Noise Audit of Device Selection for Peripheral Vascular Interventions in the Vascular Quality Initiative.

Objective: To conduct a "Noise Audit" of the Vascular Quality Initiative (VQI) Peripheral Vascular Intervention (PVI) Registry to analyze variation in device selection for similar patients and anatomies.

Methods: We conducted a retrospective cohort study analysis of the VQI PVI Registry. Eight clinical-based, and seven lesion-based scenarios were developed, and the VQI was queried for all procedures matching these descriptions from December 2010 to December 2021. All patients undergoing PVI for either claudication or chronic limb threatening ischemia (CLTI) were included. Device selection was then identified for each procedure, with a treatment hierarchy to designate the primary device class along with a separate accounting of specialty balloons including drug coated and lithoplasty devices. Variability was then quantified in each mock patient scenario using the most commonly used device as the reference standard.

Results: The number of patients matching the criteria for each mock scenario ranged from 223-1,379 across eight clinical and from 6,166-26,241 in seven lesion-based scenarios. The most commonly used device in a given patient scenario was used between 32% and 72% of procedures in the clinical based scenarios, and between 34% and 67% of procedures in the lesion-based scenarios. Variation in device selection and use of specialty balloons were found to be significant across all patient scenarios (P <0.0001).

Conclusion: Noise in PVI device selection and use of drug coated balloons in the VQI was significant across 15 mock scenarios. This indicates that similar patients with similar vascular pathologies are receiving vastly different devices and confirms that future comparative effectiveness studies are required to more effectively combat this systemic source of Noise in device selection.