR “ano-
rectal malformation” AND “diagnosis” AND “radiograph”
OR “colostography” OR “colostogram” OR “magnetic res-
onance imaging,” (Fig 1).
The search was limited to English-language articles, human,
and pediatric studies. The abstracts were reviewed and se-
lected based on well-designed methodology, clinical trials,
outcomes, and diagnostic accuracy. Additional relevant ar-
ticles were selected from the references of reviewed articles
and published guidelines.
APPRAISE
Upper GI Tract
The evidence base for the imaging of intestinal obstruction
in the pediatric population is limited. Computed tomogra-
phy is generally avoided because of concerns regarding radiation
exposure. Radiography, contrast studies, and ultrasound are
commonly used to assist with diagnosis.
Neonates and young children with intestinal obstruction
deteriorate rapidly, and prompt clinical and radiological de-
tection of obstruction is essential to reduce morbidity and
mortality. Before radiological investigation and treatment, it
is essential that children are adequately resuscitated and sta-
bilized (25). Children who are medically unstable and in whom
there is a high clinical suspicion of malrotation with volvu-
lus should proceed directly to surgery (6).
COMPARATIVE EFFECTIVENESS OF IMAGING MODALITIES
Duodenal Atresia and Stenosis, Jejunal, and Ileal Atresia
In many cases of congenital intestinal atresia and obstruc-
tion, the diagnosis is made on antenatal ultrasound (26–28).
Patients may present with polyhydramnios due to fetal bowel
atresias. Findings such as the “double bubble sign” of gastric
and duodenal dilatation suggest the diagnosis of duodenal atresia
(1). For the purpose of this study, we will focus on the proven
diagnostic modalities used after birth, which include plain ra-
diography and upper GI contrast study.
Radiography.—On searching the primary literature, we found
a retrospective observational study entitled “Congenital Duo-
denal Obstruction in Neonates: A Decade’s Experience From
One Center” by Chen et al. This was assigned level 4 evidence
using the Oxford Center for Evidence-Based Medicine (29). The
evidence regarding the diagnosis of duodenal, jejunal, and ileal
atresias is extremely limited and the literature is predomi-
nantly expert opinion (level 5 evidence). No studies provided
a sensitivity or specificity for the use of imaging in the di-
agnosis of duodenal atresia/stenosis or jejunal or ileal atresia.
The mainstay of diagnosis of duodenal atresia and stenosis
in a neonate is radiography showing the “double bubble sign”
of gastric and proximal duodenal dilatation and an absence
of gas within the distal bowel (30). This sign was demon-
strated in 68.6% of patients presenting with congenital duodenal
obstruction in one retrospective observational study (31). In
partial duodenal obstruction or duodenal stenosis, an air-
fluid level in the stomach may be demonstrated.
Similarly, the even rarer jejunal and ileal atresias are usually
first identified on radiograph showing dilated air and fluid-
filled bowel loops down to mid or distal small bowel level
with no gas beyond the level of obstruction or in the colon.
All infants with intestinal obstruction should have a supine
anteroposterior abdominal radiograph and a lateral decubi-
tus radiograph to outrule pneumoperitoneum (1).
Upper GI contrast study.—It is recommended that all medi-
cally stable suspected cases of duodenal atresia are followed
up with an upper GI contrast study to outrule malrotation
and volvulus masquerading as a double bubble sign (1). In
infants who have had the stomach decompressed either through
vomiting or through aspiration of a nasogastric tube, air can
be injected down the nasogastric tube to act as its own con-
trast agent and reproduce the double bubble sign. Barium is
useful to determine if the patient has duodenal atresia or du-
odenal stenosis.
Upper GI contrast study with barium may be useful in cases
of jejunal or ileal atresia to differentiate atresia from stenosis.
Contrast enema.—It is recommended that all medically stable
suspected cases of jejunal and ileal atresia are followed up with
a contrast enema to differentiate between MI or Hirschsprung
disease and intestinal atresia. All three pathologies may have
an associated microcolon. In cases of atresia, the contrast does
not reflux back into the dilated proximal bowel loops around
the inspissated stool (1).
3
CARROLL ET AL
Figure 1. Search strategy using patient, investigation, comparison, outcome (PICO)-focused keywords.
