Costs and complications of endovascular inferior vena cava filter retrieval.

Objective: Advanced endovascular techniques are frequently used for challenging inferior vena cava (IVC) filter retrieval. However, the costs of IVC filter retrieval have not been studied. This study compares IVC filter retrieval techniques and estimates procedural costs.

Methods: Consecutive IVC filter retrievals performed at a tertiary center between 2009 and 2014 were retrospectively reviewed. Procedures were classified as standard retrieval (SR) if they required only a vascular sheath and a snare device and as advanced endovascular retrieval (AER) if additional endovascular techniques were used for retrieval. Cost data were based on hospital bills for the procedures. Patients' characteristics, filter dwell time, retrieval procedure details, complications, and costs were compared between the groups. All statistical comparisons were performed using SAS 9.3 software.

Results: There were 191 IVC filter retrievals (SR, 157; AER, 34) in 183 patients (mean age, 55 years; 51% male). Fifteen filters (7.9%) were placed at an outside hospital. The indications for placement were mostly therapeutic (76% vs 24% for prophylaxis). All IVC filters were retrievable, with Bard Eclipse (Bard Peripheral Vascular, Tempe, Ariz; 34%) and Cook Günther Tulip (Cook Medical, Bloomington, Ind; 24%) the most common. Venous ultrasound examination of the lower extremities of 133 patients (70%) was performed before retrieval, whereas only 5 patients (2.6%) received a computed tomography scan of the abdomen. There was no difference in the mean filter dwell time in the two groups (SR, 147.9 ± 146.1 days; AER, 161.4 ± 91.3 days; P = .49). AERs were more likely to have had prior attempts at retrieval (23.5%) compared with SRs (1.9%; P < .001). The most common AER techniques used were the wire loop and snare sling (47.1%) and the stiff wire displacement (44.1%). Bronchoscopy forceps was used in four cases (11.8%); this was the only off-label device used. AERs were more likely to require more than one venous access site for the retrieval procedure (23.5% vs 0%; P < .001). AERs were significantly more likely to have longer fluoroscopy time (34.4 ± 18.3 vs 8.1 ± 7.9 minutes; P < .001) and longer total procedural time (102.8 ± 59.9 vs 41.1 ± 25.0 minutes; P < .001) compared with SRs. The complication rate was higher with AER (20.6%) than with SR (5.2%; P = .006). Most complications were abnormal radiologic findings that did not require additional intervention. The procedural cost of AER was significantly higher (AER, $14,565 ± $6354; SR, $7644 ± $2810; P < .001) than that of SR. This translated to an average increase in cost of $6921 ± $3544 per retrieval procedure for AER.

Conclusions: Advanced endovascular techniques provide a feasible alternative when standard IVC filter retrieval techniques do not succeed. However, these procedures come with a higher cost and higher rate of complications.