EFFECT OF INTERMITTENT VS CONTINUOUS SCVO2 MONITORING ON SEPSIS BUNDLE COMPLIANCE AND MORTALITY

Abstract

PURPOSE: To determine whether intermittent measurement vs. continuous measurement of ScvO2 results in a greater success rate for achieving an ScvO2 goal of > 70% within six hours of diagnosis in patients with severe sepsis/septic shock. Numerous articles describe the importance of meeting resuscitation bundle criteria in severe sepsis/septic shock, but they have not explored a preferred method for measurement of ScvO2. While continuous monitoring was used in at least one study of early goal directed therapy, it has been proposed that ScvO2 can be adequately monitored on an intermittent basis, via central venous blood gases. METHODS: This 2 year longitudinal study examined 283 patients meeting the standard criteria for severe sepsis/septic shock who were admitted to a community hospital ICU. Sepsis protocol was initiated and a central line placed for monitoring ScvO2 in all patients. Standard peripherally inserted central catheters (PICC), triple lumen catheters (TLC), or pulmonary artery catheters (SG) were inserted in 106 patients, and continuous measurement catheters (Edwards) were placed in 177 patients. All patients were evaluated to determine whether they met the six hour resuscitation goal for ScvO2 > 70%. In-hospital mortality was also assessed. RESULTS: Of patients with continuous ScvO2 monitoring 134 of 177 (75.7%) met the six-hour goal of > 70% vs. 64 of 106 (60.3%) with intermittent monitoring (p=0.0075). The mortality rate was lower in the continuous monitoring group (47/177 or 26.5%) vs. intermittent (41/106 or 38.7%) (p=0.0349). CONCLUSION: These data support continuous monitoring of ScvO2 versus intermittent monitoring. Continuous observation of ScvO2 allowed for improved bundle compliance, improved survival, and more rapid treatment of the value to improve global tissue perfusion. CLINICAL IMPLICATIONS: It is possible that continuous display of ScvO2 induces continuous activity toward the 70% goal, and intermittent, non-displayed values do not. The current results do not explain mechanisms for the differences in compliance and mortality. Validation in a larger data set will be important.