Malrotation and Volvulus
Radiography.—A plain radiograph is often performed as a first-
line investigation to outrule gross pneumoperitoneum or
obvious lower GI obstruction as an alternative diagnosis. The
most common finding in malrotation and volvulus is a gasless
abdomen, but occasionally a double bubble sign (like in du-
odenal atresia) is demonstrated (6).
Upper GI contrast study.—On searching the secondary litera-
ture, we found a study entitled “Intestinal Malrotation” on
www.uptodate.com. This recommends upper GI contrast study
as the investigation of choice for children with suspected mal-
rotation and volvulus, based on a retrospective cohort study
using surgical findings as the gold standard (32). This was
4
Academic Radiology, Vol ■, No ■, ■■ 2016
assigned level 3b evidence using the Oxford Center for
Evidence-Based Medicine (29). In this study, the sensitivity
of the upper GI contrast study for the diagnosis of malrota-
tion was 96% and the sensitivity for the diagnosis of midgut
volvulus was 79% (32). The study looked only at patients who
went to surgery for suspected malrotation and/or volvulus;
therefore, negative studies that did not go to surgery were
not assessed and an accurate specificity could not be obtained.
Findings at upper GI contrast study in malrotation include
malposition of the duodenum (D-J junction is not in its normal
position to the left of the left-sided vertebral pedicles at the
level of the duodenal bulb on the frontal view and posterior
on the lateral view) and duodenal obstruction similar to the
appearance in duodenal atresia. In acute volvulus, a beak sign
Academic Radiology, Vol ■, No ■, ■■ 2016
of duodenal obstruction is often seen, whereas the cork-
screw sign of the twisted bowel may or may not be
demonstrated.
In up to 25% of cases, the findings at upper GI contrast
study can be subtle or equivocal (6). In these cases, contrast
enema or ultrasound may be useful as an adjunct to assist in
making the diagnosis.
Contrast enema.—If the study is inconclusive, a contrast enema
may be performed to confirm the position of the cecum, aiding
the diagnosis of malrotation (33). The position of the cecum
is variable and therefore contrast enema cannot exclude mal-
rotation. If the imaging findings are inconclusive, but there
is a strong suspicion of volvulus, the patient should proceed
to exploratory surgery (11).
Ultrasound.—There is emerging evidence of the usefulness of
ultrasound in detecting malrotation and small bowel volvu-
lus, by demonstrating abnormal orientation of the superior
mesenteric vessels.
On searching the secondary literature, we found a study
entitled “Sonographic Features Related to Volvulus in Neo-
natal Intestinal Malrotation” by Chao et al. This was assigned
level 2b evidence using the Oxford Center for Evidence-
Based Medicine (29).
In Chao et al.’a 3-year prospective study of abdominal and
color Doppler ultrasonography in neonates with suspected mal-
rotation or malrotation with volvulus, water instillation was
used to detect duodenal dilatation, edema, and malrotated small
bowel. They found that duodenal dilatation with tapering con-
figuration, fixed midline bowel, and mesenteric whirlpool sign
were all suggestive of volvulus. The sensitivity and specific-
ity of duodenal dilatation with tapering configuration for the
detection of volvulus were 89% and 92%, respectively (Table 1
and Fig 2a); of fixed midline bowel were 89% and 92%
(Fig 2b); of whirlpool sign were 89% and 92% (Fig 2c); and
of dilatation of distal superior mesenteric vein were 56% and
73% (Fig 2d) (34). Other studies have used ultrasound to look
at the position of the superior mesenteric vein, which can be
seen on the left of the superior mesenteric artery in malro-
tation. However, a normally placed superior mesenteric vein
does not outrule malrotation (35,36).
TABLE 1. Test Performances for the Diagnosis of Upper and Lower GI Tract Obstruction in Neonates and Infants
Upper Gastrointestinal (GI) Tract
US with Fixed Midline
Selected Study US Volvulus Bowel US with Whirlpool Sign
Sensitivity, % 89 89
Specificity, % 92 92
Accuracy 0.81 0.81
SMV, superior mesenteric vein; US, ultrasound.
COMPARATIVE EFFECTIVENESS OF IMAGING MODALITIES
Lower GI Tract
In a vomiting neonate, the initial abdominal radiograph allows
differentiation between upper and lower GI causes and guides
subsequent investigations, with contrast enema as the inves-
tigation of choice for lower GI pathology.
