Você está na página 1de 12

Hong Kong Journal of Emergency Medicine

Accuracy of plain abdominal radiography in the differentiation between small bowel obstruction and small bowel ileus in acute abdomen presenting to emergency department

X
SH Kim, KN Park, SJ Kim, CK Eun, YM Park, MK Oh, KH Choi, HJ Kim, DW Kim, HJ Choo, JH Cho, JH Oh, HY Park

Introduction: Our purpose was to evaluate whether plain abdominal radiography (PAR) could accurately differentiate between small bowel obstruction (SBO) and small bowel ileus (SBI) in an emergency setting. We also evaluated the value of known classic signs on the PAR for differentiating between SBO and SBI. Methods: This retrospective study included 216 emergency room patients who had small bowel distension (maximal small bowel diameter 2.5 cm) on the PAR and who underwent successive abdominal computed tomography. One radiologist and one emergency physician retrospectively reviewed PAR in consensus, unaware of the patients' clinical data; they divided the patients into an SBO group and an SBI group according to the radiographic findings. Presence or numeric values of 10 radiographic signs were also recorded. Final diagnoses of SBO and SBI were established by a combined analysis of medical charts, surgical records, radiographic findings on abdominal computed tomography, and small bowel studies. The differential diagnoses based on PAR and the final diagnoses were compared, and the sensitivity and specificity of PAR were calculated. We also evaluated the differences among 10 radiographic signs between the final SBO and SBI groups. Results: Sensitivity and specificity of PAR for SBO were 82.0% and 92.4%, respectively. Among the 10 radiographic signs, all except maximal colon diameter were statistically significant predictors on the final

Correspondence to: Park Kyu Nam, MD Seoul St. Mary's Hospital, Department of Emergency Medicine, The Catholic University of Korea, 505 Banpo-dong, Seocho-gu, Seoul 137-701, Republic of Korea Email: emsky@catholic.ac.kr Kim Han Joon, MD Inje University Haeundae Paik Hospital, Department of Emergency Medicine, 1435 Jwa-dong, Haewondae-gu, Busan 612-030, Republic of Korea Kim Suk Hwan , MD Cho Joon Ho, MD Oh Je Hyuk, MD Park Ha Young, MD Inje University Haeundae Paik Hospital, Department of Radiology, 1435 Jwa-dong, Haewondae-gu, Busan 612-030, Republic of Korea Kim Suk Jung, MD Eun Choong Ki, MD

Inje University Busan Paik Hospital, Department of Radiology, Radiology, 633-165 Gaegeum 2(i)-dong, Busanjin-gu, Busan, 614-735, Republic of Korea Park Young Mi, MD Kim Dong Wook, MD Choo Hae Jung, MD Inje University Busan Paik Hospital, Clinical Trial Center, 633-165 Gaegeum 2(i)-dong, Busanjin-gu, Busan, 614-735, Republic of Korea Oh Min Kyung, PhD Uijeongbu St. Mary's Hospital, Department of Emergency Medicine, The Catholic University of Korea, 65-1 Geumo-dong, Uijeongbu-si, Gyeonggi-do, 480-717, Republic of Korea Choi Kyung Ho, MD

Kim et al./Differentiating small bowel obstruction from ileus

69

diagnosis. Conclusions: PAR is an accurate and effective initial imaging modality for differentiating between SBO and SBI in an emergency setting, and most of the classic radiographic signs have a diagnostic value. (Hong Kong j.emerg.med. 2011;18:68-79)
X PAR PAR 2.5 PAR 10 PAR 82.0% X 92.4% X PAR 10 X 10 X PAR PAR SBO SBI SBO SBO SBI SBO 216 PAR

SBI

Keywords: Dilatation, intestinal pseudo-obstruction, X-ray film


X

Introduction
Small bowel distension (maximal small bowel diameter 2.5 cm) on plain abdominal radiography (PAR) indicates the presence of a small bowel disorder with impaired transit of intestinal contents.1-7 This disorder is commonly subdivided into small bowel obstruction (SBO), which result from intrinsic luminal obstruction or extrinsic compression, and small bowel ileus (SBI), which result from intestinal autonomic nervous system dysfunction, eventually delaying the transit of intestinal contents.1,8 Abdominal computed tomography has recently been regarded as a more accurate imaging tool than PAR in differentiating between SBO and SBI,8-10 but PAR is still advocated by some authors as the initial modality of choice for such differentiation. 4,6,11-13 However, it would be inappropriate to apply the results of prior studies directly to an emergency setting, because those study populations were limited to patients who were clinically suspected of having SBO or who were in a post-operative period. Recent research has raised questions concerning the effectiveness of known classic signs (such as air-fluid

