Hypotension is strongly associated with poor patient outcome, and reversing this condition clearly should be a primary therapeutic goal when treating patients in the early stages of shock. Potent inotropic and vasoconstrictor agents are de rigueur in the treatment of shock. The therapeutic goal is to maintain the mean arterial pressure at levels above 60 mmHg, a value thought to be the minimal pressure head required for coronary and renal perfusion 1. Our predicament is how best to determine the mean arterial pressure level that will result in optimal tissue perfusion in a given patient. In other words, is a mean arterial pressure of 60 mmHg sufficient to assure adequate perfusion to all organs? In some patients this accepted minimal mean arterial pressure may not suffice to insure adequate tissue perfusion. Should we aim for higher, or perhaps even lower, mean arterial pressure values? Catecholamines, while extremely useful in treating decreases in cardiac output, may produce an unwelcome increase in myocardial O₂ consumption in cardiogenic shock, or may impede blood flow to oxygen-starved tissues in hypovolemic shock 2. In septic shock, the balance between the positive and the negative effects of vasopressor and inotropic agents are even more difficult to discern 3.

A reliable and practical method to detect the onset of tissue hypoxia in critically ill patients would be an invaluable tool in guiding the timing and aggressiveness of resuscitation efforts. Finding such a tool has bedeviled clinical investigators for many years. Given our present level of technology, our options in determining the adequacy of tissue oxygenation in the clinical setting remain limited. Direct measures of tissue oxygen concentration are not sufficient to characterize the complex interaction between cellular energy requirements and oxygen supply. More complex technology, such as magnetic resonance spectroscopy and near-infrared spectroscopy are either insensitive or impractical in the clinical setting.

A great deal of thinking has been devoted in the past to the relationship of systemic measures of O₂ delivery to O₂ consumption 4. Enthusiastic acceptance of the 'supranormal' O₂ delivery concept, produced by the infusion of fluid volume, inotropic agents and blood products 5, has been tempered by studies finding no efficacy 6, or even increased mortality 7, with this approach. Other clinical studies, however, have shown improved survival in individuals when treated immediately upon their arrival at the emergency department with a protocol designed to increase O₂ delivery 8. It appears that early efforts at resuscitation are critical to survival, whereas delays in restoring adequate O₂ delivery may result in an ischemia-reperfusion-type phenomenon and in greater mortality 9.

Measuring tissue PCO₂ with a gastric tonometer 10 or a sublingual tonometer 1112 has been proposed as a physiologically sound method of detecting decreases in organ perfusion. Although numerous clinical and experimental studies show a strong correlation between increases in tissue PCO₂ and poor patient outcome, gastric tonometry or sublingual tonometry have encountered variable clinical acceptance. Technical difficulties certainly have dampened the initial enthusiasm for PCO₂ tonometry, but a more challenging obstacle to the widespread use of this technology has been an inadequate understanding of the mechanism(s) that result in tissue CO₂ accumulation. A crucial issue regarding the physiology of tissue CO₂ accumulation is whether the rise in tissue PCO₂ results from decreases in cellular O₂ delivery (or metabolism by mitochondria) or from decreases in blood flow and the accumulation of 'aerobic' CO₂ generated in the tricarboxylic acid cycle 13.

Dubin and colleagues 14 explored this issue in the past by subjecting experimental animals to decreases in O₂ delivery produced by lowering flow (ischemic hypoxia) or by lowering arterial oxygen saturation (hypoxic hypoxia). There were increases in intestinal venous PCO₂ with ischemic hypoxia but not with hypoxic hypoxia. Given that both preparations presumably experienced similar degrees of tissue hypoxia, this finding suggests that blood flow, not dysoxia, is the primary determinant of increases in tissue PCO₂. This conclusion, moreover, was in consonance with the results obtained by Vallet and colleagues 15 in isolated dog skeletal muscle.

In the current issue of Critical Care, Dubin and colleagues 16 extend their findings by testing the hypothesis that increasing intestinal blood flow prevents a rise in tissue PCO₂ in septic animals. They subjected two groups of sheep to lipopolysaccharide infusion. One group received intravenous fluids at a rate that maintained baseline intestinal blood flow. The other group was aggressively fluid resuscitated, resulting in increases of 50% in intestinal blood flow over basal conditions. An additional group of animals served as a normal control. The intramucosal-to-arterial PCO₂ difference (ΔPCO₂) rose in the first group, whereas volume expansion prevented increases in ΔPCO₂ in the aggressively resuscitated group. Of interest, metabolic acidosis as evidenced by a widening of the anion gap was greater in the resuscitated group.

The avoidance of increases in tissue PCO₂ in this model of resuscitated sepsis supports the notion that hypoperfusion, not tissue hypoxia, is the mechanism responsible for the accumulation of CO₂ in tissues. This finding provides experimental validity to the concept developed in a mass transport model of tissue CO₂ exchange 17, in which increases in tissue and venous blood CO₂ concentration are shown to be markers of regional hypoperfusion and not of tissue hypoxia.

The mechanism responsible for the increased anion-gap acidosis due to unmeasured anions in the resuscitated animals cannot be ascertained from the measurements obtained in this study. It is conceivable that resuscitation with normal saline may have produced local tissue hypoxia, the result of O₂ radical species production. Measurements of intestinal lactate production would have been helpful in answering this question by establishing the degree of anerobiasis in the resuscitated animals.

The study of Dubin and colleagues is another milepost in our understanding of the mechanisms that govern increases in tissue PCO₂ during hypoxic and septic conditions. Since changes in tissue PCO₂ are likely to be determined by alterations in blood flow, this may explain why gastric mucosal PCO₂ improves with greater cardiac output but not with increases in O₂-carrying capacity produced by blood transfusions 18.

I believe that now is the time to renew our interest in clinical measures of tissue PCO₂, as we understand further the clinical importance of gastric and sublingual tonometry as markers of regional tissue perfusion.

Competing interests