Meconium Plug Syndrome and MI
Although prenatal diagnosis of MI is often suggested by ul-
trasound demonstrating hyperechoic bowel, the definitive
diagnosis is made after birth, as echogenic bowel can be a
normal transient occurrence in utero (37).
Radiography.—A plain radiograph is usually the first investi-
gation in the work-up of both meconium plug and MI.
In meconium plug, plain radiography shows inspissated me-
conium pellets in the large bowel only (14).
In MI, a plain radiograph demonstrates dilated small bowel
loops with the inspissated meconium in the distal ileum and
the large bowel. Plain radiography is necessary to assess for
perforation, calcifications (evidence of prenatal perforation and
meconium peritonitis), or other abnormalities. Evidence of
perforation indicates complex MI and precludes an attempt
at nonoperative decompression. In this case, the infant should
proceed directly to surgery.
Contrast enema.—Water-soluble contrast enemas with
hyperosmolar contrast have long been established as the di-
agnostic modality of choice for meconium plug (hypoplastic
left colon syndrome) and also for MI (38). Gastrografin is most
commonly used. It is meglumine diatrizoate, a hyperosmolar,
water-soluble, radiopaque solution containing 0.1% polysor-
bate 80 (Tween 80) and 37% organically bound iodine. The
osmolarity of the solution is 1900 mOsm/L. Upon instilla-
tion, fluid is drawn into the intestinal lumen, hydrating and
softening the meconium mass (37). Aside from demonstrat-
ing the level of obstruction and outlining the meconium, they
also have the therapeutic effect of often relieving the ob-
struction and avoiding the need for surgery (39).
In patients with MI, the contrast enema typically reveals
a small-caliber colon (microcolon of disuse) and meconium
pellets in the terminal ileum. The ileum proximal to the ob-
struction is dilated. For successful decompression, contrast should
be seen refluxing into the terminal ileum. Follow-up
Lower GI Tract
Contrast Enema for
US with SMV Dilatation Hirschsprung Disease
89 56 70
92 73 83
0.81 0.60 0.70
5
6

CARROLL ET AL
Academic Radiology, Vol ■, No ■, ■■ 2016
Figure 2. Graph of conditional probabilities for ultrasound diagnosis of upper gastrointestinal (GI) obstruction in neonates and infants. Graph of conditional probability and area under the
curve (AUC). The x-axis represents the pretest probability of the disease in question, and the y-axis represents the posttest probability. The pretest probability is chosen, and a vertical line
is drawn from the corresponding point on the x-axis to the “test-negative” or “test-positive” curve. From this point of intersection, a horizontal line is drawn to the y-axis. The posttest
probability is the value at the point where the horizontal line intersects the y-axis. (a) Graph of conditional probabilities for ultrasound diagnosis of volvulus in upper GI obstruction in
neonates and infants. Sensitivity = 0.89; specificity = 0.92; and AUC = 0.81. (b) Graph of conditional probabilities for ultrasound detection of fixed midline bowel in upper GI obstruction in
neonates and infants. Sensitivity = 0.89; specificity = 0.92; and AUC = 0.81. (c) Graph of conditional probabilities for ultrasound detection of the whirlpool sign in upper GI obstruction in
neonates and infants. Sensitivity = 0.89; specificity = 0.92; and AUC = 0.81. As we can see from the above three graphs (a–c), the high sensitivity and specificity means that ultrasound is
reliable in ruling in and ruling out volvulus and can reliably detect the presence or absence of a fixed midline bowel and the “whirlpool” sign. (d) Graph of conditional probabilities for
ultrasound detection of superior mesenteric vein dilatation in upper GI obstruction in neonates and infants. Sensitivity = 0.56; specificity = 0.73; and AUC = 0.60. As we can see from this
graph, the available evidence suggests that ultrasound cannot be reliably used to assess the presence or absence of superior mesenteric vein (SMV) dilatation in the setting of upper GI
obstruction in neonates and infants.
