Emergency. 2017; 5 (1): e61 OR I G I N A L RE S E A RC H Diagnostic Accuracy of Computed Tomography Scan in Detection of Upper Gastrointestinal Tract Injuries Follow- ing Caustic Ingestion Hooman Bahrami-Motlagh1, Mohammad Hadizadeh-Neisanghalb2, Hassan Peyvandi2, 3∗ 1. Radiology Department, Loghman Hakim Hospital, Shahid Beheshti University of Medical Sciences, School of Medicine, Tehran, Iran. 2. General Surgery Department, Loghman Hakim Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran. 3. Hearing Disorders Research Center, Loghman Hakim Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran. Received: February 2017; Accepted: March 2017; Published online: 10 March 2017 Abstract: Introduction: Endoscopy is an invasive procedure and finding noninvasive alternative tools in detection of probable upper gastrointestinal (GI) tract injuries following caustic ingestion is an area of interest. The present study aimed to evaluate the screening performance characteristics of thoraco-abdominal computed tomogra- phy (CT) scan in this regard. Methods: This prospective cross sectional study was conducted on patients pre- senting to emergency department following acute caustic ingestion. The findings of CT scan and endoscopy re- garding the presence of upper GI tract damage were compared and screening performance characteristics of CT scan were calculated using MedCalc software. Results: 34 patients with the mean age of 35.38±13.72 years were studied (58.8% male). The agreement rate between CT scan and endoscopy regarding the grade of esophageal and gastric injuries was moderate (K= 0.38; p = 0.001) and fair (K= 0.17; p = 0.038), respectively. The sensitivity and specificity of CT scan in detection of esophageal damage were 96.29) 79.11- 99.80) and 57.14 (20.23 - 88.19), respectively. These measures were 89.65 (71.50 - 97.28) and 40.00 (7.25 - 82.95), respectively for gastric damage. The area under the ROC curve of CT scan in detection of esophageal and gastric damages was 0.76 (95% CI: 0.52 – 1.00) and 0.64 (95% CI: 0.35 – 0.94), respectively. Conclusion: Based on the findings of the present study, CT scan could be considered as a sensitive tool in ruling out upper gastrointestinal mucosal injuries following acute caustic ingestions. However, the correlation between endoscopy and CT scan findings regarding the grading of injury is not high enough to eliminate the need for endoscopy. Keywords: Caustics; eating; Burns, Chemical; Tomography, X-Ray Computed; Endoscopy © Copyright (2017) Shahid Beheshti University of Medical Sciences Cite this article as: Bahrami-Motlagh H, Hadizadeh-Neisanghalb M, Peyvandi H. Diagnostic Accuracy of Computed Tomography Scan in Detection of Upper Gastrointestinal Tract Injuries Following Caustic Ingestion. Emergency. 2017; 5(1): e61. 1. Introduction I ngestion of caustic substances is one of the toxicology emergencies that are associated with relatively high mor- bidity and mortality (1). The severity of tissue damage de- pends on the type, concentration, volume of ingestion, and contact duration (2, 3). The gold standard tool in assessment of mucosal damage is esophagogastroduodenoscopy (EGD) within the first 12 hours of the incident (4, 5). 4 days af- ∗Corresponding Author: Hassan Peyvandi; Hearing Disorders Research Cen- ter, Loghman Hakim Hospital, Makhsus St., Qazvin St., Tehran, Iran. Email: hassan.peyvandi@gmail.com Tel: +98-912-3260870; Fax: +98-21-55416130 ter the incident, EGD is not recommended due to the risk of perforation (6, 7). Some believe that, EGD should be per- formed in all patients except those who have indication of emergent surgery (6, 8, 9). However, it is an invasive proce- dure and there are contradicting opinions about performing endoscopy in asymptomatic patients. Thoraco-abdominal computed tomography (CT) scan has been widely used in cases of caustic ingestion to gather more details about the surrounding tissues injury (10, 11). Lurie et al. showed that CT scan underestimates the severity of caustic-related gas- trointestinal injuries compared to EGD (12). However, there is little evidence about the diagnostic accuracy of CT scan in detection of upper gastrointestinal mucosal injuries fol- lowing caustic ingestion. The present study aimed to eval- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com H. Bahrami-Motlagh et al. 2 uate the screening performance characteristics of thoraco- abdominal CT scan in this regard. 