537

Introduction: Post-cardiac arrest (CA) resuscitated patients often develop a "sepsis-like" syndrome, which may be associated with organ dysfunction. We have shown that microcirculation is impaired after CA (1), however the time course of these alterations as well as their association with mortality have not been studied yet. We evaluated the evolution of microvascular reactivity disturbances during time after CA and their association with outcome. Methods: From April 2009 to April 2012, we prospectively studied 60 successfully resuscitated CA patients. Thenar oxygen saturation (StO2) was measured using a tissue spectrometer (InSpectra 650; Hutchinson, USA), and a vaso-occlusive test was performed to evaluate StO2 reperfusion rate (SRR), reflecting microvascular reactivity. Measurements were performed at 5 different time points: within the first 4 hours after CA (T0, n=38) and 4-12 hours (T1, n=49), 12-24 hours (T2, n=58), 24-48 hours (T3, n=50), 48-72 hours (T4, n=40) thereafter. The following variables were recorded: total hemoglobin index (THI), the StO2 desaturation slope during the occlusion (dSlope, %/sec) and the StO2 upslope of the reperfusion phase following the ischemic period (ascSlope, %/sec). Muscle oxygen consumption (NIR VO2; arbitrary units) was calculated as the product of the inverse StO2 desaturation slope and the mean THI over the first minute of arterial occlusion. Analysis of variance (ANOVA) for repeated measurement with correction for missing values was performed. Results: In this cohort, median age was 61 [52-67] years and 48 (80%) patients were men, 44 (73%) had a presumed cardiac origin, 29 (60%) VF/VT as initial rhythm and 55 (91%) an out-of-hospital CA. Median time from collapse to CPR and from collapse to ROSC were 3.5 [3-5] and 15 [10-20] minutes, respectively. Median APACHE II and SOFA score were 26 [22-31] and 7 [5-9] on admission, respectively. Twenty-three (38%) patients presented cardiogenic shock on ICU admission. Overall ICU stay was 4 [2-11] days and 33/60 (55%) eventually died. The cause of death was neurological (ND) in 22 patients and related to cardiac or multiple organ failure (MOF) in 11 patients. Only THI (from 11.8 [9.5-14.6] to 9.6 [8.1-11.9], p=0.02), and dSlope (from -0.11 [-0.18 to -0.09] to -0.14 [-0.20 to -0.11], p=0.004) showed significantly changes during time. When we analysed separately survivors, non-survivors for ND and non-survivors with MOF, we observed a significant differences in THI, nirVO2 and ascSlope between survivors and non-survivors with MOF, while microvascular data between survivors and non-survivors with ND were similar. Conclusions: Alterations in peripheral microvascular reactivity progressively resolved during time and were more pronounced in those patients dying from cardiac or MOF.