Martha Terrazas MD, Presenter: Nothing to Disclose

Ane Lauren McCullough MD, Abstract Co-Author: Nothing to Disclose

William Moreau Thompson MD, Abstract Co-Author: Nothing to Disclose

Jess Schwartz MD, Abstract Co-Author: Nothing to Disclose

Loren Howard Ketai MD, Abstract Co-Author: Nothing to Disclose

Determine the relative accuracy of Fluoroscopic Esophagography (FE) and CT in the detection of perforation of intrathoracic viscus (esophagus or neo-esophagus).

A  university hospital Radiology Information System was searched for patients who had undergone both FE and CT within three days. Those patients who had received both exams to evaluate suspected perforation of an intrathoracic viscus were included for study. FEs and CTs were evaluated by independent readers, each blinded to the results of the other exam and to the clinical outcome.FE and CT were scored separately using a 1-5 Likert scale, a score>3 designated positive. A composite score (CS) was also created, designating cases as positive if Likert score was> 3 on FE or CT.Cases were considered true positive if 1) free perforation was surgically repaired 2) contained perforation was confirmed by evolution on subsequent diagnostic imaging or 3) by post image evaluation consensus of both readers and an attending thoracic surgeon. Accuracy of FE, CT and CS were compared using McNemar’s test

Seventy-nine patients met the inclusion criteria, 19 of whom had free (12) or contained (7) viscus perforations. Perforations were postsurgical (6),or related to endoscopy (5), penetrating trauma (4), Boerhaave’s(3) or blunt trauma (1).CT was 69% sensitive and 88% specific, with 5 of the 6 false negatives (FNs) occurring in CTs performed without positive GI contrast.Two FNs involved penetrating trauma and two were related to endoscopy. FE was 42% sensitive and 98% specific for perforation, 5 of the FNs occurring in postsurgical perforations. Sensitivity of the composite score was 95% and specificity was 88%. Both CT and CS were significantly better than FE alone (P <.05).

CT is superior to FE in the detection of perforated intrathoracic viscus , particularly in the post-surgical setting but is not sufficiently sensitive to stand alone in all settings. Diagnostic accuracy may be improved by combining CT with FE, particularly in the diagnosis of perforations caused by penetrating or endoscopic trauma or if CT is performed without administration of positive GI contrast

In a large cohort of patients undergoing evaluation for perforated intrathoracic viscus the relative efficacy of FE and CT was dependent upon the clinical setting. Results suggest that combination of both tests is often warranted.

Terrazas, M, McCullough, A, Thompson, W, Schwartz, J, Ketai, L, Perforated Intrathoracic Viscus: Are Two Tests Better than One?.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14012402.html