Hemodilution Then Hemoconcentration after Resuscitation of Patients with Severe Sepsis or Septic Shock.

BACKGROUND The impact of fluid ri suscitation on hematologic parameters and function has been well studied in hemorrhagic shock. Similar research has not been conducted in resuscitation of septic shock. HYPOTHESIS In the absence of accompanying hemorrhage, resuscitation of patients with sepsis should be marked by hemodilution, followed by hemoconcentration during recovery. METHODS Records of patients with primary diagnoses of severe sepsis or septic shock treated in a community hospital intensive care unit (ICU) between 2009 and 2012 were extracted from an electronic d tabase for analysis. Demographic, physiologic, an laboratory values were recorded at daily intervals. RESULTS 132 patients with an average age of 70. (SD 15.1) years and Acute Physiology and Chronic Health Evaluation II (APACHE II) score of 15. (6.0) were studied. Patients spent an average of 10. (9.9) days in the ICU and 18.9 (12.0) days in hospita 19 (14.4%) did not survive hospitalization. Mean admission hematocrit was 34.8 (6.5%), and lom est hematocrit, adjusted for (average 0.2 U PRBC) transfusions, 25.3 (5.1)% (P < .001), occurred after an average of four days of treatment, and 7.2 (5.4 L of cumulative positive fluid balance. By day 10 adjusted hematocritincreased to 26.9(8.1) (P =.006' 'Ihere was a significant (P < .001) albeit loose correlation (R = .35) of cumulative positive fluid balance associated with lowest hematocrit. CONCLUSION Fluid resuscitation of patients with severe sepsis or septic shock is marked by initial reductions of hematocrit followed by increases during recovery, as fluid is mobilized.