Abstract 13334: Safety of Chest Compressions in Patients With a Durable Left Ventricular Assist Device

Abstract

Introduction: The safety of chest compressions (CC) during cardiopulmonary resuscitation (CPR) among pts with a durable left ventricular assist device (LVAD) is an area of controversy. Concerns regarding device malfunction or dislodgment with CC may result in significant delays or the withholding of life-sustaining intervention entirely. Current societal guidelines, based on small case reports, suggest that manual CC are probably not harmful, but offer no additional guidance. Furthermore, they acknowledge that there are no data regarding the use of machine-assisted CC. Methods: We retrospectively reviewed medical records of consecutive pts who underwent durable LVAD implantation at Duke University from 2013-2022. Charts were searched for “code”, “arrest”, “compressions”, and “rapid response team” to identify all instances of cardiac arrest (CA). Baseline pt characteristics, details of the arrest (including CC details), in-hospital care processes, and pt outcomes were abstracted. Results: Of 750 pts reviewed, 58 (mean age 64 years, 76% male, median LVAD support duration 19 months) had documented CA. Fifteen (26%) pts presented with an out-of-hospital CA. Twenty-four (41%) pts did not receive CC. The most commonly reported reason for avoiding CC was “LVAD in place - contraindicated.” The CPR duration was longer among those who received CC (median 30mins, range: 1-120 mins) than those who did not (median 20 mins, range: 2-79 mins). Two pts had support from a Lund University Cardiac Assist System (LUCAS) device. Among those who survived either manual and machine-assisted CC, 27% had a post-arrest TTE and none showed evidence of LVAD malfunction/dislodgment. Conclusions: Among pts with a durable LVAD who underwent CPR for CA we found no evidence of device dislodgment or safety concerns associated with either manual or LUCAS-assisted CPR. Future multicenter and prospective evaluation is necessary to both confirm these findings and to inform the guidelines.