levels and differential air-fluid levels) on PAR for differential diagnosis between SBO and SBI.2,14-16 And, to our knowledge, clinical differences between SBO and SBI have not been previously investigated. Small bowel distension on PAR is frequently found among patients with abdominal pain visiting the emergency room. For these patients, differentiation between SBO and SBI is crucial to decide treatment.17,18 Although classic plain radiographic signs for differential diagnosis between SBO and SBI are commonly used by physicians, the effectiveness of these signs on differential diagnosis has not been vigorously studied. 2,15,16,19 Awareness of their real effectiveness might facilitate correct diagnosis. In addition, awareness of the clinical differences between SBO and SBI might help physicians to establish a diagnosis and predict the patient's prognosis.17 The purpose of our study was to evaluate the accuracy of PAR in differentiating between SBO and SBI in emergency room patients complaining of abdominal pain. We modulated the classic signs into simplified forms that can be easily used in an emergency setting, and we evaluated their effectiveness. We also evaluated the clinical differences between the two groups.

70

Hong Kong j. emerg. med.

Vol. 18(2)

Mar 2011

Methods
We conducted a retrospective review of radiologic findings and medical charts for those patients who had abdominal pain and small bowel distension on PAR. This study was conducted at the emergency room of a tertiary referral university hospital in South Korea. The study period was from January 2008 to June 2008. This study was approved by our institutional review board, and informed consent was waived. We confined the study group to patients who had undergone both erect and supine PAR views, taking into consideration the possible influence of position on the radiographic findings. We further confined the study group to patients who had undergone successive abdominal computed tomography in order to enhance the accuracy of the final diagnosis and to facilitate confirmation of the causative disease.5,6,8-13,20-22 We selected 10 radiographic signs that can be easily applied in the emergency setting from among the known classical signs for differential diagnosis between SBO and SBI.1-3,7,15,16,19,23 These signs included air-fluid level, differential air-fluid level, string-of-beads sign,

stretch sign, decreased colon gas, number of air fluid levels, number of differential air-fluid levels, maximal width of air-fluid level, maximal small bowel diameter, and maximal colon diameter (Table 1). One radiologist and one emergency physician with 7 years of experience reviewed the radiographs together without any knowledge of the patients' clinical data; they evaluated the presence or numeric value of the 10 radiographic signs in consensus. Air-fluid levels, differential air-fluid levels, string-of-beads sign, stretch sign, and decreased colon gas were documented as being present or absent (existence or non-existence). The number of air fluid levels, differential air-fluid levels, maximal width of air-fluid level, maximal small bowel diameter, and maximal colon diameter were documented in the form of numeric values. Only airfluid levels that measured more than 10 mm in diameter were counted in the number of air-fluid levels, small air-fluid levels that measured less than 10 mm in diameter were classified as string-of-beads sign (Figures 1 and 2). Some ambiguity of standards was inevitable because no definite diagnostic cut-off values currently exist.

Table 1. The definitions of the 10 radiological signs for differential diagnosis between small bowel obstruction and small bowel ileus Radiological signs Air-fluid levels Differential air-fluid levels Bi-categoric signs (existence/ non-existence) String-of-beads sign Stretch sign Definitions A sharp flat horizontal line representing the interface between gas density above and fluid density below. Two different air-fluid levels in the same small bowel segment. Air-fluid levels measured as less than 10 mm in diameter. They typically arrange consecutively like string-of-beads. Abnormal distension of predominantly fluid-filled small bowel loop where the luminal gas has a striped appearance oriented perpendicular to long axis of the bowel. Diffuse decrease or absence of gas in the colon loop. The number of air-fluid levels that measured more than 10 mm in diameter. The number of differential air-fluid levels. The width of largest air-fluid level. The diameter of maximally distended small bowl loop. The diameter of maximally distended colon loop.

Decreased colon gas No. of air-fluid levels No. of differential air-fluid levels Numeric signs Maximal air-fluid level width (mm) Maximal small bowel diameter (mm) Maximal colon diameter (mm)

Kim et al./Differentiating small bowel obstruction from ileus

71

Figure 1. Erect view of an abdominal radiograph of a 55year-old man with small bowel obstruction due to postoperative adhesion. There are five air-fluid levels (short arrows) and two differential air-fluid levels (long arrows). Differential air-fluid level is depicted by two different airfluid levels in the same small bowel segment. Maximal airfluid level width is measured as 75 mm (line A). Maximal small bowel diameter is measured as 53 mm (line B). Maximal colon diameter is measured as 36 mm (line C).

