di Emanuela Capalbo, Farideh Sajadidehkordi, Anna Kluzer, Paola Mariani, Maurizio Cariati
Abstract Objectives: evaluating diagnostic performance of multi-slice computed tomography (MSCT) in emergency department (ED), in characterization of site and type of bowel obstruction and comparing with surgical findings. Methods: we selected 119 patients positive for bowel obstruction. The surgical treatment is performed up to 48 hours after imaging. We obtained the written informed consent by patients and Ethical Committee approval. We calculated sensitivity, specificity, positive (PPV) and negative (NPV) predictive value and diagnostic accuracy. Diagnostic concordance of cause and site has been evaluated using K Cohen’s coefficient. Results: the main causes of bowel obstruction, according to MSCT are: adhesion 21%, inflammation 13,4%, tumors 10,9%, volvulus-intussusception 10,9%. MSCT for diagnosis of cause has a sensitivity of 96,9%, specificity 85%, PPV 96,9%, NPV 85% and diagnostic accuracy 94,9%.The localization was in duodenum 15,9%, jejunum 12,6%, ileum 17,6%, right colon 15,1%, left colon 13,4% and sigma/rectum 9,2%. MSCT for diagnosis of site has a sensitivity of 97,5%, specificity 87,2%, PPV 97,8%, NPV 86,5% and diagnostic accuracy 96,1%.The cause and site is indeterminate in 15,9%. K=0,75 for cause and 0,71 for site of bowel obstruction. Conclusions: MSCT identifies with a good concordance both cause and localization of the bowel obstruction compared with surgical diagnosis. It is diagnostic reference-standard in ED to define the patient’s therapy planning. Keywords: MSCT, diagnosis, emergency, bowel occlusion. Introduzione Since the introduction of multi-slice computed tomography (MSCT), a progressive increase in the use of this method has been registered for both elective and emergency evaluation of intestinal pathology, and in particular where there is suspicion of bowel obstruction. Many emergency-urgency conditions occur with acute abdominal pain, among them there is the bowel obstruction that occurs also with vomiting and bowel closed by feces and gas . In many cases it is difficult to make a diagnosis based only on the clinical and laboratory tests; to this end the imaging has often a decisive role . Radiological exams used in these cases are: X-ray, ultrasound and MSCT of the abdomen, and the choice between them depends on the severity of symptoms, on the anamnesis of patient and on the decision of the clinician and the radiologist. The MSCT provides a lot of information in short scan times and allows evaluating the site of mechanical occlusion and whether it is or not associated with intestinal ischemia. This will help in the choice between medical or surgical treatment, and in the latter case between laparotomy or laparoscopy . Bowel obstruction may be localized at small or large intestine and, according to the cause; it is classified in mechanical (adhesions, internal and external hernias, diverticulitis, tumors, volvulus, gallstone ileus, endometriosis and chronic inflammatory diseases) and dynamic. The dynamic ileus is the result of a significant intestinal contraction  or the evolution of a mechanical ileus unrecognized or not promptly treated . Adhesions and hernias are the most common causes of small bowel obstruction, reported by approximately 80% of cases . Tumors, sigmoid diverticulitis and volvulus are the most common causes of large bowel obstruction and together constitute 80-85% of cases . Aim of this paper is to evaluate the diagnostic performance of MSCT carried out in the emergency department (ED) in making an accurate diagnosis of cause and location of intestinal obstruction, correlating the radiological diagnosis with surgical findings. Material and methods Selection of Patients In 2013 we recorded 8256 patients in emergency Radiology for non-traumatic acute abdominal pain; 1372 of these performed a MSCT (16,7%). For this study, we selected retrospectively 119 patients with bowel obstruction. We included patients with high suspicious of bowel occlusion at triage, and/or after surgery visit and radiological diagnosis. Then we examined the medical records considering the surgery as reference diagnostic. The surgical treatment has been performed up to 48 hours after imaging. We obtained the written informed consent by patients and Ethical