End-tidal carbon dioxide as a noninvasive indicator of cardiac index during circulatory shock

Abstract

OBJECTIVE: To document the relationships between cardiac index and end-tidal carbon dioxide tension (PetCO2 during diverse low-flow states of circulatory shock. DESIGN: Randomized, prospective, controlled studies on animal models of hemorrhagic, septic, and cardiogenic shock. SETTING: University-affiliated research laboratory. SUBJECTS: Sixteen anesthetized domestic pigs weighing 35-45 kg. INTERVENTIONS: Hemorrhagic shock was induced in five pigs by bleeding followed by reinfusion of shed blood. Septic shock was induced in five pigs by infusion of live Escherichia coli. Cardiogenic shock followed an interval of global myocardial ischemia after inducing and reversing ventricular fibrillation in six pigs. MEASUREMENTS AND MAIN RESULTS: PetCO2 was continuously measured. Cardiac index was measured intermittently by using conventional thermodilution techniques. Cardiac index was correlated with PetCO2 by polynomial regression and Bland-Altman analyses. PetCO2 was highly correlated with cardiac index during hemorrhagic shock (r2 = .69, p < .01), septic shock (r2 = .65, p < .01), and cardiogenic shock (r2 = .81, p < .01). PetCO2 predicted thermodilution cardiac index with bias of -11+/-27 (+/-2 SD) mL/min/kg during hemorrhagic shock, 1.3+/-20.4 (+/- 2 SD) mL/min/kg during septic shock, and -1+/-12 (+/-2 SD) mL/min/kg during cardiogenic shock. CONCLUSIONS: Cardiac output and PetCO2 were highly related in diverse experimental models of circulatory shock in which cardiac output was reduced by >40% of baseline values. Therefore, measurement of PetCO2 is a noninvasive alternative for continuous assessment of cardiac output during low-flow circulatory shock states of diverse causes.