ORIGINAL ARTICLE

The value of the erect abdominal radiograph for the diagnosis of mechanical
bowel obstruction and paralytic ileus in adults presenting with acute
abdominal pain
Wendy Z. M. Geng, BMedRad (Hons), Michael Fuller, ADDR, BHA, Brooke Osborne, BMedRad,
MMedSon, GradDipEd, BHlthSc (Hons),& Kerry Thoirs, Assoc Dipl Rad Tech, DMU, MMedRad, PhD
International Centre for Allied Health Evidence, University of South Australia, Adelaide, South Australia, Australia

Keywords Abstract
Abdomen, acute, diagnostic x-ray, ileus, intestinal
obstruction, sensitivity and specificity Introduction: There is discord on the value of the erect abdominal radiograph for
diagnosing acute abdominal pathologies. The erect radiograph can be
Correspondence uncomfortable for patients in pain and increases patient radiation dose. Aim: To
Wendy Z. M. Geng, International Centre for determine if including the erect abdominal radiograph in plain abdominal
Allied Health Evidence, University of South
radiography (PAR) improved diagnostic accuracy for identifying mechanical
Australia, GPO Box 2471, Adelaide, South Australia
5001, Australia.
bowel obstruction and/or paralytic ileus in adults presenting with acute abdominal
Tel: +61 08 1300 301 703; pain. Methods: PAR of 40 consecutive adults presenting with suspected bowel
Fax: +61 08 8302 2466; obstruction or paralytic ileus was retrospectively sampled and independently
E-mail: [email protected] reviewed by two emergency department (ED) consultants and two radiology
consultants for bowel obstruction and paralytic ileus across two sessions. In
Received: 15 January 2018; Revised: 22 June session 1, the assessors assessed the supine abdominal radiographs (PAR 1) and
2018; Accepted: 22 June 2018 J clinical details in a randomised order, and session 2, at least 6 weeks later, they
assessed the supine and erect radiographs (PAR 2) and clinical details of the
Med Radiat Sci xx (2018) 1–8
randomly re-ordered cases. Computed tomography was the reference standard.
Pair-wise comparisons of receiver operating characteristic curves were calculated
doi: 10.1002/jmrs.299
to assess for significant differences in participants’ diagnostic accuracy using
MedCalc 16.4.3. Results: Average sensitivity, specificity and area under the
receiver operating characteristic curves (AUROC) were 69.7%, 61.0% and 0.642
for PAR 1, respectively, and 80.0%, 53.4% and 0.632 for PAR 2 respectively. For
AUROC there were no significant differences (P> 0.05) between PAR 1 and PAR
2. Intra-rater and inter-rater agreement improved in PAR 2. Conclusion: There
was no statistically significant improvement in diagnostic accuracy when
including the erect radiograph in PAR for the acute abdomen.

