MAXIMIZING OXYGEN DELIVERY IN CRITICALLY ILL PATIENTS - A METHODOLOGIC APPRAISAL OF THE EVIDENCE

Objective: To systematically review the effect of interventions design ed to achieve supraphysiologic Values of cardiac index, oxygen deliver y (Do(2)), and oxygen consumption (Vo(2)) in critically ill patients. Data Sources: Computerized bibliographic search of published research, citation review of relevant articles, and contact with primary invest igators. Study Selection: We included all randomized clinical trials o f adult intensive care unit (ICU) patients that evaluated intervention s (fluids, inotropes, and vasoactive drugs) designed to achieve suprap hysiologic values of cardiac index, Do(2), and/or Vo(2). Independent r eview of 64 articles identified seven relevant studies of 1,016 patien ts. Data Extraction: We abstracted data on the population, interventio ns, outcomes, and methodologic quality of the studies by duplicate ind ependent review, Agreement was high (weighted kappa 0.73); differences were resolved by consensus.Data Synthesis: Targeting therapy to achie ve supraphysiologic end points in critically ill patients is associate d with a nonstatistically significant trend toward decreased mortality rates (relative risk 0.86, 95% confidence intervals 0.62 to 1.20). Fo r the two studies in which supraphysiologic goals were initiated preop eratively, the relative risk was 0.20 (95% confidence intervals 0.07 t o 0.55). This value differed significantly from the combined estimate of the remaining studies, in which the intervention was started after ICU admission (relative risk 0.98, 95% confidence intervals 0.79 to 1. 22; p < .01). However, there are several methodologic problems with th e primary studies. In no trials were caregivers or outcome assessors b linded to treatment allocation. Only three of seven trials analyzed pa tients according to the group to which they were allocated. None adequ ately controlled for cointerventions, and there was considerable cross over between groups (patients in the control group achieved the goals of the intervention group and Vice versa). Conclusions: Interventions designed to achieve supraphysiologic goals of cardiac index, Do(2), an d Vo(2) did not significantly reduce mortality rates in all critically ill patients. However, there may be a benefit in those patients in wh ich the therapy is initiated preoperatively. Methodologic limitations weaken the inferences that can be drawn from these studies and preclud e any evidence-based clinical recommendations.