Academic Radiology, Vol ■, No ■, ■■ 2016 COMPARATIVE EFFECTIVENESS OF IMAGING MODALITIES
radiographs are advised to outrule perforation as a compli-
cation and to assess for resolution of the obstruction. The enema
can be repeated once or twice at 12–24 hour intervals if nec-
essary to completely relieve the obstruction.
On searching the secondary literature, we found a study
entitled “‘Pulling the Plug’—Management of Meconium Plug
Syndrome in Neonates” by Cuenca et al. (40). This was as-
signed level 4 evidence using the Oxford Center for Evidence-
Based Medicine (29). In their study, 30% of patients had
spontaneous resolution of the meconium plug without any
specific treatment. Of those infants requiring treatment, barium
enema was successful in 97% and only one enema was re-
quired in 98% of successful cases. On searching the primary
and secondary literature, we found no studies of sensitivity
or specificity for contrast enema in the diagnosis of meco-
nium plug.
The success rate of patients with uncomplicated MI, treated
with Gastrografin enemas, historically range as high as 83%
(37). Several more recent studies report much lower success
rates closer to 36–39% (39,41), with corresponding lower com-
plication rates, possibly due to a more conservative approach
in terms of number of enemas performed per patient and pres-
sures used. However, these studies are of small size and are
prone to bias. Operative treatment is associated with signif-
icant morbidity and mortality (41). On searching the secondary
literature, we found no studies of sensitivity or specificity for
contrast enema in the diagnosis of MI.
Hirschsprung Disease
Contrast enema.—Contrast enema is the imaging modality of
choice for the diagnosis of Hirschsprung disease. It helps as-
certain the length of aganglionic colon and aids in surgical
planning. Washing out of the distal colon with enemas should
be avoided before the contrast enema as it may distort a low
Figure 3. Graph of conditional probabili-
ties for contrast enema for the diagnosis of
Hirschsprung disease in lower gastrointes-
tinal (GI) obstruction in neonates and infants.
Sensitivity = 0.70; specificity = 0.83; area under
the curve (AUC) = 0.70. Graph of condition-
al probability and AUC. The x-axis represents
the pretest probability of the disease in ques-
tion, and the y-axis represents the posttest
probability. The pretest probability is chosen,
and a vertical line is drawn from the corre-
sponding point on the x-axis to the “test-
negative” or “test-positive” curve. From this
point of intersection, a horizontal line is drawn
to the y-axis. The posttest probability is the
value at the point where the horizontal line
intersects the y-axis. As we can see from this
graph, contrast enema is reliable in patients
with a high pretest probability and a posi-
tive test. It is not so reliable in patients with
a high pretest probability and a negative test.
These patients will require further investigation.
transition zone. The catheter should be placed just inside the
anus without inflation of the balloon to avoid distortion of a
low transition zone and perforation. Radiographs are taken
immediately after hand injection of contrast and 24 hours later.
Positive findings include a transition zone with proximal dilated
bowel, a microcolon, or retention of contrast in the bowel
24 hours post the enema on the delayed radiograph. Con-
trast enema is less reliable in the neonatal period than in older
infants (42).
On searching the secondary literature, we found a study
entitled “Diagnostic Tests in Hirschsprung Disease: A Sys-
tematic Review” by de Lorijn et al. This was assigned level
3a evidence using the Oxford Center for Evidence-Based Med-
icine (29). This systematic review showed that contrast enema
had a sensitivity of 70% and specificity of 83% (12 studies for
a total of 425 patients; Table 1 and Fig 3), which was sig-
nificantly lower than those of rectal suction biopsy and anorectal
manometry (43). Further studies have confirmed this (44–46).
Imperforate Anus
In approximately 10% of patients, there are no clinical signs
to reveal the location of the anorectal anomaly after 24 hours
and radiography can be useful in this group. A prone cross-
table lateral radiograph can be performed to look for air in
the rectum and identify the distance of the rectum from the
imperforate anus at the skin. If the distance is <1 cm, ano-
plasty can be performed rather than colostomy in the absence
of urinary tract or genital tract abnormalities (47).
Radiography.—On searching the secondary literature, we found
a study entitled “Prone Cross-table Lateral View: An Alter-
native to the Invertogram in Imperforate Anus” by
Narasimharao et al. (48). This was assigned level 4 evidence
using the Oxford Center for Evidence-Based Medicine (29).
