Incidence of Increases in Pump Power Utilization and Associated Clinical Outcomes in Patients Who Have Undergone Heartmate II Implantation

Purpose The incidence and clinical significance of increases in pump power utilization (power spikes) in patients with LVADs has not been published. This project’s aim is to identify how often power spikes occur, when they occur relative to device implantation, and whether they precede clinical findings suggestive of pump thrombosis. Methods and Materials All inpatient and outpatient medical logs were reviewed for 205 patients who underwent Heartmate II implantation at the Cleveland Clinic from 2008-2012. Power spikes were defined as an increase in pump power of 2W over baseline or pump power greater than 10W. Results Of the 205 patients, 61 (29.8%) had at least one power spike; 15 (7.4%) patients had just a single power spike and 46 (22.4%) had two or more. 41 patients had spikes in the first post-operative year, 14 patients in the second, 7 in the third, and one in the fourth; two patients had spikes in multiple years. Of the patients with power spikes in post-operative year one, 14 (34.1%) had them in the first twenty days after implantation and 26 (63.4%) had them in the first 100 days. 13 patients (6.3%) had hemolysis, 26 patients (12.7%) had TIA/embolic stroke, and 4 patients (2%) had peripheral thromboembolism (PTE). 3 patients (23%) with hemolysis had a power spike preceding the event, 4 patients (15%) with TIA/embolic stroke had a power spike before the event, and 1 patient (25%) with PTE had a preceding power spike. For those patients with power spikes preceding the clinical event, the mean and median amount of time that the power spikes occurred prior to the event was 79 and 20 days for hemolysis, 286 and 321 days for TIA/embolic stroke, and 579 and 579 days for PTE. Conclusions Power spikes are a common occurrence in patients with Heartmate II LVADs. They occur frequently in the first year after implantation but their initial presentation can be more than three years later. They do not frequently precede clinical events suggestive of pump thrombosis and their proximity to the clinical event varies.