Figure 2. Erect and supine view abdominal radiographs of a 58-year-old man with small bowel obstruction due to bezoar. Stringof-beads sign (arrow in erect view) is depicted by air-fluid levels measured as less than 10 mm in diameter. Stretch sign is depicted by abnormal distension of predominantly fluid-filled small bowel loop where the luminal gas has a striped appearance oriented perpendicular to the long axis of the bowel (arrows in supine view).

72

Hong Kong j. emerg. med.

Vol. 18(2)

Mar 2011

The final diagnoses of SBO and SBI were established by a combined analysis of medical charts, surgical records, and radiographic findings on abdominal computed tomography studies and small bowel series. We placed surgical records and radiographic findings on abdominal computed tomography first. 7,8,24-27 In addition, we recorded any history of previous abdominal surgery, performance of emergency abdominal surgery, occurrence of death, and causative disease noted on medical records. The differential diagnoses based on PAR were compared with the final diagnoses, and the sensitivity and specificity of PAR were calculated. The relationships between the radiographic signs and the final diagnoses were studied by univariate analysis and logistic regression analysis. A receiver operating characteristic (ROC) curve was drawn to calculate the cut-off values of numeric signs that had shown statistically significant results. Finally, a stepwise binary logistic regression analysis was performed to establish the significant predictors for diagnosis. For this process, numeric form signs were converted to categorical form by using the cut-off values. We compared the existence or non-existence of previous abdominal surgery history, performance or non-performance of emergency abdominal surgery, and death or life between the two groups by a Pearson's chi-square test or a Fisher's exact test. Compositional differences of causative diseases between the two groups were simply compared. Statistical analyses were performed using SPSS 11.5 (SPSS Inc, Chicago, Illinois, USA) and SAS 9.1.3 (SAS Institute, Car y, NC, USA). We assigned the significance level as p<0.05.

small bowel distension on PAR were included in the final study population. Of the 216 patients enrolled in this study, 127 patients were male and 89 patients were women. Their mean age was 58 years (range=16-88 years, standard deviation=16.4 years). On PAR, 47 patients were assessed as having SBO and 169 patients were assessed as having SBI. Final diagnoses were established by analysing clinical courses and abdominal computed tomography findings in 142 patients; clinical courses, abdominal computed tomography, and surgical findings in 73 patients; and clinical course, abdominal computed tomography, and small bowel study findings in one patient. Fifty patients were finally diagnosed as having SBO and 166 patients as having SBI (Figure 3). Among the 50 patients whose final diagnosis was SBO, 41 were assessed as SBO on PAR. Among the 166 patients whose final diagnosis was SBI, 160 were assessed as SBI on PAR. Consequently, the sensitivity and specificity of PAR were 82.0% and 96.4%, respectively. All the signs, except maximal colon diameter, showed significant statistical association with the final diagnoses (p<0.001) (Table 2). The AUCs of ROC curves were 0.884 for number of air-fluid levels, 0.805 for number of differential air-fluid levels, 0.881 for maximal air-fluid level width, and 0.871 for maximal small bowel diameter (Figure 4). As determined by ROC curve analysis of the statistically significant numeric form signs, cut-off values for discriminating SBO were as follows: more than 2 levels in number of air-fluid levels, more than 1 level in number of differential air-fluid levels, more than 31.5 mm in maximal air-fluid level width, and more than 37.5 mm in maximal small bowel diameter (Table 3). In the stepwise binary logistic regression analysis, presence of differential air-fluid levels and presence of stretch sign, maximal air-fluid level width more than 31.5 mm, maximal small bowel diameter more than 37.5 mm were significant independent predicting factors for final diagnosis (Table 4).

Results
From January 2008 to June 2008, 1355 adult patients with abdominal pain visited the hospital's emergency room. Seven hundred and forty one patients had undergone PAR (both erect and supine views) and successive abdominal computed tomography within 48 hours. Of these 741 patients, 216 patients who had

Kim et al./Differentiating small bowel obstruction from ileus

73

PAR=plain abdominal radiography; SBO=small bowel obstruction; SBI=small bowel ileus.

Figure 3. Flow-chart showing summary of the study result.

Figure 4. Receiver operation characteristic curves of numeric signs with statistical differences between small bowel obstruction and small bowel ileus.

74

Hong Kong j. emerg. med.

