End-tidal CO2 changes under constant cardiac output during cardiopulmonary resuscitation.

Objectives: To evaluate a) whether end-tidal CO2 values change under constant cardiac output during cardiopulmonary resuscitation (CPR), and b) what factors are responsible for the change.

Design: A cohort study.

Setting: University research laboratory.

Subjects: Nine mongrel dogs.

Interventions: Ventricular fibrillation was electrically induced. After 2 mins, open-chest cardiac massage was initiated to maintain cardiac output at 0.2 L/min (23% of baseline cardiac output) by the measurement of blood flow with an electromagnetic flow probe on the ascending aorta. The cardiac massage was kept constant until 50 mins after the induction of ventricular fibrillation.

Measurements And Main Results: Before and during ventricular fibrillation, end-tidal CO2, minute volume of alveolar ventilation, and CO2 excretion were continuously monitored. Blood gases and oxygen saturation values were also measured in arterial and the mixed venous blood samples. CO2 content was calculated. After induction of ventricular fibrillation, end-tidal CO2 decreased and thereafter continued to increase until the end of the experiment. Two mechanisms may have contributed to the early reduction in end-tidal CO2. One mechanism is a further decrease in CO2 excretion compared with the reduction in alveolar ventilation and the other is an increase in alveolar deadspace (estimated from the increase in the difference between PaCO2 and end-tidal CO2). The subsequent increase in end-tidal CO2 was mainly due to a change in CO2 excretion. There are two hypotheses concerning the subsequent increase in CO2 excretion: the increase in pulmonary capillary blood flow (estimated from the change in the arteriovenous CO2 content gradient) and the increase in CO2 production itself.

Conclusions: End-tidal CO2 changes under constant cardiac output during CPR. When end-tidal CO2 is used to estimate the effectiveness of the cardiac massage, this type of change must be recognized.