LAUSR

Background In the absence of robust evidence from randomized controlled trials, clinicians treating cardiogenic shock (CS) often make decisions on short-term Mechanical Circulatory Support (s-t MCS) based on individual center or clinical team preferences. We sought to investigate whether description of device-related adverse events (AE) could help identify targets to advance the field. Methods Our institution's CS database was queried for CS patients who underwent placement of s-t MCS based on predefined criteria, between Jan 2014 - Mar 2017. Patients who needed Intra-Aortic Balloon Pump, central cannulation or had a post-cardiotomy CS were excluded from the analysis. We identified 61 who required s-t MCS with Impella, peripheral Veno-Arterial Extra-Corporeal Membrane Oxygenation (VA-ECMO) or combination of these devices as a sequela of escalating support strategy. Results Of the total patients, 23 (37.7%) were treated with Impella, 10 (16.4%) with VA-ECMO and the remaining 28 (45.9%) with combinations of devices. Groups were of similar demographics and CS etiology. Severity of CS pre-MCS revealed no differences in vital signs, number of vasoactive agents used, left ventricular ejection fraction, use of mechanical ventilation and CPR. The VA-ECMO compared with Impella and Combination groups had higher baseline lactate (11.1 vs 4.9 vs 5.3; p=0.002) and lower baseline pH (7.05 vs 7.27 vs 7.27; p=0.004) pre-MCS. The mean length of support was similar between Impella and VA-ECMO groups, however it was longer in the Combination vs Impella group (p=0.001). Major AE are shown in the Table . The mean ICU stay was 15 days and did not differ among the groups. Exchange to durable LVAD was performed in 6 patients: 3/23 (Impella) and 3/28 (Combination) group. The overall survival to hospital discharge was 43% and did not differ between the groups (p=0.2). Conclusions Despite significant advances in the technology of s-t MCS options, morbidity and mortality remain high and the device-related AE could provide practical quality improvement targets for CS programs.