Effect of access site choice on inferior vena cava filter angulation and outcomes

Background: In the present study, we compared the outcomes of inferior vena cava (IVC) filter placement between the femoral vein (FV) and internal jugular (IJ) vein access sites. Methods: We performed a retrospective study using the Vascular Quality Initiative database to assess patients who had undergone IVC filter placement from 2013 to 2019. The patients were placed into two groups according to the access site location: FV and IJ vein. The FV group included patients with access via the right and left FVs and other leg veins, and the IJ group included patients with access via the right or left IJ vein. The primary outcome was the rate of filter angulation. The secondary outcomes included access site complications such as deep vein thrombosis, hematoma, and bleeding requiring transfusion. Results: Of 13,221 patients, 8214 (63%) had undergone IVC filter placement via FV access and 4789 (37%) via IJ access. The remaining 218 patients had had an unknown access site or were excluded. Within the IJ group, 4696 (98.0%) had undergone access via the right IJ and 93 (2%) via the left IJ. Within the FV (common femoral, femoral, or other infrainguinal veins) group, 7007 (85.3%) had undergone access via the right FV and 1207 (14.6%) via the left FV. The mean patient age was 63 +/- 15.9 years, the mean body mass index was 30.9 +/- 9.60 kg/m(2), and 6788 of the patients were men (52.0%). The most common indication for filter placement was a contraindication to anticoagulation because of a recent or active bleeding episode (30%), followed by planned surgery (22%), new deep vein thrombosis/pulmonary embolism (7%), fall risk (5%), and trauma (4%). Infrarenal filters had been placed in 97.9% of the patients. Univariate analysis identified body mass index and suprarenal placement as independent risk factors for angulation. The final multivariate analysis showed a significant increase in angulation (0.9% vs 0.34%; odds ratio, 1.46; 95% confidence interval, 1.02-2.11; P = .04) and increased access site complications (0.25% vs 0.07%; odds ratio, 2.068; 95% confidence interval, 1.01-4.23; P = .048) in the FV access group. No significant correlation between the access site and retrieval rate was found ( P = .9270). Conclusions: Placement of IVC filters via IJ access showed a lower rate of filter angulation in the IVC and fewer access site complications compared with FV access.