2. Methods 2.1. Study design and setting This prospective cross sectional study was conducted on patients presenting to emergency department of Loghman Hakim Hospital, Tehran, Iran, in 2015, following acute caus- tic ingestion. All participants provided written informed con- sent, and the study protocol was approved by the ethics committee of Shahid Beheshti University of Medical Sci- ences. Researchers adhered to the Helsinki recommenda- tions throughout the study period. 2.2. Participants All adult (> 15 years old) patients, presented to the emergency department of the mentioned hospital during the study pe- riod were enrolled. This hospital is the biggest poisoning center of Tehran, Iran capital. Patients with unstable hemo- dynamics, third degree burns of the hypopharynx, respira- tory distress, and positive history of a chronic disease or le- sion in stomach or esophagus as well as those suspected to gastrointestinal perforation were excluded. 2.3. Procedures Upper gastrointestinal tract endoscopy was performed by ex- pert gastroenterologists within 24 hours of admission to hos- pital. Concurrently, thoraco-abdominal CT scan with intra- venous urografin 76% contrast material was carried out for all patients. CT scan slides were reviewed by an expert radi- ologist who was blinded to endoscopic and clinical findings of patients. The endoscopy and CT scan grading of gastroin- testinal mucosal injury were done based on appendix 1 (13, 14). 2.4. Data gathering A checklist that consisted of demographic data (sex, age), type of ingested substance (acid, alkaline), volume of inges- tion, duration of hospital stay, and time from event to arriv- ing at the hospital as well as endoscopy and CT scan grading of upper gastrointestinal injuries was used for data gather- ing. A trained surgery resident was responsible for collection of patients’ data. 2.5. Statistical Analysis Data were analyzed using SPSS version 21. Continuous vari- ables were presented as mean ± standard deviation and qualitative ones as frequency and percentage. The corre- lation between esophagus and stomach grading of injuries was calculated using Spearman rank correlation coefficient. The inter-rater agreement between CT scan and endoscopy grading was measured using calculation of Cohen’s kappa coefficient (K). K=0 defined as no agreement; 0= 55 5 (14.7) Sex Male 20(58.8) Female 14(41.2) Type of substance Acid 30(88.2) Alkaline 3(8.8) Volume of ingestion (mL) 92.42 ± 89.78 Hospital stay (day) 3.88 ± 2.11 Time to hospital (hours) 9.69 ± 15.04 Data were presented as mean ± standard deviation or fre- quency and percentage. pertechnetate has high agreement with EGD in detection of gastric injury, but this method is not affordable and avail- able in all health centers (17). There is little evidence regard- ing the accuracy of CT scan in this regard (6). Some studies have mentioned various advantages of CT scan like its avail- ability, feasibility and ability to specify the extension of extra gastrointestinal tract involvement (10, 11, 13). CT scan has been more valuable in the evaluation and approval of endo- scopic gastric perforation (13, 18). Ryu et al. showed that CT scan has high sensitivity and specificity in predicting com- plications such as esophageal stricture in patients with caus- tic substances