The Impact of Implementation of a Commercial Inferior Vena Cava Filter Database Program on Filter Retrieval versus Physician Tracking over a 9-Year Period: A Retrospective, Observational Study

Abstract

PurposeTo compare the impact of a commercial tracking database on inferior vena cava filter retrievals with that of physician tracking and no tracking. Materials and MethodsFrom January 2013 to December 2021, 532 filters were placed at a single institution and followed in 3 phases: (a) Phase 1, pretracking (January 1, 2013, to February 28, 2015); (b) Phase 2, commercial database tracking (March 1, 2015, to June 30, 2019); and (c) Phase 3, commercial database tracking with separate tracking by an interventional radiologist (July 1, 2019, to December 31, 2021). Patients excluded from the commercial database due to human error served as a control group. Outcomes of commercial database entry, 2-year filter retrieval rates, dwell times, and factors contributing to retrieval candidacy were collected. ResultsTwo-year retrieval rates in Phases 1, 2 and 3 were 20%, 31%, and 46%, respectively (Phase 1 vs 2, P = .04; Phase 2 vs 3, P = .009). Median dwell times across Phases 1, 2, and 3 were 168 days (4–1,313 days), 140 days (3–1,988 days), and 188 days (13–734 days) (P = .33), respectively. There was no difference in retrieval rates (P = .86) and dwell times (P = .50) between patients enrolled in the database group and those enrolled in the control group. Across all phases, 48% of patients enrolled in the database were not successfully contacted, and only 6% were categorized as “likely to consult” filter retrieval. During Phase 3, 100% of patients achieved a retrieval disposition. ConclusionsA commercial tracking database had low success rates of contacting patients and did not increase filter retrieval rates relative to those in the control group; however, physician tracking increased retrieval rates.