Vol. 18(2)

Mar 2011

Table 2. Result of univariate analysis of radiographic signs for differentiating small bowel obstruction by using logistic regression test Signs Presence of air-fluid levels Presence of differential air-fluid levels Presence of string-of-beads sign Presence of stretch sign Presence of decreased colon gas Number of air-fluid levels Number of differential air-fluid levels Maximal air-fluid level width Maximal small bowel diameter Maximal colon diameter Odds ratio 25.87 40.38 28.81 35.14 13.42 2.02 8.39 1.07 1.17 0.98 95% Confidence interval 10.97 15.59 6.23 7.69 6.32 1.65 4.03 1.05 1.12 0.96 61.04 104.63 133.16 160.66 28.46 2.46 17.49 1.09 1.23 1.01 p-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.2516

Table 3. Best cut-off values of numeric signs on plain abdominal radiography for differentiating small bowel obstruction Signs No. of air-fluid levels No. of differential air-fluid levels Maximal air-fluid level width Maximal small bowel diameter
*Real value calculated by SPSS

Cut-off value 2 (1.5*) 1 (0.5*) 31.5 mm 37.5 mm

Sensitivity 80% 64% 78% 74%

Specificity 89% 96% 93% 89%

Odds ratio 30.95 40.38 45.50 23.40

95% Confidence interval 13.34 15.59 18.68 10.52 71.81 104.63 110.84 52.04

p-value <0.0001 <0.0001 <0.0001 <0.0001

Table 4. The result of the stepwise selection of logistic regression test to select more effective radiographic signs for differentiating small bowel obstruction Signs Presence of differential air-fluid levels Presence of stretch sign Maximal air-fluid level width 31.5 mm Maximal small bowel diameter 37.5 mm Odds ratio 4.90 26.67 8.56 4.00 95% Confidence interval 1.34 3.67 2.53 1.29 17.97 193.56 28.95 12.40 p-value 0.0165 0.0012 0.0006 0.0163

Table 5. Comparison of clinical factors between small bowel obstruction and small bowel ileus Clinical factor Previous abdominal surgery Emergency abdominal surgery Death Total Small bowel obstruction No. of patients (%) 36 (72) 28 (56) 2 (4) 50 (100) Small bowel ileus No. of patients (%) 32 (19.3) 53 (31.9) 4 (2.4) 166 (100) p-value <0.001* 0.002* 0.624

*Result of Pearson's chi-square test; Result of Fisher's exact test

Kim et al./Differentiating small bowel obstruction from ileus

75

The existence or non-existence of previous abdominal surgery history and performance or non-performance of emergency abdominal surgery showed statistically significant differences (p=0.001, 0.002). In contrast, death or life did not show statistically significant differences between the SBO and SBI groups on final diagnosis (p=0.624) (Table 5). Po s t - o p e r a t i ve a d h e s i ve i l e u s c a s e s we re t h e overwhelming majority for causative disease of SBO (56%), and inguinal hernia, umbilical hernia, intussusception, inflammatory bowel disease, tumour, abscess, and foreign body causing mechanical obstruction in the bowel cavity were less frequent causes. The causative diseases of SBI were relatively evenly distributed although hepato-biliary disease, appendicitis, and gastroenteritis were slightly more prevalent than others (Table 6).

absence of structural obstruction is the most important consideration in clinical decision making, and this simple bisectional classification is widely used in the clinical field and can prevent confusion regarding the use of the term "ileus" among physicians. In our study of emergency department patients who had abdominal pain and small bowel distension on the PAR, the sensitivity and specificity of PAR for SBO were 82.0% and 96.4%, respectively. These results are far superior to the results of other plain radiographic studies conducted with patients clinically suspected of having SBO (sensitivity 63-77%, specificity 50-78%).4,9,12 Moreover these results nearly approach some of the reported sensitivities and specificities of abdominal computed tomography studies.4,9,12,32 According to our results, PAR should be regarded as a valuable initial tool in the emergency setting for differential diagnosis of small bowel disorders with impaired transit of intestinal contents. PAR is better in excluding SBO (diagnosing SBI) than in diagnosing SBO, because its specificity is superior to its sensitivity. Since Frimann's classic study, 14 air-fluid level and differential air-fluid level have been used to detect SBO. However, succeeding investigators viewed these signs as non-specific, although more frequent in SBO, because these signs could also be seen in cases of SBI.1 Other signs such as the string-of-beads sign, the stretch sign, and decreased colon gas have not been verified as specific.2 Recently, Lappas et al2 and Thompson et al16 evaluated the effectiveness of various classic signs on PAR through subdivision and numeric conversion of these signs. The subjects of these studies were limited to patients clinically suspected of having SBO, in contrast to our study, which included any emergency room patients with small bowel distension on PAR and abdominal pain. The study by Lappas et al 2 was also different from ours in that they subdivided small bowel disorders into high-grade SBO versus low-grade SBO and SBI. In our study, most of the plain radiographic signs were found to occur significantly more frequently in the final SBO group than in the final SBI group. These differences were more prominent in our study than in the results of other studies. 2,15,16 In our study, only