abdominal radiograph (EAR) may not be necessary

Introduction because most findings are demonstrated on the supine
Plain abdominal radiography (PAR) is often the initial abdominal radiograph (SAR). 2,3,6–8 However, air–fluid diagnostic
imaging tool for patients presenting with acute levels are only seen on the EAR and are a significant abdominal pain. 1 PAR
in the acute setting may consist of radiological sign for diagnosing acute small bowel supine and erect abdominal
radiographs and an erect obstruction. 9,10 Bowel obstruction is one of the most chest radiograph. 2,3 The erect chest view is
recommended common diagnoses in patients presenting with acute for diagnosing chest pathologies, such as pneumonia,
that abdominal pain, accounting for 12.6–21.8% of emergency mimic the symptoms of an acute abdomen. 4 The
admissions.11–13 This condition prevents the distal flow of exposure parameters of the chest x-ray also give greater intestinal
contents, and can be of a mechanical or visualisation of free gas under the diaphragm when a functional nature. Mechanical
bowel obstruction occurs hollow visceral perforation is suspected. 5 The erect when a physical barrier prevents intestinal
 The Erect Abdominal Radiograph W. Z. M. Geng et al.
flow, whereas
ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of 1
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology. This is an
open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction
in any medium, provided the original work is properly cited.
paralytic ileus is a functional impairment of the bowel radiology information system (RIS) of Flinders Medical
wall or nervous system.14 Immediate life-saving surgical Centre. Consecutive patient cases from 29 November
intervention is required in some cases of bowel 2016 were sampled in a backwards time direction until
obstruction.15 Therefore, a rapid diagnosis is required for 40 cases were retrieved. Case inclusion criteria were that
these conditions. It would be helpful for radiographers to the patient was 18–65 years old, presented to the
know the value of the EAR, as it can be a difficult emergency department (ED) with acute abdominal pain
radiograph for both the radiographer and the patient to and clinical suspicion of mechanical bowel obstruction
achieve when the patient is in pain or disabled and adds or paralytic ileus, have undergone SAR and EAR within
to patient radiation dose. There has also been limited 24 h of presentation to the ED and undertaken an
investigation into the diagnostic value of the EAR, with abdominal computed tomography (CT) scan within 4 h
most studies undertaken over two decades ago and after PAR. Cases were excluded if the patient was
without using a standardised reference standard.2,7,16–21 institutionalised, had psychiatric or neurological
The diagnostic value of the EAR may vary depending disorders or had an equivocal CT result. Our sample size
upon the experience of the interpreting doctor. Doctors of 40 cases was based on an estimation that 20 of the 40
with less experience may find it difficult to interpret cases would have mechanical bowel obstruction or
abdominal radiographs that appear to have normal paralytic ileus. According to a published table 24 this
anatomy but have an unusual bowel gas pattern. 22 The would provide an 80% probability of detecting a 20%
addition of an EAR or decubitus abdominal radiograph difference in diagnostic accuracy (P< 0.05).
may be helpful to these doctors as it provides more
information or, as others argue, it may provide Index tests
misleading information.17,23
Two PAR protocols were compared for each case. PAR 1
We undertook this study to (1) determine if the consisted of only SAR, and PAR 2 consisted of SAR and
inclusion of the EAR in PAR improves diagnostic EAR.
accuracy in identifying mechanical bowel obstruction Two radiology consultants and two ED consultants, all
and/or paralytic ileus in adults presenting with acute of whom were experienced in interpreting PAR in the acute
abdominal pain and (2) to determine if there is a setting, participated in this study. Post-registration
difference in the interpretations of abdominal experience ranged from 25 to 27 years, and 10 to 13 years
radiographs between doctors. for the radiologists and ED consultants respectively. Each
assessor independently assessed the two PAR protocols,
with a minimum 6-week interval between starting PAR 2
Methods and completing PAR 1. Clinical information from the
This study was undertaken at Flinders Medical Centre, request form for each case was included with the images.
South Australia. Ethical approval was granted by the Cases were presented to the assessors in a randomised
Southern Adelaide Clinical Human Research Ethics order using an electronic survey tool (SurveyMonkey Inc.,
Committee and the Human Research Ethics Committee, California, USA) (Fig. 1). The assessors could only view
University of South Australia. the series in a forward direction to limit comparisons with
previous cases. The assessors were instructed on how to
Study design indicate their diagnostic assessment using the visual
analogue scale (VAS) in the survey. This consisted of a 0–
The diagnostic accuracy of two different plain abdominal 100 continuous scale which provided an indication of how
radiography (PAR) protocols was compared by sure each participant was of their assessment. The left and
retrospectively reviewing patient cases who had right ends of the scale were labelled with ‘definitely no
presented with acute abdominal pain and were clinically obstruction/paralytic ileus’ and ‘definite
suspected of mechanical bowel obstruction or paralytic obstruction/paralytic ileus’ respectively. The assessors
ileus. could undertake the assessments at their convenience and
complete the survey across multiple sittings.
Case selection
Cases were retrospectively selected from the picture Reference standard
archive and communication system (PACS) and

