Strategies Associated With High Inferior Vena Cava Filter Retrieval: A Positive Deviance Study.

Trivedi, Premal; Barnard, Juliana G; Gurfinkel, Dennis; Baron, Maria Puello; Lindquist, Jonathan; Bai, Harrison; Ho, P Michael · J Am Coll Radiol · 2026

other · Level V

Where this comes from

Abstract

Inferior vena cava (IVC) filter retrieval is recommended once mechanical protection against pulmonary embolism is no longer needed, ideally within 3 months of implantation. Building on prior work demonstrating low aggregate retrieval and substantial interfacility variation in the United States, this qualitative study aimed to describe barriers to timely IVC filter retrieval and surveillance processes and implementation strategies associated with exceptionally high (>95th percentile) retrieval rates. This qualitative analysis was embedded within a mixed-methods positive deviance study. After risk-adjusted 1-year IVC filter retrieval rates were estimated for US hospitals serving Medicare beneficiaries, we examined four high-performing (retrieval rates 40%-74%) and three intermediate-performing (retrieval rates 7%-40%) hospitals. Semistructured interviews with surveillance personnel were analyzed using team-based rapid qualitative analysis focused on barriers to timely retrieval and high-retrieval surveillance processes. Barriers to timely retrieval were similar across sites and included dependence on individual clinicians to recognize implanted filters, labor-intensive tracking, difficulty obtaining provider input, challenges contacting patients, and competing clinical obligations. High-retrieval hospitals converged on several implementation strategies: an empowered nonphysician champion, protocolized surveillance, presumptive rather than open-ended communication regarding retrieval eligibility with involved clinicians, and reserving additional imaging or clinic evaluation for medically complex patients. Electronic medical record dashboards facilitated workflow but did not differentiate high- from intermediate-retrieval hospitals. High IVC filter retrieval seemed less dependent on a specific surveillance model than on how programs operationalized follow-up into routine care. Scalable interventions should prioritize accountable nonphysician ownership, risk-stratified escalation, protocolized reassessment, and streamlined clinician decision making, rather than registries or dashboards alone.