Meta-analysis of hemodynamic optimization in high-risk patients

Objective: The aim of this evidence-based report was to review pertinent randomized controlled studies that describe hemodynamic goals in acute, critically ill patients and to evaluate outcome of resuscitation therapy in association with physiologic, clinical, and therapeutic influences. Methods: MEDLINE was the source of randomized controlled studies written in English. The inclusion criteria were acutely ill, high-risk elective surgery, trauma, and septic patients. The goals of therapy were to resuscitate to either normal or supranormal values; the latter were described as a cardiac index of >4.5 L.min(-1).m(-2), pulmonary,artery occlusion pressure of <18 mm Hg, oxygen delivery of >600 mL.min(-1).m(-2), and oxygen consumption of >170 mL.min(-1).m(-2). The outcome criterion was survival or death. We found 21 randomized clinical trials described in 20 articles. The studies were divided into groups based on the time that goals were implemented (i.e., "early," 8 to 12 hrs postoperatively or before organ failure, vs. "late," or after onset of organ failure) and the severity of illness, determined by the control group mortality as >20% (12 studies) or <15% (nine studies). Results: In severely ill patients (control mortalities group >20%), six studies had a 23% mortality difference (p < .05) between the control and protocol groups with early optimization, but seven studies optimized after the development of organ failure did not have significantly improved mortality. Moreover, outcome was not significantly improved in less severely ill patients (control mortalities group <15%) and normal values as goals or when therapy did not improve oxygen delivery. Conclusion: Review of 21 randomized controlled trials with various approaches to treatment revealed statistically significant mortality reductions, with hemodynamic optimization, when patients with acute critical illness were treated early to achieve optimal goals before the development of organ failure, when there were control group mortalities of >20% and when therapy produced differences in oxygen delivery between the control and protocol groups.