The Value of Sigmoidoscopy to Detect Colonic Ischaemia After Ruptured Abdominal Aortic Aneurysm Repair.

Diagnosing colonic ischaemia (CI) after ruptured abdominal aortic aneurysm (RAAA) repair is challenging. This study determined the diagnostic value of sigmoidoscopy in patients suspected of CI after RAAA repair. This was a retrospective multicentre cohort study. Patients who underwent RAAA repair in three hospitals in Amsterdam, the Netherlands, between 2004 and 2011 (AJAX cohort) were included. Sigmoidoscopies were carried out based on clinical judgment. Endoscopy results were classified as "no ischaemia," "mild CI," or "moderate to severe CI." The surgical diagnosis was classified as "transmural" or "no transmural" CI. The value of sigmoidoscopy was assessed with calculation of positive and negative predictive values (PPV, NPV) with 95% CI for transmural CI. Logistic regression analysis was used to express the association of risk factors with CI as adjusted OR. Transmural CI was diagnosed in 23 of 351 patients (6.6%). Thirteen of sixteen patients (81%) who underwent direct laparotomy for high suspicion of CI indeed had transmural CI. Forty-six patients (13%) underwent sigmoidoscopy. The prevalence of transmural CI was 22% (10/46; 95% CI 12-36%) in these patients. The PPV for transmural CI of "moderate to severe CI" on sigmoidoscopy was 73% (8/11; 95% CI 43-90%). The PPV of "mild CI" on sigmoidoscopy was 11% (2/19; 95% CI 2.9-31%). The NPV of "no ischaemia" on sigmoidoscopy was 100% (95% CI 78-100%). Cardiac comorbidity (OR 3.1, 95% CI 1.19-7.97), low first haemoglobin (OR 0.6, 95% CI 0.47-0.87), and high vasopressor administration (OR 9.4, 95% CI 1.99-44.46) were independently associated with CI. Sigmoidoscopy increases the likelihood of correctly identifying the presence or absence of transmural CI, especially in patients with a moderate clinical suspicion for CI after RAAA repair.

Informations de copyright

Copyright © 2018 European Society for Vascular Surgery. Published by Elsevier B.V. All rights reserved.