7
CARROLL ET AL
This showed that cross-table lateral radiograph is superior to
invertogram radiograph in delineating the most caudal point
of the rectum in neonates. Prone cross-table lateral radio-
graphs were performed with the infant in prone position with
the hips flexed in a genupectoral position for 3 minutes before
a radiograph centered over the greater trochanters. Invertograms
were performed holding the infant upside down for 3 minutes
before a radiograph centered over the greater trochanters. Infants
were much less distressed in the prone lateral position.
However, one study showed a low sensitivity of invertogram
of only 27% when compared to surgical findings as a gold
standard (49).
Infants with imperforate anus should also have radiogra-
phy of the spine and sacrum performed to assess for associated
anomalies.
Ultrasound.—Ultrasound of the perineum has been shown to
have a sensitivity of 86% in assessing the bowel to skin mea-
surement distance when compared to surgical findings as the
gold standard (49). This is far superior to plain radiography.
Ultrasound of the abdomen should be performed to outrule
associated urinary obstruction or hydrocolpos or a persistent
cloaca. Ultrasound of the spine may be performed in infants
less than 3 months to assess for associated tethered cord or
other spinal anomalies (47).
Colostography.—Colostography is a contrast study of the distal
colon through the mucus fistula in patients who have had a
colostomy for imperforate anus. Water soluble contrast is in-
jected by hand at a pressure sufficient to distend the distal
rectum and demonstrate any fistulas. The injection is con-
tinued until the infant voids, and images are taken during
micturition to reveal, in a single radiograph, the sacrum, rectum
height, perineum, fistula location, bladder, vesicoureteral reflux
(if present), and urethra (47). This is very important in pre-
operative planning (50). Sensitivity of augmented pressure distal
colostography has been shown to be 100% in one study using
surgical findings as the gold standard (49).
Magnetic Resonance Imaging (MRI)
MRI is being increasingly used in the evaluation of infants
with complex anorectal malformations and fistulas (51). Pre-
natal MRI often delineates the anatomy and can distinguish
high anorectal malformation from cloaca before birth (52).
It is useful at detecting the presence of fluid in the rectum
suggesting a rectovesical fistula.
APPLY
Upper GI Tract
Duodenal Atresia and Stenosis, Jejunal, and Ileal Atresia
The evidence for imaging of duodenal atresia/stenosis and
jejunal and ileal atresia is very limited. The recommenda-
tions, based on low-quality evidence, are to commence with
plain radiography. Contrast swallow and contrast enema should
8
Academic Radiology, Vol ■, No ■, ■■ 2016
also be performed to outrule alternate pathology that may mas-
querade as a more proximal atresia.
Malrotation and Volvulus
The evidence supports the use of upper GI contrast study for
the diagnosis of malrotation ± volvulus, once the infant is stable
and perforation has been outruled on plain radiography. More
research is needed to evaluate the specificity of upper GI con-
trast study, to allow accurate comparison to ultrasound for the
ruling in of malrotation and volvulus. Equivocal cases are
common after upper GI contrast study, and ultrasound and
contrast enema are used as adjuncts to assist in the diagnosis
of these cases.
Lower GI Tract
Meconium Plug Syndrome and MI
The evidence for the postnatal diagnosis of meconium plug
and MI is severely limited and appears to be based mainly on
consensus opinion or low-quality evidence. The use of plain
radiography for the diagnosis of large bowel obstruction, fol-
lowed by hyperosmolar contrast enema under fluoroscopic
guidance for diagnosis and treatment, is the accepted algo-
rithm for the investigation and management of meconium plug
and simple, uncomplicated MI. Repetition of the enema may
be required to fully relieve the obstruction. Infants with ev-
idence of perforation or meconium peritonitis should proceed
directly to surgery after plain radiography and should not
undergo contrast enema.
Hirschsprung Disease
The main imaging tool for the diagnosis of Hirschsprung disease
is contrast enema. It is inferior to both rectal suction biopsy
and anorectal manometry in terms of sensitivity and speci-
ficity (43). When positive, contrast enema is a useful test and
less invasive than rectal suction biopsy. Contrast enema is limited
by false-negative results and cannot reliably outrule Hirschsprung
disease, especially in cases with a high pretest probability and
a negative test (Fig 3).