Committee approval for this research study. MSCTs have been evaluated by many radiologists, with years of experience ranging from 5 to 20, during different work shifts at ED. All medical reports described the location and cause of bowel obstruction. Subsequently, the images were reviewed in blind by two experienced radiologists (18 and 20 years of experience) who pointed out the presence of signs in the MSCT, suggestive of bowel obstruction. For the radiological diagnosis of obstruction cause, we grouped the lesions as follows: cancer, inflammation (mainly diverticulitis and appendicitis), adhesions, internal and external hernias, volvulus and intussusception, gallstone ileus, endometriosis, inflammatory bowel disease (Crohn’s disease and ulcerative colitis), fitobezoar/ tricobezoar, and external bodies. The site of occlusion was divided into: duodenum, jejunum, ileum, right colon (from cecum to transverse colon included), left colon (from splenic flexure to the sigmoid excluded) and rectum/sigma. TC method The examinations were performed using a 16-slice CT Lightespeed (GE Medical System, Milwauke WI, USA). The protocol foresees an acquisition from the diaphragmatic dome to the pelvis in basal conditions and, based on the clinical suspicion and according to the indications given in literature , intravenously administration of non-ionic iodinated contrast medium (concentration 370 mg/dl, dosage 1,5-2ml/kg, flow 2,5-3ml/sec). In such cases after injection of contrast medium we completed the exam with an early acquisition after 35 sec (arterial phase) and/or late acquisition after 90 seconds (portal phase). We set the following parameters: collimation 16x0.75mm, FOV 320-360mm, pitch 1,75, voltage 120kV, automatic modulation of tube current (mA), slice thickness 3,75 mm, imaging reconstruction to 1,25 or 2,5mm with a standard filter for multiplanar evaluations, display window for abdomen. In no case the contrast medium was administered orally or rectally. Statistical Analysis The data were analyzed using SPSS version 18. The baseline characteristics of the study population were expressed as mean±standard deviation for continuous variables, and as percentages for categorical variables. The diagnostic concordance between radiological and surgical diagnosis was evaluated by using the K test of Cohen, both for the cause and the site of the occlusion. The concordance was also calculated for the retrospective diagnosis performed by radiologists A and B, as well as between the original and the retrospective diagnosis. Finally, we calculated sensitivity, specificity, diagnostic accuracy, positive (PPV) and negative (NPV) predictive value. Results The general characteristics of the patients included in this study are summarized in Tab. 1. In 97 patients out of the 102, the cause of the occlusion described by MSCT was confirmed by surgical diagnosis (Fig. 1-2). In 3 patients there was a misdiagnosis of endometriosis, adherence and inflammation which were not seen in the intervention. In 2 patients we had an incorrect diagnosis of the cause: an inflammation with abscess and an external body were found to be respectively a tumor and a gallstone ileus in surgery In the remaining 2 patients, the persistence and severity of the symptoms guided the surgeons to perform an explorative laparoscopy, identifying an occlusive disease: it was one case of fitobezoar located in the cecum in a young patient and one tumor in the rectum in a woman with diverticular disease. The diagnosis of cause performed with MSCT has a sensitivity of 97%, with a specificity of 77,2%, PPV and NPV of 95% and 85% respectively, with a diagnostic accuracy of 95,7%. The concordance between MSCT and surgery has a k=0,75 and p<0.05 (Tab. 2). For the diagnosis of site, MSCT has a sensitivity of 96,9%, a specificity of 85%, a PPV of 96,9% and VPN of 85% and a diagnostic accuracy of 94,9%. The concordance was found to have a k=0,71 p<0,05 (Tab. 3). The radiologists A and B evaluated retrospectively the presence or not of the principal bowel obstruction signs (Tab. 4). These statistical parameters, both for cause and site were subsequently evaluated also by radiologists A and B (Tab. 5): there was a high diagnostic concordance of cause and site, respectively k=0,83 and k=0,86. Sensitivity, specificity, NPV,PPV and diagnostic