ingestion (19). Lurie and their colleagues as- sessed the role of CT scan in detecting the severity of mucosal injury due to corrosive ingestion and concluded that, deci- sion regarding the need for surgery should not be made solely based on CT scan findings (12). In a recent review by Chirica et al. 2016, CT scan was superior to endoscopy in screening of patients in need for emergent surgery (20). In this study, screening performance characteristics of CT scan for detec- tion of esophageal and gastric injuries were about the same range. Despite the high sensitivity and negative predictive value, specificity and positive predictive value of the test were not that high and therefore, this introduces CT scan as a tool for ruling out injury and not for ruling in it. The overall ac- curacy of the test based on area under the ROC curve is poor to moderate. These findings are in line with the results of the study by Lurie et al. (12). In addition to being a less invasive, easy and fast method, CT scan can give important informa- tion in the field of pulmonary infiltration and surrounding thoracic soft tissue involvement. Additionally, in some situa- tions such as upper airway inflammation, delayed visit (after 4 days and the risk of perforation), and absence of a skilled endoscopist, CT scan could be the right choice for ruling out upper gastrointestinal tract injuries. 5. Limitation Low sample size was one of the limitations of this study. All CT scans were interpreted by one expert radiologist, while it was better if two radiologists reviewed the CT slides and cases of disagreement were discussed with a third radiologist. 6. Conclusion Based on the findings of the present study, CT scan could be considered as a sensitive tool in screening upper gas- trointestinal mucosal injuries following caustic ingestions. In other words, CT scan could be used for ruling out mucosal injury in this setting. However, the correlation between en- doscopy and CT scan findings regarding the grading of injury is not high enough to eliminate the need to endoscopy. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com H. Bahrami-Motlagh et al. 4 Table 2: Endoscopy and CT scan grading of patients’ upper gastrointestinal tract injuries Location Grading of injuries number (%) Normal I II III IV Endoscopy Esophagus 7 (20.6) 15 (44.1) 10 (29.4) 2 (5.9) 0 (0) Stomach 5 (14.7) 5 (14.7) 21 (61.8) 3 (8.8) 0 (0) CT scan Esophagus (proximal) - 17 (50.0) 11 (32.4) 6 (17.6) 0 (0) Esophagus (distal) - 14 (41.2) 12 (35.3) 8 (23.5) 0 (0) Stomach (fundus) - 15 (41.64) 12 (35.3) 5 (14.7) 2 (5.9) Stomach (body) - 13 (38.2) 12 (35.3) 9 (26.5) 0 (0) Stomach (antrum) - 13 (38.2) 18 (52.9) 3 (8.8) 0 (0) Table 3: Screening performance characteristics of CT scan in detection of upper gastrointestinal tract injuries following caustic ingestion Characteristics Esophagus (95% CI) Stomach (95% CI) True positive 26 (76.47) 26 (76.47) True negative 4 (11.76) 2 (5.88) False positive 3 (8.82) 3 (8.82) False negative 1 (2.94) 3 (8.82) Sensitivity 96.29(79.11- 99.80) 89.65 (71.50 - 97.28) Specificity 57.14 (20.23 - 88.19) 40.00 (7.25 - 82.95) Positive Predictive Value 89.65 (71.50- 97.28) 89.65 (71.50 - 97.10) Negative Predictive Value 80.00 (29.87 - 98.94) 40.00 (7.25 - 82.95) Positive Likelihood Ratio 8.66 (2.94 - 25.43) 8.66 (2.94 - 25.43) Negative Likelihood Ratio 0.25 (0.03 - 1.59) 1.50 (0.41 - 5.52) CI: confidence interval. 7. Appendix 7.1. Acknowledgements The authors would like to acknowledge the Clinical Research Development Center (CRDC) of Loghman Hakim Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran, for consultation. 7.2. Author’s contribution All authors passed four criteria for authorship contribution based on recommendations of the International Committee of Medical Journal Editors. 