Imperforate Anus
The investigation and management of imperforate anus and
the associated malformations depends on the clinical find-
ings in the neonate, the results of urinalysis, and assessment
of associated anomalies such as VACTERL or trisomy 21. In
a select group of patients with no clinical signs to reveal the
location of the anorectal anomaly after 24 hours, cross-table
lateral radiographs are useful when they show air in the rectum
<1 cm from the skin. Ultrasound is important to assess for as-
sociated urinary or genital tract anomalies, the presence of which
will impact on the surgical management. Perineal ultra-
sound has been shown to be superior to radiography in
delineating the location of the distal rectal pouch.
The majority of patients will require diversion colos-
tomy. After the colostomy has healed, this group benefit from
preoperative augmented pressure distal colostography to de-
lineate the surgical anatomy and identify fistulas prior to
Academic Radiology, Vol ■, No ■, ■■ 2016
definitive surgical repair. MRI has a role in prenatal diagno-
sis and an emerging role in postnatal preoperative evaluation,
especially in complex cases.
EVALUATE
Upper GI Tract
Duodenal Atresia and Stenosis, Jejunal, and Ileal Atresia
More research is needed into the optimal imaging of these
rare upper GI atresias and stenosis. At present, we should follow
the algorithm of plain abdominal radiography, followed by
contrast swallow and contrast enema, based on the limited
low-quality evidence.
Malrotation and Volvulus
With regard to malrotation and acute volvulus, the investi-
gation of choice is an upper GI contrast study (6,36,53). Plain
radiography is usually the initial investigation and is useful to
outrule obvious perforation, in which case the child should
proceed directly to surgery. Other useful adjuncts to the upper
GI contrast study are ultrasound and barium enema, espe-
cially in equivocal cases (6).
Lower GI Tract
Meconium Plug Syndrome and MI
More research is needed into the optimal imaging of meco-
nium plug and MI, as well as the use of contrast enema to
relieve the obstruction. At present, it is recommended to follow
the established algorithms of plain radiography followed by
hyperosmolar contrast enema under fluoroscopic guidance for
the diagnosis and treatment of meconium plug syndrome and
uncomplicated, simple MI.
Hirschsprung Disease
There is still a role for contrast enema in the work-up of infants
with suspected Hirschsprung disease. It is less invasive than
rectal suction or full thickness biopsies and, when positive,
can guide treatment. Radiologists and referring physicians should
be aware of its pitfalls and seek further investigations in pa-
tients with negative enemas whose clinical symptoms give them
a high pretest probability.
Imperforate Anus
Multiple imaging modalities including radiography, ultra-
sound, augmented pressure distal colostography, and MRI have
a role to play in the work-up of imperforate anus, depend-
ing on the clinical findings and presence of associated
malformations. More research is needed to compare these mo-
dalities and identify the relevant sensitivities and specificities
in the different cohorts of these patients.
REFERENCES
1. Wesson DE. Intestinal atresia. Version 11.0. Wolters Kluwer, Up to date
2014. Available at: http://uptodate.com.
COMPARATIVE EFFECTIVENESS OF IMAGING MODALITIES
2. Gomella TC, Cunningham M, Eyal FG. Neonatology: management, pro-
cedures, on-call problems, diseases, and drugs. 6th ed. New York: McGraw
Hill, 2009.
3. Best KE, Tennant PW, Addor MC, et al. Epidemiology of small intesti-
nal atresia in Europe: a register-based study. Arch Dis Child Fetal Neonatal
Ed 2012; 97:F353–F358.
4. Applegate KE, Anderson JM, Klatte EC. Intestinal malrotation in chil-
dren: a problem-solving approach to the upper gastrointestinal series.
Radiographics 2006; 26:1485–1500.
5. Millar AJ, Rode H, Cywes S. Malrotation and volvulus in infancy and child-
hood. Semin Pediatr Surg 2003; 12:229–236.
6. Brandt ML. Intestinal malrotation. Wolters Kluwer, Up to date 2015. Avail-
able at: http://www.uptodate.com.
7. Torres AM, Ziegler MM. Malrotation of the intestine. World J Surg 1993;
17:326–331.
8. Filston HC, Kirks DR. Malrotation—the ubiquitous anomaly. J Pediatr Surg
1981; 16(4 suppl 1):614–620.