accuracy were slightly higher than those obtained from the diagnosis made by radiologists with experience rangin Discussion Intestinal occlusion is frequent in emergency practice, with incidence of the hospitalizations for acute abdomen of 20% [8, 9]. It is a significant cause of morbidity and mortality and an appropriate treatment depends by early diagnosis and by accurate identification of patients who require surgery . In literature the MSCT is described able to exclude the disease or identify site, cause and severity [9, 10] and it is useful in the therapeutic decision-making [11, 12] as it helps the surgeon to determine the type of intervention . It is a highly accurate method in the diagnosis of occlusion, with a sensitivity, specificity and accuracy of respectively 90-96%, 96% and 95% [14, 10]. In our study, MSCT (Fig. 1, 2) performed in patients with clinical suspicion of intestinal obstruction confirmed high sensitivity, specificity and diagnostic accuracy (Tab. 6). The diagnosis of occlusion by MSCT is based on the identification of dilated proximally intestinal loops and collapsed distally intestinal loops; identifying the point where there is change of intestinal caliber is critical (Fig. 1, 2) [15, 11]. The interpretation of axial images is essential (Fig. 1a, 2a, 2b), although the associated analysis of the MPR (Fig. 1b, 1c, 2c) helps in recognizing the obstruction localization [16, 17, 12,13]. The two modalities improve the visualization of the transition zone (Fig. 1c) [15, 18, 19, 20, 11, 14, 15, 16] with accuracy between 90 [20, 16] and 94% [15, 11], in agreement with values obtained in our study (Tab. 6). According to Filippone [21, 17], small intestine occlusions (Fig. 2c) are better diagnosed by axial acquisitions than the coronal reconstructions; however large intestine occlusions (Fig.1c) are better diagnosed by coronal reconstructions. Among the various radiological signs of intestinal obstruction the “sign of the feces” (Fig. 1C) (presence of structured material within dilated loop of small intestine, immediately upstream of the transition zone) has been described as little sensible and specific for the diagnosis of occlusion especially of the small intestine [21, 22, 23, 17, 18, 19], in fact it can be present even in the absence of occlusion [15, 11]. While according to Wang  it would be predictive of intestinal ischemia [13, 2,3 20, 19]. In our study, we found it in 7 of 19 patients (36,8%) with ischemia (Tab. 1). Pneumatosis has been described as predictive sign of ischemia, but its absence does not exclude the presence of ischemia . Among 8 patients with pneumatosis described in our study, 5 had ischemia (Tab. 1). The “whirlpool sign”, originally defined by Fisher as “loop of the small intestine coiled around superior mesenteric artery” [24, 21], was also associated with volvulus of the large intestine [25, 22]. According to literature, diagnostic value for volvulus of “whirlpool sign” is limited: we found it only in 23% of patients with volvulus (Tab. 1, 3). The intra-abdominal effusion is a sign of a serious occlusive condition and it is often associated with ischemia . 80% of patients with intra-abdominal effusion had ischemia; the other 20% had cancer or inflammation. In several studies the accuracy of MSCT was difficult to assess objectively because of the variable elapse between the radiological examination and surgical intervention [13, 20]. This bias is not present in our study because the maximum time between the MSCT and the surgery is 48 hours. Tabella 7. Types of vascular occlusion. VASCULAR OCCLUSION N (%) Mesenteric artery 9 (47,4) Celiac origin 3 (15,8) Portal Vein 4 (21,1) Mesenteric vein 2 (10,5) Mesenteric vein+splenic vein 1 (5,2) Total 19 (100%) 14 Special Articles Italian Journal of Emergency Medicine - Febbraio 2015 In this study we found a good concordance between the retrospective and the initial diagnosis, according to Wang [13, 20]. A good knowhow of the TC signs and TC characteristics of the bowel obstruction [26, 23] is essential for a correct diagnosis and appropriate treatment  Limitations The limitations of this paper is the relatively small number of patients analysed and the possible selection bias due to the retrospective study. Conclusions MSTC performed in emergency has a high diagnostic accuracy in diagnosis of cause and site of intestinal obstruction, reducing morbidity and mortality. We showed a high concordance between radiological and surgical diagnosis, also thanks to the use of the MPR. Therefore, nowadays the MSCT represents the reference-standard for the diagnosis of intestinal obstruction. References 1. Harrison TR (2005) Principi di Medicina Interna.Vol II, 16rd edn, McGraw-Hill Companies, srl Publishing Italia, Milano. 