7.3. Funding/Support None. 7.4. Conflict of interest None. References 1. Quingking CG, Dioquino C, Pascual J. Predictive Factors of Gastrointestinal Caustic Injury According to Clinical and Endoscopic Findings. Asia Pacific Journal of Medical Toxicology. 2013;2(1):19-22. 2. Alipour Faz A, Arsan F, Peyvandi M, Oroei M, Shafagh O, Yousefi M, et al. Epidemiologic Features and Outcomes of Caustic Ingestions; a 10-Year Cross-Sectional Study. Emergency. 2017;5(1): e56. 3. Cheng H-T, Cheng C-L, Lin C-H, Tang J-H, Chu Y-Y, Liu N-J, et al. Caustic ingestion in adults: the role of en- doscopic classification in predicting outcome. BMC gas- troenterology. 2008;8(1):31. 4. Cello JP, Fogel RP, Boland CR. Liquid caustic inges- tion spectrum of injury. Archives of internal medicine. 1980;140(4):501-4. 5. Mittal BR, Bhoil A, Kashyap R, Bhattacharya A, Dutta U, Kochhar R. 99mTc-Pertechnetate Scintigraphy and En- doscopy in Assessment of Caustic-Induced Gastric Mu- cosal Injury. Clinical nuclear medicine. 2013;38(3):e146- e7. 6. Bonnici K, Wood D, Dargan P. Should computerised to- mography replace endoscopy in the evaluation of symp- tomatic ingestion of corrosive substances? Clinical Toxi- cology. 2014;52(9):911-25. 7. Tewfik T, Schloss M. Ingestion of lye and other corrosive agents–a study of 86 infant and child cases. The Journal of otolaryngology. 1980;9(1):72-7. 8. Azadegan M, Ranjbar H. Clinical signs and endoscopic findings of the upper gastro intestinal tract in patients This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 5 Emergency. 2017; 5 (1): e61 admitted at Toxication Department of Loghman Hakim Hospital in Tehran with the history of caustic sub- stances ingestion (acid or alkaline) in 2007. Pejouhan- deh. 2009;14(3). 9. Celik B, Nadir A, Sahin E, Kaptanoglu M. Is esophagoscopy necessary for corrosive ingestion in adults? Diseases of the Esophagus. 2009;22(8):638-41. 10. Ananthakrishnan N, Parthasarathy G, Kate V. Acute cor- rosive injuries of the stomach: a single unit experience of thirty years. ISRN gastroenterology. 2010;2011. 11. Keh SM, Onyiioc N, McManus K, McGuigan J. Corro- sive injury to upper gastrointestinal tract: Still a ma- jor surgical dilemma. World journal of gastroenterology. 2006;12(32):5223. 12. Lurie Y, Slotky M, Fischer D, Shreter R, Bentur Y. The role of chest and abdominal computed tomography in as- sessing the severity of acute corrosive ingestion. Clinical Toxicology. 2013;51(9):834-7. 13. Contini S, Scarpignato C. Caustic injury of the upper gas- trointestinal tract: a comprehensive review. World J Gas- troenterol. 2013;19(25):3918-30. 14. Zargar SA, Kochhar R, Mehta S, Mehta SK. The role of fiberoptic endoscopy in the management of corrosive in- gestion and modified endoscopic classification of burns. Gastrointestinal endoscopy. 1991;37(2):165-9. 15. Munoz SR, Bangdiwala SI. Interpretation of Kappa and B statistics measures of agreement. Journal of Applied Statistics. 1997;24(1):105-12. 16. Palao R, Monge I, Ruiz M, Barret J. Chemical burns: pathophysiology and treatment. Burns. 2010;36(3):295- 304. 17. Bhoil A, Mittal B, Kashyap R, Bhattacharya A, Kochhar R. Meckel’s scan for triage in caustic injury patients: Alternative to endoscopy. Journal of Nuclear Medicine. 2012;53(supplement 1):2128-. 18. Ramasamy K, Gumaste VV. Corrosive ingestion in adults. Journal of clinical gastroenterology. 2003;37(2):119-24. 19. Ryu HH, Jeung KW, Lee BK, Uhm JH, Park YH, Shin MH, et al. Caustic injury: can CT grading system enable prediction of esophageal stricture? Clinical Toxicology. 2010;48(2):137-42. 20. Chirica M, Bonavina L, Kelly MD, Sarfati E, Cattan P. Caustic ingestion. Lancet (London, England). 2016. 21. Billiar T, Andersen D, Hunter J, Brunicardi F, Dunn D, Pol- lock RE. Schwartz’s principles of surgery: McGraw-Hill Professional; 2004. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com Introduction Methods Results Discussion Limitation Conclusion Appendix References