Purpose: To assess interobserver and intraobserver agreement on presence of rupture and determining suitability for endovascular repair (EVAR) on CT angiography (CTA) of patients with a suspected ruptured abdominal aortic aneurysm (RAAA).
Methods: For the Amsterdam Acute Aneurysm study, a randomised multicenter trial (ISRCTN66212637), we register all patients with suspected RAAA in the Amsterdam region. For the current analysis 51 consecutive patients were included from this prospective database. Pre operative CT scans were assessed twice with a six-week interval by three vascular surgeons and two interventional radiologists. Variables recorded were presence of rupture, diameter and length of the infrarenal aortic neck, diameters of both iliac arteries and final judgement on anatomical suitability for EVAR. Kappa values for dichotomous outcomes were calculated as a measure of agreement above chance. Continuous outcomes were investigated by calculating the intraclass correlation coefficient (ICC) and by Bland-Altman plots.
Results: For presence of rupture group kappa was 0.59 (CI: 0.42-0.77). The group kappa of suitability for EVAR was 0.38 (CI: 0.24-0.51). The ICC for diameter and length of the infrarenal aortic neck and diameters of left and right iliac arteries were 0.40 (CI: 0.26-0.56), 0.47 (CI: 0.32-0.62), 0.61 (CI: 0.48-0.74) and 0.35 (CI: 0.21-0.50) respectively. The Bland-Altman plots confirmed the large variation among observers. Intraobserver kappa ranged from 0.57-0.78 for presence of rupture and 0.40-0.80 for suitability for EVAR.
Conclusion: Moderate interobserver agreement was found for presence of rupture and fair agreement for suitability for EVAR. Intraobserver agreement ranged from moderate to almost perfect. Based on this data, optimization of the protocol is mandatory to identify uniform measuring techniques of well defined anatomical criteria for endovascular repair of ruptured aneurysms.