2. Unni K, Udayasankar Jianhai Li DA, Baumgarten WC et al. Acute abdominal pain: value of non contrast enhanced ultra-low-dose multi detector row CT as a substitute for abdominal radiographs. Emerg Radiol 2009; 16: 61-70. 3. Hayakawa K, Tanikake M, Yoshida S et al. Radiological diagnosis of large-bowel obstruction: non neoplastic etiology. Jpn J Radiol 2012; 30: 541-552. 4. Passariello R, Simonetti G. Compendio di Radiologia, 3rd edn, Idelson-Gnocchi,Napoli 2010. 5. Santillan CS. Computed Tomography of Small Bowel Obstruction. Radiol Clin N Am 2013; 51: 17-27. 6 aourel P, Kessler N, Lesnik A et al. Helical CT of large bowel obstruction. Abdom Imaging 2001; 28: 267-75. 7. Leschka S, Alkadhi H, Wildermuth S et al. Multi-detector computed tomography of acute abdomen. Eur Radio 2005; 15: 2435-2447. 8. Scaglione M, Grassi R, Pinto A et al. Positive predictive value and negative predictive value of spiral CT in the diagnosis of closed loop obstruction complicated by intestinal ischemia. Radiol Med 2004; 107 (1-2): 69-77. 9. Nicolaou S, Kai B, Ho S et al Imaging of acute small-bowel obstruction. AJR Am J Roentgenol 2005; 185 (4): 1036- 44. 10. Qalbani A, Paushter D, Dachman AH. Multidetector Row CT of Small Bowel Obstruction. Radiol Clin North Am. 2007; 45 (3): 499-512. 11. Rosen MP, Sands DZ, 3rd et al. Impact of abdominal CT on the management of patients presenting to the emergency department with acute abdominal pain. AJR Am J Roentgenol 2000; 174: 1391-1396. 12. Udayasankar UK, Li J, Baumgarten DA et al. Acute abdominal pain: value of non-contrast enhanced ultra-low-dose multi-detector row CT as a substitute for abdominal radiographs. Emerg Radiol 2009; 16: 61-70. 13. Wang Q, Chavhan GB, Babyn PS et al. Utility of CT in the diagnosis and management of small-bowel obstruction in children. Pediatr Radiol. 2012; 42: 1441-1448. 14. Silva AC, Pimenta M, Guimaraes LS. Small bowel obstruction: what to look for. Radiographics. 2009; 29: 423–439 15. Angelelli G, Moschetta M, Cosmo T et al. CT Diagnosis of the nature of bowel obstruction: morphological evaluation of the transition point. Radiol Med. 2012; 117: 749-758. 16. Horton KM, Fishman EK. The current status of multidetector row CT and three-dimensional imaging of the small bowel. Radiol Clin North Am. 2003; 41 (2): 199-212. 17. Jaffe TA, Martin LC, Thomas J et al. Small-bowel obstruction: coronal reformations from isotropic voxels at 16-section multi-detector row CT. Radiology. 2006; 238 (1): 135-42. 18. Aufort S, Charra L, Lesnik A et al. Multidetector CT of bowel obstruction: value of post-processing. Eur Radiol. 2005; 15: 2323-2329. 19. Sinha R, Verma R. Multidetector row computed tomography in bowel obstruction. Part 1. Small-bowel obstruction. Clin Radiol. 2005; 60: 1058-1067. Italian Journal of Emergency Medicine - Febbraio 2015 Special Articles 15 20. Hodel J, Zins M, Desmottes L et al. Location of the transition zone in CT of of bowel obstruction: added value of multiplanar reformations. Abdon Imaging. 2009; 34: 35-41. 21. Filippone A, Canci R, Storto ML. Bowel obstruction: comparison between multidetector-row Ct axial and coronal planes. Abdon Imaging. 2007; 32-310-316. 22. Catalano O. The feces sign: a Ct findings in small bowel obstruction. Radiology. 1997; 37: 417-419. 23. Lazarus DE, Slywotsky C, BBennett GL et al. Frequency and relevance of the “small-bowel feces” sign on CT in patients with small bowel-obstruction. AJR. 2004; 183: 1361-1366. 24. Fisher JK. Computed tomographic diagnosis of volvulus in intestinal malrotation. Radiology. 1991; 140: 145-146. 25. Gollub MJ, Sora Y, Smith L McG et al. Does the CT whirl sign really predict small bowel volvulus? Experience in an oncologic population. J Comput Assist Tomogr. 2006; 30: 25-32. 26. Hayakawa K, Tanikake M, Yoshida S et al. Radiological diagnosis of large-bowel obstruction: non neoplastic etiology. Jpn J Radiol. 2012; 30: 541-52.