LAUSR

We read with great interest the systematic review and meta-analysis by Gendreau and colleagues [1] examining the relation between hypercapnia and mortality in patients affected by acute respiratory distress syndrome (ARDS). We would like, however, to raise two concerns. The first pertains to the methodological choice of dividing hypercapnia in “permissive hypercapnia” and “induced hypercapnia”, based on the study design. Acknowledging the effort to establish an operational basis for the meta-analysis, we nevertheless feel the categories created do not reflect actual clinical/physiological differences. As such, the conclusion clinical effects of hypercapnia were conflicting depending on the mechanism of hypercapnia seems inaccurate, for two reasons. The mechanism is the same – reduced alveolar ventilation leading to reduced CO2 removal; and those differences more probably reflect study protocol or standard clinical care employed (time lapse of studies included spans over 20 years). The second concern refers to the true meaning of hypercapnia when we could not find a single reference to pH throughout the paper. In biological systems, the CO2 molecule seems to be relatively nontoxic as long as its hydration does not lead to an acute accumulation of excess amounts of hydrogen ions [2]. Paramount to homeostatic functions, pH is kept at narrow range by shifts in HCO3/CO2 balance, with acute adjustments relying on CO2 removal regulation (through alveolar ventilation). Moreover, pH shift leads to important physiologic derangements mainly in protein based reactions, whose failure ultimately leads to multiorgan dysfunction [2, 3]. In ARDS, evidence based practice entails limiting volume and pressure delivered through mechanical ventilation to less compliant lungs, these measures jointly referred to as protective ventilation. Such volume limitation frequently leads to incomplete CO2 removal and hypercapnia. This can in itself have diverse effects in organ systems [2], with different analysis yielding conflicting results [4, 5]. Nevertheless, current guidelines recommend permissive hypercapnia, with some Authors specifically referring a pH target. In any two random patients, mechanically ventilated in an intensive care unit (ICU) we can have the same PaCO2 value but very different physiological status brought by associated acute kidney injury, shock or other metabolic imbalances leading to acidosis. It will be the overall homeostatic capacity of the patient the true determinant of the ability to overcome protective ventilation associated hypercapnia and subsequent acid/base buffer imbalance. This is why we feel pH is a better target, when talking about respiratory acidosis in ARDS mechanically ventilated patients. We acknowledge CO2 direct effects on some organs, leading to considerable dysfunction in these critically ill patients, but we feel that only a comprehensive analysis of gas exchange and acid–base status allows for correct physiological integration of disease severity and therapeutic strategic planning—should we be considering/analyzing permissive hypercapnia or permissive acidosis? And this is precisely why we feel the analysis by Gendreau and colleagues should have somehow considered pH, either as a variable or at the very least as a confounder. Regardless of these concerns, we applaud the authors for the breadth of data reviewed, bringing a new look into existing data regarding an important aspect of ARDS patient care.