2 ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
W. Z. M. Geng et al. The Erect Abdominal Radiograph
The radiology report for the CT scan from each case was between PAR protocols and assessors using the statistical
used to categorise each case as ‘positive’ or ‘negative’ and method of Delong, Delong and ClarkePearson.30 An
served as the reference standard. The assessors were AUROC of ‘0’ indicates that the diagnostic test is
blinded to the CT results. Each CT scan was reported as consistently incorrect at differentiating diseased from non-
per the department protocol, with an available consultant diseased states, ‘1’ indicates the test to be always correct

(a) (b)

Figure 1. (a) An example of a patient case presented on the online survey. (b) A patient’s supine abdominal x-ray presented on the survey.
radiologist or radiology registrar producing the report. All and ‘0.5’ indicates a chance level of differentiation. 31 The
reports produced by registrars were checked by consultants Youden’s index was calculated to determine the optimal
with an addendum report issued if necessary. CT has a threshold point (criterion value), and its associated
reported sensitivity, specificity and accuracy of 90–94%, sensitivity and specificity.32,33
93–100% and 94–95%, respectively, for the detection of
mechanical bowel obstruction,25–28 and has the highest
accuracy for the differential diagnosis of mechanical SBO Agreement testing
and post-operative paralytic ileus.29 Ten duplicate cases for each PAR protocol were used to
test for intra-rater agreement of each assessor’s
diagnostic interpretations. Each assessor was given
Statistical analysis
different duplicate cases randomly mixed into the case
All statistical calculations were performed using MedCalc series. Intra-class correlation coefficients (ICC) were
16.4.3 (MedCalc Software, Ostend, Belgium). Two calculated using a twoway mixed-effects model based on
receiver operating characteristic (ROC) curves were a single measure and absolute agreement. Inter-rater
generated for each assessor using continuous data from the agreement was tested by comparing diagnostic
VAS scale (index test) and binomial data from the CT assessments between the assessors for each PAR.
radiology report (reference test). The first ROC for each Agreement was tested by calculating the ICC using a
participant represented their diagnostic assessments when two-way mixed-effects model based on the average of
using only the supine abdominal radiograph (PAR 1). The two raters and absolute agreement. 34 An ICC of less than
second ROC represented their diagnostic assessments 0.5, between 0.5 and 0.75, between 0.75 and 0.9 and
when using both the supine and erect abdominal more than 0.9 demonstrates poor, moderate, good and
radiograph (PAR 2). Area under the receiver operating excellent agreement respectively.34
characteristic curves (AUROC) were calculated for each
participant and pair-wise comparisons (P< 0.05) were made

ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of 3
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
 The Erect Abdominal Radiograph W. Z. M. Geng et al.
Results Moderate-to-good agreement (ICC of 0.413–0.733)
between the assessors was achieved for PAR 1, and
The 40 cases included 17 females and 23 males (mean age
goodto-excellent agreement was achieved for PAR 2
49.0 9.42 years). Table 1 demonstrates clinical
(Table 4).
presentations of all cases. Fifteen (38%) cases had bowel
obstruction or paralytic ileus diagnosed by CT. The Discussion
average time between patient admission and PAR was Both PAR protocols demonstrated low-to-moderate
194155 min and between PAR and CT was diagnostic accuracy for identifying mechanical bowel
obstruction and/or paralytic ileus in adults presenting with
13760.5 min. The time interval between the two testing
acute abdominal pain. We found no significant differences
sessions was 7–8 weeks for the ED doctors, 6 weeks for in the overall accuracy between the two protocols. This is
one radiologist and 10 weeks for the second radiologist. consistent with other studies which have demonstrated
limited value of the EAR.16,19,21
Diagnostic accuracy of consultants’ interpretations We found no significant differences in overall diagnostic
accuracy between the assessors. This is in contrast to the
Diagnostic accuracy data for each assessor for PAR 1
study by Thompson et al.,10 which found more senior and
and PAR 2 are presented in Table 2. Across all assessors,
experienced radiologists to be more accurate and confident
the AUROC ranged from 0.581 to 0.712 with an average
in diagnosing SBO using PAR than radiologists with less
of 0.632 for PAR 1 and from 0.565 to 0.673 with an
than 5 years experience. The assessors in our study all had
average of 0.632 for PAR 2. There were no significant
over 10 years of experience suggesting that the effect of
differences (P> 0.05) in AUROC between the two PAR
experience on learning diminishes after 10 years for both
protocols. Average sensitivity and specificity were
radiologists and ED doctors. Other authors have compared
69.7% and 61.0% for PAR 1, respectively, and 80.0%
the interpretations made by radiologists and non-
and 53.4% for PAR 2 respectively (Table 3). There was
radiologists, finding that non-radiology doctors mostly
a wide variation in optimum criterion values, sensitivity
missed, misinterpreted or identified irrelevant radiological
and specificity values between assessors and between
features.7,17 Improved image interpretation training for non-
PAR protocols.
radiologic doctors since the 1980s is a potential reason for
the discrepancy in our findings and these earlier studies.
Intra-rater agreement Intra-rater and inter-rater agreement increased when the
EAR radiograph was added to the protocol. This
Moderate-to-excellent intra-rater agreement (ICC of
improvement was most profound for inter-rater agreement
0.551–0.939) was achieved for PAR 1 (Table 4). Adding
between the two ED consultants, which more than doubled
EAR to PAR 2 increased the intra-rater agreement of
when the EAR radiograph was added. Factors for this
diagnostic interpretations for all assessors except one
result may include both the doctors’ speciality or years of
radiology consultant.
experience which was different from the radiologists.
However, the wide confidence interval for some results
Inter-rater agreement indicates that the 10 duplicates cases used to test reliability
may not have been enough to give a true indication of
reliability.
Table 1. Clinical symptoms and computed tomography diagnosis.
Computed tomography diagnosis
(n)
Clinical symptoms (n)
Abdominal pain (31) Bowel obstruction (13)
? bowel Paralytic ileus (2)
obstruction/ileus (38)
Abdominal distension (9) Appendicitis (6)
Decrease/no flatus or bowel Inflammation of the bowel (2)
not open (13)
Nausea/vomiting (15) Perforation (2)
? perforation (16) Hernia (3)
Known hernia (4) Other abnormalities (8)

4 ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
W. Z. M. Geng et al. The Erect Abdominal Radiograph
? hernia (2) No intra-abdominal abnormality (4)
Other clinical details indicative
of bowel obstruction/ileus
(13)

We asked the assessors to rate, on a continuous scale, the definite presence or absence of the conditions rather than to
dichotomise their assessment into ‘positive’ or ‘negative’. This reflects radiologic practice, where descriptors such as
‘probable’, ‘unlikely’ or ‘apparent’ are commonly used. 35 There were wide variations in the sensitivity and specificity
values between the assessors. This variation was also demonstrated in previous studies reporting wide ranges of
sensitivity (19–96.2%) and specificity (57–100%) for diagnosing SBO.10,36–40 Our sensitivity and specificity values for
SAR (40–86.7%) and SAR combined with EAR (73.3–86.7%) were lower than that reported by Tie and Edwin, 21 who
reported 88.5% and 92.5% sensitivity for SAR and SAR combined with EAR respectively. These differences may be
accounted for by disease prevalence, which was higher in our study (37.5%), compared to 11.6% reported by Tie and
Edwin.21,41
Table 2. Diagnostic accuracy of assessments made by each doctor for each protocol.

AUROC (95% CI)

Significant level
Assessor PAR 1 PAR 2 Difference in AUROC (95% CI) (P-value)

1 (ED)1 0.642 (0.472 to 0.788) 0.565 (0.397 to 0.723) 0.0764 (0.0904 to 0.243) 0.370

2 (ED) 0.581 (0.415 to 0.735) 0.673 (0.507 to 0.813) 0.0920 (0.0637 to 0.248) 0.247

3 (radiology) 0.712 (0.547 to 0.844) 0.651 (0.484 to 0.794) 0.0613 (0.112 to 0.235) 0.489

4 (radiology)2 0.634 (0.465 to 0.782) 0.637 (0.468 to 0.785) 0.00286 (0.212 to 0.218) 0.979

ED, emergency department; >, greater than; %, percentage; CI, confidence interval; AUROC, area under the receiver operating characteristic curve. 1
Case 1 in PAR 2 removed from statistical analysis due to data management error. 2 Case 13 in PAR 1 removed from statistical analysis due to data
management error.

