Role of duplex scan in endoleak detection after endoluminal abdominal aortic aneurysm repair

Objective: to validate the role of duplex scan in endoleak detection in postoperative surveillance of endoluminal abdominal aneurysm repair (EAAR). Patients and methods: between April 1997 and March 1999, 103 patients were eligible for duplex and computed tomography (CT) scan after EAAR. Mean follow-up was 8 months (range 1-24 months). The study protocol comprised concurrent examination with colour-duplex and CT scan at 1, 6, and 12 months after EAAR, for a total of 198 concurrent examinations. All duplex scan examinations were performed by two vascular surgeons with the same machine (ATL HDI 3000). Interobserver agreement in endoleak detection (kappa = 1) and in type of endoleak (kappa = 0.7) was evaluated in 50 random duplex examinations. Endoleak detection was examined comparatively in duplex and CT scan, the latter being the gold standard. Sensitivity and specificity tests together with negative- and positive-predictive values (NPV and PPV) were calculated. Results: duplex scan was not feasible in one patient. On CT scan the endoleak rate was 4% at one month, 3% at 6 months, and 4% at one year. Overall, CT scan detected 12 endoleaks. With respect to endoleak detection, duplex scan revealed a great ability in ruling out false-negative results (sensitivity 91.7%, NPV 99.4%), but overestimated the presence of endoleak (specificity 98.4%, PPV 78.6%). Regarding type of endoleak, the ability of duplex scan to identify the source of endoleak was low (sensitivity 66.7%). Conclusions: duplex scan, if validated, appears to be a reliable means for excluding the presence of endoleak after EAAR.