292

Abstract

Introduction: Hemodynamic parameters derived via transpulmonary thermodilution are frequently used by clinicians for goal-directed fluid management. In this context, typical pathologies include sepsis or ARDS. However, little is known about its predictive value in patients in cardiogenic shock. Methods: After IRB approval, we analyzed electronic records of patients that received a PiCCO device between 01/01/2006 and 31/12/2011 on one of our intensive care units (n=1,389). Thereof, 212 patients presented with symptoms of acute cardiac decompensation. Patients were attributed in two groups according to hospital mortality. After matching based on APACHE2 admission score on ICU (score ± 10%), a total of 164 patients were included for analysis. Statistical analysis was performed by using SPSS 20.0 software, considering a p-value below 0.05 to be significant. Results: Patients of both groups (survivors and non-survivors) did not present statistically significant differences for age, BMI and ICU admission scores (APACHE2, SAPS2, SOFA, TISS). For cardiac index, non-survivors had a lower minimal value (2.0 vs. 2.4; p=0.04) and last value measured (2.7 vs. 3.2; p=0.04). EVLWI was significantly higher in non-survivors for minimum (8.7 vs. 6.9; p<0.01), maximum (12.5 vs. 10.5; p<0.01) and last value (10.6 vs. 7.8; p<0.01). Although first measurement of both CI and ELWI were not significant, minimal and maximal EVLWI provided significant data at the earliest point of time (both p=0.02 on second day of PiCCO measurements). No statistical association was found for ITBV (intrathoracic blood volume) and mortality. Conclusions: In the present setting of cardiogenic shock, a lower CI and elevated EVLWI are associated with higher rates of death. EVLWI, especially, seems to have a predictive value for mortality.