Abstract 4360153: Adverse Events and Outcomes Among Patients with Temporary Mechanical Circulatory Support Devices Placed at Referring Versus Hub Cardiogenic Shock Centers

Introduction: Temporary mechanical support (tMCS) devices (intra-aortic balloon pump [IABP], Impella CP, Impella 5.5) are commonly used in cardiogenic shock (CS) and are often placed at regional centers (RC) to stabilize patients for transfer to a CS hub center (HC). Aims: To assess whether initial tMCS device placement for CS at an RC before transfer to HC is associated with more device-related adverse events (DRAEs) and worse outcomes than with initial device placement at a HC. Methods: All patients admitted to a single center from 8/2021 to 8/2023 with CS requiring tMCS were identified using CPT codes. Baseline characteristics, CS severity, DRAEs, and outcomes were collected. Outcomes included in-hospital mortality and “unfavorable outcome” (death prior to heart transplant, durable LVAD implant, or discharge). Patients were stratified by location of initial tMCS device placement (RC vs HC). Characteristics, DRAEs, and outcomes were compared. Multivariable logistic regression was performed to account for baseline differences. Sensitivity analyses were performed to assess consistency of results by device exposure. Results: 268 patients were included, 214 (79.9%) with initial device placed at HC and 54 (20.1%) at RC. The initial device was IABP in 54.1%, Impella CP in 28.0%, and Impella 5.5 in 17.9%. Median age was 60 (44, 67) years, 59.3% were Black, and 71.6% male (similar between groups). Compared to HC implanted patients, RC patients more commonly had non-ischemic CM (63% vs 37%, p=0.02), de novo HF-CS (41% vs 14%, p=0.02), and cardiac arrest (32% vs 11%; p<0.01) but less commonly had prior HFrEF (51% vs 69%, p=0.02) and kidney disease (9% vs 25%, p=0.01). RC implanted patients had more advanced CS (Stage E CS in 30% vs 8%; p<0.01). DRAEs were more common in RC than HC implanted patients (1.19 vs 0.46 events per patient-week on tMCS support; p<0.01)(Fig 1). RC implanted patients had higher unadjusted odds of in-hospital mortality and unfavorable outcome (Fig 2). This finding was similar after adjustment for baseline differences and CS etiology but no longer statistically significant after adjusting for CS severity and cardiac arrest. These findings were consistent in the sensitivity analyses among patients with any IABP and any Impella exposure. Conclusions: Patients with tMCS devices placed for CS at RC have more DRAEs and worse outcomes than those with initial device placed at a HC. The higher mortality in RC implanted patients may be due to greater CS severity.