Discussion
The definition and classification of small bowel disorders with impaired transit of intestinal contents have been confusing. Diverse classifications and terms have been used in the clinical field. Some apply "ileus" only to the functional disorder and others apply it to both structural and functional disorders.1,2,8,28,29 SBO, mechanical obstruction, mechanical ileus, and obstructive ileus have been used as terms that mean a small bowel disorder with impaired transit of intestinal contents due to intrinsic luminal obstruction or extrinsic compression. In contrast, SBI, nonobstructive ileus, adynamic ileus, paralytic ileus, and postoperative ileus have all been used as terms that mean a small bowel disorder with impaired transit of intestinal contents due to functional impairment.1,2,7,8,28,29 Some authors subdivide SBI into spastic ileus, hypotonic ileus, and paralytic ileus according to the mechanism of abnormal motility. Some authors do not use the bisectional concept of SBO and SBI, but instead use gradual classification of normal small bowel gas pattern, abnormal but nonspecific small bowel gas pattern, probable small bowel obstruction pattern, and definitive small bowel obstruction pattern.7,30,31 Despite the controversies mentioned above, using the terms SBO and SBI seems rational, because the presence or

76

Hong Kong j. emerg. med.

Vol. 18(2)

Mar 2011

Table 6. Comparison of causative diseases in small bowel obstruction and small bowel ileus Small bowel obstruction Causative disease No. of patients (%) 1. Post operative adhesion 28 (56%) 2. Inguinal hernia 3 (6%) 3. Umbilical hernia 2 (4%) Intussusception 2 (4%) Inflammatory bowel disease 2 (4%) Peritoneal carcinomatosis 2 (4%) Colon cancer 2 (4%) 4. Appendicitis 1 (2%) Bezoar 1 (2%) Enterocolitis 1 (2%) Foreign body 1 (2%) Intestinal adhesion 1 (2%) Intraperitoneal abscess 1 (2%) Postoperative peritonitis 1 (2%) Radiation enteritis 1 (2%) Small bowel cancer 1 (2%) Small bowel ileus Causative disease No. of patients (%) 1. Gallstone 22 (13.3%) 2. Appendicitis 13 (7.9%) Common bile duct stone 13 (7.9%) 3. Enterocolitis 11 (5.7%) 4. Hepatocellular carcinoma 10 (5.1%) 5. Cholangiocarcinoma 9 (5.5%) Liver cirrhosis 9 (5.5%) 6. Peptic ulcer 7 (4.3%) Liver abscess 7 (4.3%) 7. Non-specific abdominal pain 6 (3.7%) 8. Acute pyelonephritis 5 (3.1%) Pancreatitis 5 (3.1%) 9. Intrahepatic duct stone 4 (2.5%) Ulcer perforation 4 (2.5%) 10. Ampulla of vater cancer 3 (1.8%) Pancreatic cancer 3 (1.8%) Pelvic inflammatory disease 3 (1.8%) 11. Bile peritonitis 2 (1.2%) Colon cancer 2 (1.2%) Colon perforation 2 (1.2%) Hepatitis 2 (1.2%) Gastrointestinal bleeding 2 (1.2%) Liver laceration 2 (1.2%) Ovarian tumour 2 (1.2%) Peritoneal carcinomatosis 2 (1.2%) Stomach cancer 2 (1.2%) 12. Aortic dissection 1 (0.6%) Biliary hamartoma 1 (0.6%) Biliary stricture 1 (0.6%) Cecal cancer 1 (0.6%) Corpus luteum rupture 1 (0.6%) Esophageal rupture 1 (0.6%) Intestinal Behcet's disease 1 (0.6%) Irritable bowel syndrome 1 (0.6%) Ischaemic colitis 1 (0.6%) Malaria 1 (0.6%) Mesenteric tumour 1 (0.6%) Ovarian cancer 1 (0.6%) Peritoneal tuberculosis 1 (0.6%) Total 50 patients (100%) Total 166 patients (100%)