Table 3. Sensitivity and specificity at criterion values.

1 Specificity (%) (95%
Criterion value Sensitivity (%) (95% CI)
CI)
PAR2
Assessor PAR1 PAR1 PAR2 PAR1 PAR2
2
1 (ED) >52 >26 73.3 (44.9–92.2) 86.7 (59.5–98.3) 56.0 (34.9–75.6) 37.5 (18.8–
59.4)
2 (ED) >38 >51 86.7 (59.5–98.3) 73.3 (44.9–92.2) 36.0 (18.0–57.5) 56.0 (34.9–
75.6)
3 (radiology) >86 >81 40.0 (16.3–67.7) 73.3 (44.9–92.2) 96.0 (79.6–99.9) 68.0 (46.5–
85.1)
4 (radiology)3 >29 >8 78.6 (49.2–95.3) 86.7 (59.5–98.3) 56.0 (34.9–75.6) 52.0 (31.3–
72.2)
ED, emergency department; >, greater than; %, percentage; CI, confidence interval.
1
– 1 is maximum.
Criterion value is the value on the receiver operating characteristic curve where sensitivity and specificity 2 Case 1
in PAR 2 removed from statistical analysis due to data management error. 3 Case 13 in PAR 1 removed from
statistical analysis due to data management error.

Table 4. Participants’ agreement in sessions 1 and 2.

ICC for PAR 2 (95%
Assessor(s) ICC for PAR 1 (95% CI)
CI)
1 (ED) 0.676 (0.144 to 0.907) 0.993 (0.976 to 0.998)1
2 (ED) 3 (radiology) 0.551 (0.0964 to 0.867) 0.794 (0.363 to
0.945)
0.774 (0.307 to 0.939)
0.976 (0.906 to
0.994)
4 (radiology) 0.939 (0.787 to 0.984) 0.667 (0.142 to
0.903)

ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of 5
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
 The Erect Abdominal Radiograph W. Z. M. Geng et al.
Radiology consultants (combined) 0.630 (0.288 to 0.807)2 0.617 (0.0844 to 0.823)
ED consultants (combined) 0.413 (0.116 to 0.690) 0.859 (0.579 to 0.940)2
All consultants (combined) 0.8650 (0.7534 to 0.9275)2
0.733 (0.558 to 0.846)2
ICC, intra-class correlation coefficient; %, percentage; CI, confidence interval; ED, emergency department.
1
11 duplicate cases analysed for intra-rater agreement. raising the possibilitythat the sample was not large
2
39 duplicate cases analysed for intra-rater agreement. enough to detect true significant differences.
Alternative radiographs to the EAR, such as decubitus
abdominal or erect chest radiographs, were not
Our results do not strongly support the inclusion of EAR considered in this study. This study used CT as the sole
in PAR, with the likelihood of additional confirmatory and standardised reference standard due to its high
imaging such as CT still being required. This study builds accuracy, however, it is not 100% accurate for
on the existing limited body of evidence investigating the diagnosing bowel obstruction and paralytic ileus.27,28
value of the EAR when bowel obstruction or paralytic ileus We minimised the risk of memory recall bias and
is suspected. We used CT as a consistent and sole reference cross-referencing between assessors by randomising the
standard. Compared to other studies where clinical history order of case presentation for each assessor, and a time
was not revealed to interpreting doctors, the assessors in interval of at least 6 weeks between interpretation of
our study reviewed clinical details together with the each protocol.
radiograph, reflecting normal practice. 7,16,17,21 We did not The survey tool restricted the use of ‘windowing’ of
seek to identify radiographic signs of bowel obstruction images, making measurements which may have been
and paralytic ileus. used to facilitate the determination of the degree of
PAR is still used in many practices as the initial bowel loop distension, and a standardised film reading
imaging modality for patients experiencing acute environment. In normal practice, radiographs are viewed
abdominal pain due to its low cost and wide on highdefinition computer screens. However, the
availability.10,21,42 Based on a wide range of diagnostic assessors’ interpretations were unlikely to have been
values in previous studies, and the low diagnostic affected as radiologic signs for bowel obstruction and
accuracy and variations in sensitivity and specificity paralytic ileus do not need high resolution.46 Another
across assessors reported in this study, patients with a potential limitation is that the departmental CT reporting
negative or positive PAR are still likely to undergo process did not control for intra-reader and inter-reader
another confirmatory test such as CT. Thus, the use of variability between different radiology consultants.
PAR for patients presenting with an acute abdomen
should be reviewed. Rather than investing in more
rigorous prospective studies with larger sample sizes in Conclusion
PAR, perhaps consideration should be given to studying Both PAR protocols demonstrated low diagnostic
the feasibility of other diagnostic tools such as low-dose accuracy for the identification of mechanical bowel
CT (LDCT) in place of PAR. LDCT has been shown to obstruction and paralytic ileus in adults presenting with
give significantly higher diagnostic yield than PAR for acute abdominal pain raising questions about the value
adults with acute abdominal pain and can potentially of PAR in this setting. The addition of the EAR to the
reduce the number of further imaging investigations with SAR gave a slight but insignificant increase in diagnostic
almost equal or only slightly higher radiation dose.43–45 accuracy, and improved the intra-rater and inter-rater
agreement, particularly for ED consultants.
Radiographers performing PAR in the investigation of
Limitations mechanical bowel obstruction and paralytic ileus should
The retrospective study design and sampling methods are be aware of the limited value of the erect radiograph,
limitations. The criteria of including patients who had especially in situations where it is technically difficult to
both achieve, patient tolerance is low and the radiographs are
PAR and CT may have created bias to the sample to to be viewed by an experienced consultant radiologist.
include more cases referred for CT investigation due to
equivocal abdominal radiographs, and more cases with
unequivocal CT findings. Patients who were
Conflict of Interest
institutionalised, or who had psychiatric or neurological The authors declare no conflict of interest.
disorders were also excluded and therefore results cannot
be applied to these patient groups. Confidence intervals References
for many of our outcome measurements were wide,

