Diagnostic Value of Abdominal Free Air Detection on a Plain Chest Radiograph in the Early Postoperative Period: a Prospective Study in 648 Consecutive Patients Who Have Undergone Abdominal Surgery

Abstract

BackgroundTo the best of our knowledge, this is the first study to evaluate the predictive value of free air (on a plain radiograph) for bowel perforation in a large prospective cohort of surgical patients. MethodsAll consecutive patients undergoing abdominal surgery between January 2011 and June 2012 were screened for this study. We performed an upright chest radiograph on the second and third postoperative day. Thereafter, additional radiographic evaluations were performed every 2 days until the disappearance of abdominal free air. ResultsOf the 648 subjects enrolled in our study, free abdominal air was found in 65 subjects on the first radiographic evaluation (2 days after surgery), 51 on the second (3 days after surgery), three on the third (5 days after surgery), and none on the fourth (7 days after surgery). The presence of free abdominal air was associated with an increased risk of gastrointestinal perforation. The presence of free air was associated with a hazard ratio (HR) of 21.54 (95% CI 9.66–48.01, p<0.001) and a HR of 23.87 (95% CI 10.68–53.34, p<0.001) at 2 and 3 days after surgery, respectively. Sensitivity, specificity, positive predictive value, and negative predictive value were 70, 93, 33, and 98%, respectively, at 2 days after surgery, and similar results were confirmed at 3 days after surgery. ConclusionWe believe that the presence of free air at 3 days after surgery should not be considered a common finding. Here, we demonstrate that the detection of free air has a remarkable predictive value for gastrointestinal perforation, which has been overestimated in previous experience.