Kim et al./Differentiating small bowel obstruction from ileus

77

maximal colon diameter was not significantly different between the two groups (p=0.2516); this result was contrary to the study of Thompson et al 16 in which expanded colon findings were significantly more frequent with SBI than with SBO. Further investigation is warranted to determine the effectiveness of using a finding of collapsed colon as a diagnostic factor for SBO. Lappas et al2 reported that the incidences of string-ofbeads sign were not significantly different between high-grade SBO versus low-grade SBO and SBI (39% and 23%, respectively; p=0.15). Thompson et al16 also reported that the incidence of string-of-beads sign was merely 10% in SBO and 1.7% in SBI although a significant difference was present between the two groups (p<0.001). Contrary to these results, in our study the difference in the incidences of string-of-beads sign was significant and prominent (26% in the final SBO group, 1.2% in the final SBI group; p<0.001). Harlow et al 15 reported that the frequencies of differential air-fluid levels were 52% in the SBO group and 29% in the SBI group. In our study, a more pronounced difference in the frequencies of differential air-fluid levels was demonstrated between the two groups (64% for SBO and 4.2% for SBI; p<0.001). The specificity of each numeric sign exceeded its sensitivity at cut-off value points on the ROC curves. This suggests that numeric signs are more effective in the exclusion of SBO than in its inclusion. This conclusion is consistent with the overall greater specificity than sensitivity of the PAR reading in our study. The incidences of previous abdominal surgery history and emergency abdominal surgery performance were significantly more frequent in the SBO group than in the SBI group. This seems to be reflective of the fact that a large number of the final SBO-group patients underwent adhesionolysis due to post-operative adhesions. The incidence of death was not significantly different between the two groups, which may imply an absence of a direct relation between the general condition of the patient and presence or absence of intestinal obstruction.

Miller et al 33 reported the most frequent causative diseases of SBO in this order: post-operative adhesion, inflammatory bowel disease, and malignant tumour. Laws and Aldrete et al34 reported the frequency in the order of post-operative adhesion, malignant tumour, and hernia. In our study, post-operative adhesion was the most common causative disease of SBO, although the frequency we found (56%) was lower than in the frequencies found in the prior studies (74% and 69%). Our study found hernia and malignant tumour were the next most common causative diseases (10% each). The frequency of inflammatory bowel disease was relatively low compared to the results of prior studies; this is likely due to racial differences in the distribution of this disease between Oriental and Western patients. SBI is caused by dysfunction of the intestinal autonomic ner vous system due to peritoneal inflammation, ascites, pain, haemodynamic instability, drug reaction, electrolyte imbalance, trauma, or other causes. 1,2,35 Individual frequencies of SBI in patients with abdominal pain, appendicitis, and cholecystitis have been mentioned in previous studies,18,36,37 but we believe our study is the first to investigate the types and frequencies of causative diseases in SBI. In our study, there was an even distribution of the abdominal diseases that were found to be causative of SBI. This might be explained by the fact that most abdominal diseases are accompanied by one or more factors that can cause dysfunction of the intestinal autonomic nerve system, including peritoneal inflammation, ascites, pain, haemodynamic instability, and electrolyte imbalance. The incidences of biliary disease and tumour were relatively high in our study, while the incidence of non-specific abdominal pain was relatively low. This finding may be the result of the generally high disease severity of this study group originating from the emergency room of a tertiary referral hospital. Although small bowel studies such as enteroclysis have been regarded as one of the most accurate radiographic examinations for the diagnosis of SBO, these studies are not suitable for emergency services because of they are time-consuming.7,24-27 With the recent widespread use of multi-detector computed tomography, the sensitivity of computed tomography in the diagnosis of SBO has been reported as nearly 100%. 6,38 Given that computed tomography can also identify the cause

78

Hong Kong j. emerg. med.

Vol. 18(2)

Mar 2011

of SBO, it is regarded as the most appropriate final radiographic examination available during emergency services.5,7-10,12,13,20-22,39,40 Among the non-surgical cases in our study, computed tomography was used as the actual standard for final diagnosis, with the exception of one patient who underwent a small bowel study. Although PAR is the most rapid and the most convenient radiographic examination in the initial evaluation of abdominal pain, there have recently been many controversies concerning its effectiveness. Some authors advocate computed tomography as a more effective initial radiographic examination than PAR for patients with abdominal pain.18,40-42 However, routine performance of computed tomography on all patients with abdominal pain is difficult to carry out, due to practical limitations such as delay of the treatment process, increased radiation exposure, and increased medical expense. Using PAR to examine components that it best detects or differentiates can be effective for reducing the demand for abdominal computed tomography. There were several limitations to our study. First, because it was a retrospective design, selection bias was inevitable. Furthermore, selection bias could have been generated from the inclusion criteria of successive abdominal computed tomography, which might imply that our study population showed increased disease severity. Second, only one radiologist participated as a reviewer. However, adding the clinical experience of an emergency physician to the interpretation might lead to more relevant results for the intention of this study than would the co-review of multiple radiologists since our study was targeted to an emergency room setting. Third, the degree of bowel obstruction was not assessed. However, from the point of view of emergency treatment, accurate screening of patients with SBO and rapid referral of these patients to a s u r g e o n w a s m o re i m p o r t a n t t h a n f u r t h e r discrimination between high-grade SBO and partial SBO. In conclusion, PAR is a valuable initial radiographic examination in the differentiating SBO from SBI, and the diagnostic accuracy of PAR nears that of computed tomography. Most of the classic radiographic signs

(except decrease of colon diameter) are useful on differentiating between SBO and SBI among patients suffering from small bowel dilatation as shown on X-ray.

