A Comprehensive Review of IVC Filters in a Community Tertiary Center

Introduction Retrievable inferior vena cava (IVC) filters are frequently used to prevent pulmonary embolus (PE). Indwelling filters can be associated with complications; timely removal is recommended. In our institution, activated trauma (AT) patients are enrolled in a prospective protocol and tracked by the trauma team, who refer for retrieval as soon as feasible. Interventional radiology maintains a database for traumatic injury (TI) and medical (M) patients and calls them for retrieval at 90 days. To this end, the primary goal of this study was to assess different retrieval patterns between Level 1 and 2 trauma patients, traumatic injury patients not triggering trauma response, and those with medical indications for IVC filters. An additional analysis to examine rates of recurrent thrombosis stratified for filter retention and anticoagulation status was performed.

Methods

A retrospective chart review was performed for all patients who had an IVC filter placed between January 2005 and March 2015 at Gundersen Medical Center in La Crosse, WI. Data were collected regarding patients' demographics, risk factors for thrombosis, indications for and details of filter type and placement, post-placement anticoagulation, filter retrieval, and complications arising from filter placement, retention, and retrieval. Patients were divided into AT, TI, and M groups. AT consisted of Level 1 or 2 traumas triggering management by trauma service, TI patients had lesser trauma not triggering trauma team referral, and M patients had filters for medical indications. Those who were lost to follow up or died within 30 days were excluded from all analysis except demographics. Those with permanent filters were omitted from retrieval analysis.

Results

Of the 633 charts reviewed, 4 were excluded for lack of data, 41 were lost to follow-up and 68 died within 30 days. Table 1 summarizes patient category characteristics. Activated trauma patients were younger and had their filter retrieved more often (p<0.0001) than medical or TI patients. The medical population were older, had more risk factors for recurrent thrombosis (median 2 risk factors, range 0-8) and more likely to remain on chronic anticoagulation (27.9%). Those receiving anticoagulation after filter placement had greater rates of recurrent thrombosis (p<0.0001), but more risk factors (p<0.0001)for thrombosis compared to those receiving no anticoagulation. No significant differences were observed in thrombosis-free survival with regard to filter retention. However, rates of DVT after filter placement were 8% higher if the filter was retained rather than removed across all 3 groups(p=0.0011, McNemar's Chi-square).

Discussion

Higher rates of DVT in patients with a filter in place are in line with current research associating IVC filters with DVT formation. We observed that patients on lifelong anticoagulation had a higher rate of recurrent thrombosis but increased risk factors for thrombosis, thus identifying a higher risk group for thrombotic events.

AT patients in a protocol setting had higher IVC filter retrieval rates compared to the M and IT patients with routine efforts at retrieval. The large gap between M/TI and AT filter retrieval is associated with but not fully explained by the increased risk factors and age of the M and IT groups. Protocolized retrieval of IVC filters increases retrieval rates. Further research is warranted in assessing the impact of protocolized retrieval in the medical and traumatic injury populations.