9. Strouse PJ. Disorders of intestinal rotation and fixation (“malrotation”).
Pediatr Radiol 2004; 34:837–851.
10. Maxson RT, Franklin PA, Wagner CW. Malrotation in the older child: sur-
gical management, treatment, and outcome. Am Surg 1995; 61:135–
138.
11. Powell DM, Othersen HB, Smith CD. Malrotation of the intestines in chil-
dren: the effect of age on presentation and therapy. J Pediatr Surg 1989;
24:777–780.
12. Spigland N, Brandt ML, Yazbeck S. Malrotation presenting beyond the
neonatal period. J Pediatr Surg 1990; 25:1139–1142.
13. Burge D, Drewett M. Meconium plug obstruction. Pediatr Surg Int 2004;
20:108–110.
14. Keckler SJ, St Peter SD, Spilde TL, et al. Current significance of meco-
nium plug syndrome. J Pediatr Surg 2008; 43:896–898.
15. Sabharwal S, Borowitz D. Cystic fibrosis: overview of gastrointestinal
disease. Up to date 2015.
16. Cuschieri A. Anorectal anomalies associated with or as part of other
anomalies. Am J Med Genet 2002; 110:122–130.
17. Haynes RB. Of studies, summaries, synopses, and systems: the “4S”
evolution of services for finding current best evidence. Evid Based Ment
Health 2001; 4:37–39.
18. Kerr JM. Small bowel imaging: CT enteroclysis or barium enteroclysis?
Critically appraised topic. Abdom Imaging 2008; 33:31–33.
19. Daskalogiannaki M, Prassopoulos P, Katrinakis G, et al. Splenic involve-
ment in lymphomas. Evaluation on serial CT examinations. Acta Radiol
2001; 42:326–332.
20. American College of Physicians Journal Club Web site. Available at:
http://www.acpjc.org. Accessed February 7, 2012.
21. Deignan RW, Malone DE, Taylor S, et al. Improving visualization of distal
and terminal ileum during the small bowel meal: an evaluation of fluo-
roscopic manoeuvres. Clin Radiol 1995; 50:548–552.
22. Gourtsoyiannis N, Grammatikakis J, Prassopoulos P. Role of conven-
tional radiology in the diagnosis and staging of gastrointestinal tract
neoplasms. Semin Surg Oncol 2001; 20:91–108.
23. Elsayes KM, Al-Hawary MM, Jagdish J, et al. CT enterography: prin-
ciples, trends, and interpretation of findings. Radiographics 2010; 30:1955–
1970.
24. Thipphavong S, Millward SF, Elliott JA. Balloon angioplasty versus primary
stenting of ostial atherosclerotic renal arterial stenoses: critically ap-
praised topic. Can Assoc Radiol J 2004; 55:108–110.
25. Hajivassiliou CA. Intestinal obstruction in neonatal/pediatric surgery. Semin
Pediatr Surg 2003; 12:241–253.
26. Virgone C, D’antonio F, Khalil A, et al. Accuracy of prenatal ultrasound
in detecting jejunal and ileal atresia: systematic review and meta-
analysis. Ultrasound Obstet Gynecol 2015; 45:523–529.
27. Brantberg A, Blaas HG, Salvesen KA, et al. Fetal duodenal obstruc-
tions: increased risk of prenatal sudden death. Ultrasound Obstet Gynecol
2002; 20:439–446.
28. Grosfeld JL, Rescorla FJ. Duodenal atresia and stenosis: reassess-
ment of treatment and outcome based on antenatal diagnosis,
pathologic variance, and long-term follow-up. World J Surg 1993; 17:301–
309.
29. Ha HK, Ko GY, Yu ES, et al. Intestinal tuberculosis with abdominal com-
plications: radiologic and pathologic features. Abdom Imaging 1999; 24:32–
38.
30. Kadian YS, Rattan KN. Congenital pyloric atresia with distal duodenal
atresia—role of CT scan. J Neonatal Surg 2014; 3:37.
9
CARROLL ET AL
31. Chen QJ, Gao ZG, Tou JF, et al. Congenital duodenal obstruction in neo-
nates: a decade’s experience from one center. World J Pediatr 2014;
10:238–244.
