Back to 2016 Annual Symposium ePosters

OBJECTIVES:
Follow up computed tomography angiography (CTA) scans add considerable post implantation costs to endovascular aneurysm repairs (EVAR) of abdominal aortic aneurysms (AAA). By building a risk model, we hope to identify patients at low risk for postoperative complications in order to minimize unnecessary CT scans.
METHODS:
557 consecutive patients were reviewed. Probit models were created for four outcomes: aneurysm sac enlargement, reintervention, type I/III endoleaks and type II endoleaks, using preoperative aneurysm morphology, patient demographics and operative details as covariates. Patients with an abnormal 30 day post operative CT scan were excluded. Time points chosen for analysis were 1-year, 2-year, 3-years and 10-year post EVAR. A combined model including three outcomes was also created.
RESULTS:
Individual models that showed most significance were aneurysm sac enlargement at 1-year post EVAR (n=411; Sensitivity=1; Specificity=0.47; Accuracy 0.48) and reintervention at 2-year post EVAR (n=494; Sensitivity=1; Specificity=0.42; Accuracy=0.45). (Table 1) Notably, our models reported that increasing iliac artery diameter and increasing aortic neck angle increased the risk for a reintervention, while increasing iliac artery tortuosity increased the risk for a type I/III endoleak. Similarly, almost all statistically significant covariates were qualitatively supported by prior literature. Our final combined model would prevent the need for post-EVAR CTA in 59 patients (16%) at 1 year, 26 patients (7%) at 2 years and 6 patients (2%) at 3 years. (Table 2)
CONCLUSIONS:
Our model is a robust predictor of patients at low risk for post AAA EVAR complications. With additional validation and refinement, it could be applied to practices to cut down on the overall need for post-implantation CTA. Reducing the number of CT scans will reduce post-implantation cost and radiation exposure.

Table 1: Individual Model Analysis

Model	1 year (Aneurysm Sac Enlargement)	2 year (Aneurysm Sac Enlargement)	1 year (Reintervention)	2 year (Reintervention)	1 year (Type I/III Endoleak)	2 year (Type I/III Endoleak)	1 year (Type II Endoleak)	2 year (Type II Endoleak)
n	411	411	494	494	505	505	470	470
True Positive	11	16	18	26	4	5	19	24
False Positive	212	278	288	273	123	194	415	412
True Negative	188	117	188	195	378	306	36	34
False Negative	0	0	0	0	0	0	0	0
Sensivity	1	1	1	1	1	1	1	1
Specificity	0.47	0.30	0.40	0.42	0.75	0.61	0.08	0.08
Accuracy	0.48	0.32	0.42	0.45	0.76	0.62	0.12	0.12

Table 2: Combined Model Analysis

Combined Model (n=377)	1 year (%)	2 year (%)	3 year (%)
True Positive	24	35	50
False Positive	294	316	321
True Negative	59	26	6
False Negative	0	0	0
Sensitivity	1	1	1
Specificity	0.16	0.07	0.02
Accuracy	0.22	0.16	0.15