6 ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
W. Z. M. Geng et al. The Erect Abdominal Radiograph
1. Gans SL, Stoker J, Boermeester MA. Plain routine use be abandoned. Br Med J (Clin Res Ed) 1985; 290:
abdominalradiography in acute abdominal pain; past, present, 1934–6.
and future. Int J Gen Med 2012; 5: 525–33. 18. Tan TX, Mueller CF, Vaswani KK, Augustyn
2. Mirvis SE, Young J, Keramati B, McCrea ES, Tarr R. MA,Holtzlander JD, Spigos DG. The erect abdominal
Plainfilm evaluation of patients with abdominal pain: Are radiograph: Is it necessary for emergency department patients
three radiographs necessary? Am J Roentgenol 1986; 147: with acute abdomen? Radiology 1998; 209P: 172.
501–3. 19. Ukrisana P, Yenarkarn P. Evaluation of the necessity ofthe
3. Stoker J, van Randen A, Lameris W, Boermeester MA. three-film abdominal series in the diagnosis of abdominal
Imaging patients with acute abdominal pain. Radiology 2009; pain. J Med Assoc Thai 2002; 85: 998–
253: 31–46. 1002.
4. Mayumi T, Yoshida M, Tazuma S, et al. The 20. Hoffstetter P, Schleder S, Jung EM, et al. Plain
practiceguidelines for primary care of acute abdomen 2015. J abdominalradiograph is there any additional clinical value
Gen Fam Med 2016; 17: 5–52. based on the supine projection? Dtsch Med Wochenschr
5. Alazzawi S, De Rover WS, Morris-Stiff G, Lewis M. 2011; 136: 2589–93.
Erectchest radiography in the setting of the acute abdomen: 21. Tie H, Edwin P. The role of plain erect abdominalradiography
Essential tool or an unnecessary waste of resources? Ann R in the diagnosis of small bowel obstruction. J Med Imaging
Coll Surg Engl 2010; 92: 697–9. Radiat Oncol 2016; 60(S1): 187.
6. De Lacey GJ, Wignall BK, Bradbrooke S, Reidy J, Hussain 22. Lim CBB, Chen V, Barsam A, Berger J, Harrison RA.
S,Cramer B. Rationalising abdominal radiography in the Plainabdominal radiographs: Can we interpret them? Ann R
accident and emergency department. Clin Radiol 1980; 31: Coll Surg Engl 2006; 88: 23–6.
453–5. 23. Whitley AS, Sloane C, Hoadley G, Moore AD.
7. Hayward M, Hayward C, Ennis W, Roberts C. A Clark’sPositioning in Radiography, 12th edn. CRC Press,
pilotevaluation of radiography of the acute abdomen. Clin London, 2005.
Radiol 1984; 35: 289–91. 24. Hanley JA, McNeil BJ. The meaning and use of the areaunder
8. Prasannan S, Zhueng TJ, Gul YA. Diagnostic value of a receiver operating characteristic (ROC) curve. Radiology
plainabdominal radiographs in patients with acute abdominal 1982; 143: 29–36.
pain. Asian J Surg 2005; 28: 246–51. 25. Megibow A, Balthazar E, Cho K, Medwid S, Birnbaum
9. Lappas JC, Reyes BL, Maglinte DD. Abdominalradiography B,Noz M. Bowel obstruction: Evaluation with CT. Radiology
findings in small-bowel obstruction: Relevance to triage for 1991; 180: 313–8.
additional diagnostic imaging. Am J Roentgenol 2001; 176: 26. Fukuya T, Hawes D, Lu CC, Chang PJ, Barloon TJ.
167–74. CTdiagnosis of small-bowel obstruction: Efficacy in 60
10. Thompson WM, Kilani RK, Smith BB, et al. Accuracy patients. Am J Roentgenol 1992; 158: 765–9.
ofabdominal radiography in acute small-bowel obstruction: 27. Suri S, Gupta S, Sudhakar P, Venkataramu N, Sood B,Wig J.
Does reviewer experience matter? Am J Roentgenol 2007; Comparative evaluation of plain films, ultrasound
188: W233–8. and CT in the diagnosis of intestinal obstruction. Acta Radiol
11. Irvin TT. Abdominal pain: Surgical audit of 1190 emergency 1999; 40: 422–8.
admissions. Br J Surg 1989; 76: 1121–5. 28. Shakil O, Zafar SN, Saleem S, Khan R, Pal KMI. The roleof
12. Ohene-Yeboah M. Acute surgical admissions forabdominal computed tomography for identifying mechanical bowel
pain in adults in Kumasi, Ghana. ANZ J Surg 2006; 76: 898– obstruction in a Pakistani population. J Pak Med Assoc 2011;
903. 61: 871–4.
13. Soressa U, Mamo A, Hiko D, Fentahun N. Prevelance,causes 29. Wu Z, Boersema GS, Dereci A, Menon AG, Jeekel J, Lange
and management outcome of intestinal obstruction in Adama JF. Clinical endpoint, early detection, and differential
Hospital, Ethiopia. BMC Surg 2016; 16: 38–46. diagnosis of postoperative ileus: A systematic review of the
14. Kulaylat MN, Doerr RJ. Small bowel obstruction. In: literature. Eur Surg Res 2015; 54: 127–38.
Holzheimer RG, Mannick JA (eds). Surgical Treatment: 30. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparingthe
Evidence-Based and Problem-Orientated. Zuckschwerdt, areas under two or more correlated receiver operating
Munich, 2001; pp. 102–14. characteristic curves: A nonparametric approach. Biometrics
15. Dıte P, Lata J, Novotny I. Intestinal obstruction and 1988; 44: 837–45.
perforation–the role of the gastroenterologist. Dig Dis 2003; 31. Hajian-Tilaki K. Receiver operating characteristic (ROC)
21: 63–7. curve analysis for medical diagnostic test evaluation. Caspian
16. Simpson A, Sandeman D, Nixon SJ. The value of an J Intern Med 2013; 4: 627–35.
erectabdominal radiograph in the diagnosis of intestinal 32. Ruopp MD, Perkins NJ, Whitcomb BW, Schisterman
obstruction. Clin Radiol 1985; 36: 41–2. EF.Youden index and optimal cut-point estimated from
17. Field S, Guy PJ, Upsdell SM, Scourfield AE. The observations affected by a lower limit of detection. Biom J
erectabdominal radiograph in the acute abdomen: Should its 2008; 50: 419–30.

ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of 7
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology
 The Erect Abdominal Radiograph W. Z. M. Geng et al.
33. Youden WJ. Index for rating diagnostic tests. Cancer 1950; 3:
32–5.
34. Koo TK, Li MY. A guideline of selecting and
reportingintraclass correlation coefficients for reliability
research. J Chiropr Med 2016; 15: 155–63.
35. Hall FM. Language of the radiology report: Primer
forresidents and wayward radiologists. Am J Roentgenol
2000; 175: 1239–42.
36. Frager D, Medwid S, Baer JW, Mollinelli B, Friedman M.CT
of small-bowel obstruction: Value in establishing the
diagnosis and determining the degree and cause. Am J
Roentgenol 1994; 162: 37–41.
37. Frager DH, Baer JW, Rothpearl A, Bossart PA.
Distinctionbetween postoperative ileus and mechanical small-
bowel obstruction: Value of CT compared with clinical and
other radiographic findings. Am J Roentgenol 1995; 164:
891–4.
38. Kim S, Park K, Kim S, et al. Accuracy of plain
abdominalradiography in the differentiation between small
bowel obstruction and small bowel ileus in acute abdomen
presenting to emergency department. Hong Kong J Emerg
Med 2011; 18: 68–79.
39. Maglinte D, Reyes B, Harmon B, et al. Reliability and roleof
plain film radiography and CT in the diagnosis of smallbowel
obstruction’. Am J Roentgenol 1996; 167: 1451–5.
40. Shrake PD, Rex DK, Lappas JC, Maglinte DD.Radiographic
evaluation of suspected small bowel obstruction’. Am J
Gastroenterol 1991; 86: 175–8.
41. Leeflang MM, Bossuyt PM, Irwig L. Diagnostic testaccuracy
may vary with prevalence: Implications for evidence-based
diagnosis. J Clin Epidemiol 2009; 62: 5–12.
42. Paulson EK, Thompson WM. Review of small-
bowelobstruction: The diagnosis and when to worry.
Radiology 2015; 275: 332–42.
43. Alshamari M, Norrman E, Geijer M, Jansson K, Geijer
H.Diagnostic accuracy of low-dose CT compared with
abdominal radiography in non-traumatic acute abdominal
pain: Prospective study and systematic review. Eur Radiol
2016; 26: 1766–74.
44. Haller O, Karlsson L, Nyman R. Can low-dose abdominalCT
replace abdominal plain film in evaluation of acute abdominal
pain? Ups J Med Sci 2010; 115: 113–20.
45. Nguyen LK, Wong DD, Fatovich DM, et al. Low-
dosecomputed tomography versus plain abdominal
radiography in the investigation of an acute abdomen. ANZ J
Surg 2012; 82: 36–41.
46. Baker SR, Cho KC. The Abdominal Plain Film
withCorrelative Imaging, 2nd edn. Appleton & Lange,
Stamford, 1999.

8 ª 2018 The Authors. Journal of Medical Radiation Sciences published by John Wiley & Sons Australia, Ltd on behalf of
Australian Society of Medical Imaging and Radiation Therapy and New Zealand Institute of Medical Radiation Technology