Reasons for Failure to Remove Inferior Vena Cava Filters: Retrospective Study in a Tertiary Care and Trauma Centre

Inferior vena cava filter (IVCF) use has been rising exponentially since the introduction of percutaneous image guided IVCF placement and the subsequent development of optional recovery models. Optional recovery IVCFs are preferred over permanent devices as an alternative or adjunct to anticoagulation to prevent pulmonary embolism (PE) in patients with venous thromboembolism (VTE) due to their ability to be retrieved once they are no longer needed. In 2010, the US Food and Drug Administration issued a safety alert for optional recovery filters following reports of adverse events associated with their widespread use. Attempted removal rates reported in the literature are suboptimal, ranging from 15 to 70% in single centre studies. We conducted a retrospective study to determine IVCF complications, retrieval rates and barriers to removal in our institution.

Consecutive patients who had a retrievable IVCF inserted or removed by the Interventional Radiology department from January 1^st 2007 to June 30^th 2010 were identified from a prospective database. Data were extracted from chart review using standardized forms. Outcomes collected included indication for IVCF, frequency of removal, documentation of an IVCF plan (for retrieval or not), reasons for non-retrieval, and all-cause mortality. A bivariate comparison of patients who had an attempted IVCF retrieval or not was performed. Patient characteristics were compared using 2-sided t-tests for continuous variables and chi-square analysis for categorical variables. Multivariate analysis by binary logistic regression was performed for significant variables (p<0.20) on univariate analysis. All statistics were performed using SPSS software.

242 patients were identified with a median age of 60 years (range 14–93 y) of which 52.9% were male. Underlying thrombotic conditions included: acute (<3 months) VTE (76.9%), malignancy (35.1%) and trauma (21.5%). Indications for IVCF placement were acute (63.6%) or prior (8.3%) VTE with contraindication to anticoagulation, high risk of PE (14.5%) and primary prophylaxis (13.6%). A total of 37 patients died at a median of 13 d (range 1–66 d) after IVCF insertion. Retrieval was attempted in 58.1% (140/242 patients) and was successful in 89.9% (124 patients). The most common reason for failed retrieval attempt was filter thrombus (N=12). In the 33 patients with filter placement for primary prophylaxis 20 (60.6%) had a removal attempt with a success rate of 100%. Reasons for IVCF non-retrieval (102 patients) included death or limited life expectancy (N=39, 38.2%), plan to make the filter permanent (N=17, 16.7%), persistent/permanent contraindication to anticoagulation (N=14, 13.7%), persistent high risk of PE despite anticoagulation (N=2, 2%), planned indefinite anticoagulation (N=1, 0.98%), transfer to another province (N=2, 2%) and unknown/unspecified in chart (N=27, 26.5%). Predictors of attempted retrieval on multivariate analysis included documentation of filter plan (p<0.001) and age <70 years (p=0.016). In 27 patients (11.1%), removal was not attempted despite a lack of indication for IVCF to remain in situ; of these, 16 patients had no documented retrieval plan. Patients with no identifiable reason for non-retrieval were more likely to be cancer-free (p=0.010), admitted under a surgical service (p=0.009) and not have a consulting hematologist's involvement (p=0.041).

Our single-centre IVC filter recovery rate compares favorably to published reports but remains suboptimal. Failure to document a filter plan and age > 70 years were the only predictors of a lower rate of attempted IVC filter recovery. A significant proportion of patients (11.1%) who did not undergo an attempted IVC filter recovery had no identifiable reason not to do so. Improper documentation may play a significant role in failure to recover devices, particularly in patients with no contraindications to removal. Further studies are needed to identify barriers to retrieval and improve removal rates.

No relevant conflicts of interest to declare.