References
1. Messmer JM. Gas and soft tissue abnormalities. In: Gore RM, Levine MS, eds. Textbook of gastrointestinal radiology. 3rd ed. Philadelphia: Saunders; 2008: 208-5. 2. Lappas JC, Reyes BL, Maglinte DD. Abdominal radiography findings in small-bowel obstruction: relevance to triage for additional diagnostic imaging. AJR Am J Roentgenol 2001;176(1):167-74. 3. Maglinte DD, Reyes BL, Harmon BH, Kelvin FM, Turner WW, Hage JE, et al. Reliability and role of plain film radiography and CT in the diagnosis of small-bowel obstruction. AJR Am J Roentgenol 1996;167(6): 1451-5. 4. Gazelle GS, Goldberg MA, Wittenberg J, Halpern EF, Pinkney L, Mueller PR. Efficacy of CT in distinguishing small-bowel obstruction from other causes of smallbowel dilatation. AJR Am J Roentgenol 1994;162(1): 43-7. 5. Frager D, Medwid SW, Baer JW, Mollinelli B, Friedman M. CT of small-bowel obstruction: value in establishing the diagnosis and determining the degree and cause. AJR Am J Roentgenol 1994;162:37-41. 6. Maglinte DD, Heitkamp DE, Howard TJ, Kelvin FM, Lappas JC. Current concepts in imaging of small bowel obstruction. Radiol Clin North Am 2003;41(2):26383, vi. 7. Frager DH, Baer JW, Rothpearl A, Bossart PA. Distinction between postoperative ileus and mechanical small-bowel obstruction: value of CT compared with clinical and other radiographic findings. AJR Am J Roentgenol 1995;164(4):891-4. 8. Batke M, Cappell MS. Adynamic ileus and acute colonic pseudo-obstruction. Med Clin North Am 2008;92(3): 649-70, ix. 9. Suri S, Gupta S, Sudhakar PJ, Venkataramu NK, Sood B, Wig JD. Comparative evaluation of plain films, ultrasound and CT in the diagnosis of intestinal obstruction. Acta Radiol 1999;40(4):422-8. 10. To r re g g i a n i WC , H a r r i s AC , Ly b u r n I D , a l Nakshabandi NA, Zwirewich CV, Brenner C, et al. C o m p u t e d t o m o g r a p h y o f a c u t e s m a l l b owe l obstruction: pictorial essay. Can Assoc Radiol J 2003; 54(2):93-9. 11. Maglinte DD, Balthazar EJ, Kelvin FM, Megibow AJ. The role of radiology in the diagnosis of small-bowel obstruction. AJR Am J Roentgenol 1997;168(5):117180. 12. Miyazaki O. Efficacy of abdominal plain film and CT in bowel obstruction. Nippon Igaku Hoshasen Gakkai Zasshi 1995;55(4):233-9.

Kim et al./Differentiating small bowel obstruction from ileus

79

13. Taourel PG, Fabre JM, Pradel JA, Seneterre EJ, Megibow AJ, Bruel JM. Value of CT in the diagnosis and management of patients with suspected acute smallbowel obstruction. AJR Am J Roentgenol 1995;165(5): 1187-92. 14. Frimann-Dahl J. Roentgen findings in intestinal knots. Acta Radiol 1942;23:22-33. 15. Harlow CL, Stears RL, Zeligman BE, Archer PG. Diagnosis of bowel obstruction on plain abdominal radiographs: significance of air-fluid levels at different heights in the same loop of bowel. AJR Am J Roentgenol 1993;161(2):291-5. 16. Thompson WM, Kilani RK, Smith BB, Thomas J, Jaffe TA, Delong DM, et al. Accuracy of abdominal radiography in acute small-bowel obstruction: does reviewer experience matter? AJR Am J Roentgenol 2007; 188(3):W233-8. 17. Vicario SJ, Price TG. Intestinal obstruction. In: Tintinalli JE, Kelen GD, Stapczynski JS, eds. Emergency medicine: a comprehensive study guide. 6th ed. New York: McGraw-Hill 2004:523-6. 18. Eisenberg RL, Heineken P, Hedgcock MW, Federle M, Go l d b e r g H I . Eva l u a t i o n o f p l a i n a b d o m i n a l radiographs in the diagnosis of abdominal pain. Ann Surg 1983;197(4):464-9. 19. Thompson WM. Gasless abdomen in the adult: what does it mean? AJR Am J Roentgenol 2008;191(4):1093-9. 20. Fukuya T, Hawes DR, Lu CC, Chang PJ, Barloon TJ. CT diagnosis of small-bowel obstruction: efficacy in 60 patients. AJR Am J Roentgenol 1992;158(4):765-9. 21. Ha HK, Park CH, Kim SK, Chun CS, Kim IC, Lee HK, et al. CT analysis of intestinal obstruction due to adhesions: early detection of strangulation. J Comput Assist Tomogr 1993;17(3):386-9. 22. Obuz F, Terzi C, Sokmen S, Yilmaz E, Yildiz D, Fuzun M. The efficacy of helical CT in the diagnosis of small bowel obstruction. Eur J Radiol 2003;48(3):299-304. 23. Rubesin SE, Gore RM. Small bowel obstruction. In: Gore RM, Levine MS, eds. Textbook of gastrointestinal radiology. 3rd ed. Philadelphia: Saunders; 2008:87199. 24. Maglinte DD, Herlinger H, Nolan DJ. Radiologic features of closed loop obstruction: analysis of 25 confirmed cases. Radiology 1991;179(2):383-7. 25. Maglinte DD, Kelvin FM, OConnor K, Lappas JC, C h e r n i s h S M . Cu r re n t s t a t u s o f s m a l l b owe l radiography. Abdom Imaging 1996;21(3):247-57. 26. Maglinte DD, Lappas JC, Kelvin FM, Rex D, Chernish SM. Small bowel radiography: how, when, and why? Radiology 1987;163(2):297-305. 27. Shrake PD, Rex DK, Lappas JC, Maglinte DD. Radiographic evaluation of suspected small bowel obstruction. Am J Gastroenterol 1991;86(2):175-8.

