Arterial-venous carbon dioxide tension difference during severe hemorrhage and resuscitation

To define clinically useful markers for determining the adequacy of resuscitation after hemorrhage.

Prospective study of 20 Yorkshire swine, using an established model for hemorrhagic shock in swine.

Clinically useful markers for assessing the adequacy of resuscitation after hemorrhage do not exist. We assess variables, such as BP and arterial blood pH. However, these variables do not correlate well with restoration of organ perfusion. In this study, 14 anesthetized swine were hemorrhaged to a mean arterial pressure of less than 25 mm Hg. After 30 mins without interventions, each animal was resuscitated with whole blood, hydroxyethyl starch, or normal saline. At baseline and during shock and resuscitation, hemodynamic variables, arterial and mixed venous blood gases, and arterial lactate concentrations were measured to determine which variables correlated most closely with the restoration of blood flow.

The correlation between cardiac index and arterial or mixed venous pH and the correlations with arterial lactate values were poor. The correlation between cardiac index and mixed venous hemoglobin saturation also was weak (r2 = .28). Cardiac index correlated best with the arterial-venous PCO2 difference (r2 = .67) and the arterial-venous pH difference (r2 = .38). Using multiple regression, a linear correlation was established between the cardiac index and the arterial-venous pH and PCO2 differences throughout shock and resuscitation (r2 = .91).

These findings suggest that the adequacy of resuscitation after hemorrhage can be assessed using paired arterial and mixed venous pH and PCO2 values.