Single Center Experience in the Management of Cardiogenic Shock Using Temporary Mechanical Circulatory Support Devices - Impella CP, Impella 5.0, IABP, and ECMO

Abstract

Purpose Mechanical Circulatory Support (MCS) devices are generally needed in the management of cardiogenic shock (CS). The goal of this study was to assess the 30 day mortality among patients supported by 3 types of MCS devices: IABP, ECMO and Impella. Methods Between May 1, 2017 and May 31, 2018, 111 patients were treated with MCS for CS at 2 sites in Milwaukee, an academic institution and a community hospital. Preexisting conditions and indicators of CS severity included diabetes, cerebrovascular disease, renal insufficiency, dialysis, heart failure, prior MI, prior PCI, transfer from an outside hospital, and need for inotropes or vasopressors. Multivariable logistic regression was used to model mortality with the device types as predictors. Adjustment for significant predictors based on the univariate tables was performed. Patients who received multiple devices were assessed independently. Results From these 111 patients, 28% received IABP, 33% received ECMO, and 23.4% received Impella 5.0. Overall there was a 66% survival to 30 days. There was a trend in increased severity of CS in patients who received Impella 5.0, but not significant. Dialysis was a strong predictor of mortality across all device types, 33% of patients with dialysis survived to 30 days, compared to 83% of patients who did not require dialysis (p-value = Conclusion ECMO had the highest mortality (75% with ECMO, 24.4% without ECMO, p-value = 0.001), but when controlled for disease severity, there was no significance. Heart rate and dialysis were the strongest indicators of mortality across all device types. After adjusting for heart rate and dialysis, the odds of 30-day mortality comparing patients who received vs. did not receive ECMO were not significantly different (aOR 1.6, p = 0.77). There was a trend which showed Impella to have best in-hospital survival after controlling for severity of disease, but this was not significant.