Inferior Vena Cava Filter Retrieval Rates Associated With Passive and Active Surveillance Strategies Adopted by Implanting Physicians.

Importance: Inferior vena cava filters are commonly implanted and infrequently retrieved. Nonretrieval contributes to significant morbidity, motivating US Food and Drug Administration and multisociety communications emphasizing the need for improved device surveillance. Current guidelines suggest that implanting physicians and referring physicians should be responsible for device follow-up, but it is not known whether shared responsibility contributes to lower retrieval. Objective: To determine if primary responsibility for follow-up care assumed by the implanting physician team is associated with increased device retrieval. Design, Setting, and Participants: This retrospective cohort study examined a prospectively collected registry of patients with inferior vena cava filters implanted from June 2011 to September 2019. Medical record review and data analysis was completed in 2021. The study included 699 patients who underwent implantation of retrievable inferior vena cava filters at an academic quaternary care center. Exposures: Prior to 2016, implanting physicians had a passive surveillance strategy whereby letters highlighting indications for and the need for timely retrieval were mailed to patients and ordering clinicians. Starting in 2016, implanting physicians assumed active responsibility for surveillance, whereby candidacy for device retrieval was assessed periodically via phone calls and retrieval scheduled when appropriate. Main Outcomes and Measures: The main outcome was the odds of inferior vena cava filter nonretrieval. Within regression modeling of the association between the surveillance method and nonretrieval, additional covariates of patient demographics, concomitant malignant neoplasm, and presence of thromboembolic disease were included. Results: Of the 699 patients who received retrievable filter implants, 386 (55.2%) were followed up with passive surveillance, 313 (44.8%) with active surveillance, 346 (49.5%) were female, 100 (14.3%) were Black individuals, and 502 (71.8%) were White individuals. The mean (SD) age at filter implantation was 57.1 (16.0) years. Mean (SD) yearly filter retrieval increased following the adoption of active surveillance, from 190 of 386 (48.7%) to 192 of 313 (61.3%) (P < .001). Fewer filters were deemed permanent in the active group vs passive group (5 of 313 [1.6%] vs 47 of 386 [12.2%]; P < .001). Age at the time of implantation (OR, 1.02; 95% CI, 1.01-1.03), concomitant malignant neoplasm (OR, 2.18; 95% CI, 1.47-3.24), and passive contact method (OR, 1.70; 95% CI, 1.18-2.47) were associated with increased odds of filter nonretrieval. Conclusions and Relevance: The findings of this cohort study suggest that active surveillance by implanting physicians is associated with improved inferior vena cava filter retrieval. These findings support encouraging physicians who implant the filter to take primary responsibility for tracking and retrieval.