Survey of Anticoagulation Practices with the Impella Percutaneous Ventricular Assist Device at High-Volume Centers

Abstract

Objectives. To characterize anticoagulation practices with the Impella percutaneous ventricular assist device (pVAD). Background. Managing anticoagulation in patients being supported by the Impella pVAD is made challenging by several unique features of the device. These include the release of a dextrose-based purge solution containing unfractionated heparin (UFH), the need to concurrently administer systemic anticoagulation with intravenous UFH, and the lack of an alternative strategy in patients with contraindications to UFH. Methods. To characterize anticoagulation practices with the Impella pVAD, we conducted a survey of centers in the United States performing a high volume of Impella cases, which we defined as > 1 per month. Centers were contacted via email or phone and individuals who agreed to participate were provided with a link to complete the survey online. The primary measures of interest were variations in practice across centers and variations from the manufacturer's recommendations. Results. Practices varied considerably among respondents (65 of 182 centers, or 35.7%) and often diverged from manufacturer recommendations. Approximately half of centers (52.4%) reported using a UFH concentration of 50 units/mL in the purge solution, whereas most of the remaining centers (41.3%) reported using lower concentrations. Strategies for the initiation and adjustment of systemic therapy also varied, as did practices for routinely monitoring for hemolysis. Nearly one-fifth of centers (16.7%) had not developed an alternative strategy for the purge solution in patients with contraindications to UFH. Most centers (58.4%) reported using argatroban or bivalirudin in this scenario, a strategy that diverges from the manufacturer's recommendations. Conclusions. Given these findings, studies to determine a systematic approach to anticoagulation with the Impella device are warranted. Copyright 2019 Brent N. Reed et al.