32. Sizemore AW, Rabbani KZ, Ladd A, et al. Diagnostic performance of the
upper gastrointestinal series in the evaluation of children with clinically
suspected malrotation. Pediatr Radiol 2008; 38:518–528.
33. Stringer DA. Pediatric gastrointestinal imaging. 1st ed. Ontario: BC Decker,
1989.
34. Chao HC, Kong MS, Chen JY, et al. Sonographic features related to vol-
vulus in neonatal intestinal malrotation. J Ultrasound Med 2000; 19:371–
376.
35. Dufour D, Delaet MH, Dassonville M, et al. Midgut malrotation, the re-
liability of sonographic diagnosis. Pediatr Radiol 1992; 22:21–23.
36. Quail MA. Question 2. Is Doppler ultrasound superior to upper gastro-
intestinal contrast study for the diagnosis of malrotation? Arch Dis Child
2011; 96:317–318.
37. Carlyle BE, Borowitz DS, Glick PL. A review of pathophysiology and man-
agement of fetuses and neonates with meconium ileus for the pediatric
surgeon. J Pediatr Surg 2012; 47:772–781.
38. Noblett HR. Treatment of uncomplicated meconium ileus by Gastrografin
enema: a preliminary report. J Pediatr Surg 1969; 4:190–197.
39. Copeland DR, St Peter SD, Sharp SW, et al. Diminishing role of con-
trast enema in simple meconium ileus. J Pediatr Surg 2009; 44:2130–
2132.
40. Cuenca AG, Ali AS, Kays DW, et al. “Pulling the plug”—management of
meconium plug syndrome in neonates. J Surg Res 2012; 175:e43–e46.
41. Karimi A, Gorter RR, Sleeboom C, et al. Issues in the management of
simple and complex meconium ileus. Pediatr Surg Int 2011; 27:963–
968.
42. Wagner JP, Shekherdimian S. Hirschsprung disease. Emedicine
2015. http://emedicine.medscape.com/article/178493-overview.
Accessed November 29, 2015.
10
Academic Radiology, Vol ■, No ■, ■■ 2016
43. de Lorijn F, Kremer LC, Reitsma JB, et al. Diagnostic tests in Hirschsprung
disease: a systematic review. J Pediatr Gastroenterol Nutr 2006; 42:496–
505.
44. Noviello C, Cobellis G, Romano M, et al. Diagnosis of Hirschsprung’s
disease: an age-related approach in children below or above one year.
Colorectal Dis 2010; 12:1044–1048.
45. de Lorijn F, Reitsma JB, Voskuijl WP, et al. Diagnosis of Hirschsprung’s
disease: a prospective, comparative accuracy study of common tests.
J Pediatr 2005; 146:787–792.
46. Diamond IR, Casadiego G, Traubici J, et al. The contrast enema for
Hirschsprung disease: predictors of a false-positive result. J Pediatr Surg
2007; 42:792–795.
47. Levitt M, Pena A. Pediatric imperforate anus surgery workup. Emedicine
2013. http://emedicine.medscape.com/article/933524-overview.
Accessed November 29, 2015.
48. Narasimharao KL, Prasad GR, Katariya S, et al. Prone cross-table lateral
view: an alternative to the invertogram in imperforate anus. AJR Am J
Roentgenol 1983; 140:227–229.
49. Niedzielski JK. Invertography versus ultrasonography and distal
colostography for the determination of bowel-skin distance in children
with anorectal malformations. Eur J Pediatr Surg 2005; 15:262–267.
50. Wetherill C, Sutcliffe J. Hirschsprung disease and anorectal malforma-
tion. Early Hum Dev 2014; 90:927–932.
51. Podberesky DJ, Towbin AJ, Eltomey MA, et al. Magnetic resonance
imaging of anorectal malformations. Magn Reson Imaging Clin N Am 2013;
21:791–812.
52. Farhataziz N, Engels JE, Ramus RM, et al. Fetal MRI of urine and me-
conium by gestational age for the diagnosis of genitourinary and
gastrointestinal abnormalities. AJR Am J Roentgenol 2005; 184:1891–
1897.
53. Applegate KE. Evidence-based diagnosis of malrotation and volvulus.
Pediatr Radiol 2009; 39(suppl 2):S161–S163.