28. Galanis I, Dragoumis D, Tsolakis M, Zarampoukas K, Zarampoukas T, Atmatzidis K. Obstructive ileus due to a giant fibroepithelial polyp of the anus. World J Gastroenterol 2009;15(29):3687-90. 29. Pa p a v r a m i d i s TS , Po t s i S , Pa r a m y t h i o t i s D , Michalopoulos A, Papadopoulos VN, Douros V, et al. Gallstone obstructive ileus 3 years post-cholecystectomy to a patient with an old ileoileal anastomosis. J Korean Med Sci 2009;24(6):1216-9. 30. Grassi R, Di Mizio R, Pinto A, Romano L, Rotondo A. Serial plain abdominal film findings in the assessment of acute abdomen: spastic ileus, hypotonic ileus, mechanical ileus and paralytic ileus. Radiol Med 2004; 108(1-2):56-70. 31. Patel NH, Lauber PR. The meaning of a nonspecific abdominal gas pattern. Acad Radiol 1995;2(8):667-9. 32. Maglinte DD, Gage SN, Harmon BH, Kelvin FM, Hage JP, Chua GT, et al. Obstruction of the small intestine: accuracy and role of CT in diagnosis. Radiology 1993;188(1):61-4. 33. Miller G, Boman J, Shrier I, Gordon PH. Etiology of small bowel obstruction. Am J Surg 2000;180(1):33-6. 34. Laws HL, Aldrete JS. Small-bowel obstruction: a review of 465 cases. South Med J 1976;69(6):733-4. 35. Delgado-Aros S, Camilleri M. Pseudo-obstruction in the critically ill. Best Pract Res Clin Gastroenterol 2003; 17(3):427-44. 36. Olutola PS. Plain film radiographic diagnosis of acute appendicitis: an evaluation of the signs. Can Assoc Radiol J 1988;39(4):254-6. 37. Rothrock SG, Gorrhuis H, Howard RM. Efficacy of plain abdominal radiography in patients with biliary tract disease. J Emerg Med 1990;8(3):271-5. 38. Yaghmai V, Nikolaidis P, Hammond NA, Petrovic B, Gore RM, Miller FH. Multidetector-row computed tomography diagnosis of small bowel obstruction: can coronal reformations replace axial images? Emerg Radiol 2006;13(2):69-72. 39. Balthazar EJ, Birnbaum BA, Megibow AJ, Gordon RB, W h e l a n C A , Hu l n i c k D H . C l o s e d - l o o p a n d strangulating intestinal obstruction: CT signs. Radiology 1992;185(3):769-75. 40. Mindelzun RE, Jeffrey RB. Unenhanced helical CT for evaluating acute abdominal pain: a little more cost, a lot more information. Radiology 1997;205(1):43-5. 41. Baker SR. Unenhanced helical CT versus plain abdominal radiography: a dissenting opinion. Radiology 1997;205(1):45-7. 42. MacKersie AB, Lane MJ, Gerhardt RT, Claypool HA, Keenan S, Katz DS, et al. Nontraumatic acute abdominal pain: unenhanced helical CT compared with three-view acute abdominal series. Radiology 2005;237 (1):114-22.