Diagnostic accuracy of computed tomography in patients with suspected abdominal aortic aneurysm rupture.

OBJECTIVES

This study was planned to evaluate the accuracy of computed tomography angiography (CTA) for suspected rupture of abdominal aortic aneurysm (AAA).

DESIGN

Retrospective, observational study.

PATIENTS: A total of 97 patients who underwent open aneurysm repair for suspected rupture of AAA.

METHODS

The accuracy of preoperative and post hoc evaluation of CTA scans was evaluated by using the intra-operative findings as reference.

RESULTS

At surgery, 58 patients were found to have a ruptured aneurysm. Interpretation of CTA findings at admission resulted in one false-negative and two false-positive diagnoses of aneurysm rupture as observed at surgery (sensitivity 98.3%, specificity 94.9%, positive predictive value 96.6%, negative predictive value 97.4%). Post hoc, blinded review of CTA findings resulted in one false-negative and three false-positive diagnoses of aneurysm rupture (sensitivity 98.3%, specificity 92.3%, positive predictive value 95.0%, negative predictive value 97.3%). Agreement between initial and post hoc assessment of CTA-findings was statistically significant (kappa 0.978, p < 0.0001).

CONCLUSIONS

CTA is highly accurate in the diagnosis of suspected rupture of AAA. The risk of false-positive diagnosis is very low and it is not likely to affect the correct analysis of the results of comparative studies on endovascular versus open repair of ruptured AAA.