Estimation of dead space fraction can be simplified in the acute respiratory distress syndrome

Acute lung injury and acute respiratory distress syndrome are characterized by a non-cardiogenic pulmonary edema responsible for a significant impairment of gas exchange. The pulmonary dead space increase, which is due primarily to an alteration in pulmonary blood flow distribution, is largely responsible for carbon dioxide retention. Previous studies, computing the pulmonary dead space by measuring the expired carbon dioxide and the Enghoff equation, found that the dead space fraction was significantly higher in the non-survivors; it was even an independent risk of death. The computation of the dead space not by measuring the expired carbon dioxide but by applying a rearranged alveolar gas equation that takes into account only the weight, age, height, and temperature of the patient could lead to widespread clinical diffusion of this measurement at the bedside.

In the previous issue of Critical Care, Siddiki and colleagues [1] presented the dead space fraction data collected at admission and on day 3 from two acute lung injury/acute respiratory distress syndrome (ALI/ARDS) databases (109 patients in the Mayo Clinic and 1,896 patients in the ARDS Network). Th e hospital mortality increased in direct proportion to an increase in the dead space fraction. For every 0.05 increment of the dead space fraction, the odds ratios for hospital mortality were 1.07 at day 1 and 1.12 at day 3. Th us, at fi rst sight, the results of Siddiki and colleagues represent merely a repetition of previous studies [2][3][4]. However, their study added a novel element in that the dead space fraction was computed more simply than in the previous studies [2][3][4] and, unlike in those studies, without monitoring of the expired carbon dioxide (CO 2 ). So that the current results may be better understood, a short summary of the theoretical aspects of the dead space computation is presented.
ALI and ARDS are characterized by a non-cardiogenic pulmonary edema with signifi cant impairment of gas exchange. Th e increases in the right-to-left intrapulmonary shunt and in low ventilation-to-perfusion ratio lead to hypoxemia, whereas the increase in pulmonary dead space reduces CO 2 removal [5,6]. Th e increase in pulmonary dead space is due mainly to alterations in the distribution of pulmonary blood fl ow originating from vascular obstruction and to regional overdistension of ventilated alveoli induced by the application of positive end-expiratory pressure (PEEP) and sometimes by the reduction in cardiac output [7][8][9].
Nuckton and colleagues [10] found that in patients with ARDS the pulmonary dead fraction measured at admission was signifi cantly higher in the non-survivors than in the survivors (0.63 ± 0.09 versus 0.54 ± 0.09); for every increase of 0.05 in the dead space fraction, the odds ratio of death increased by 45%. Subsequent studies showed that the dead space fraction, measured during the fi rst 6 days of mechanical ventilation, was signifi cantly higher in the non-survivors than in the survivors and in patients with ARDS compared with patients with ALI [2][3][4]. Th e odds ratios of death were 1.59 and 2.87 in the early and intermediate phases of ARDS, respectively [4].
Th e original Bohr's equation required the mean alveolar (PA CO2 ) and mixed expired (PE CO2 ) partial pressures of CO 2 in order to estimate the physiologic dead space: (PA CO2 − PE CO2 )/PA CO2 . Subsequently, given the diffi culty of measur ing the PA CO2 and given that PA CO2 and arterial pressure of CO 2 (Pa CO2 ) are nearly identical in normal subjects, Enghoff suggested that Pa CO2 be used instead of PA CO2 ; consequently, the new equation was (Pa CO2 -PE CO2 )/Pa CO2 . Th us, any changes in intra pulmonary shunting, ventilation/perfusion ratio, or CO 2 solubility (Haldane eff ect) can aff ect the pulmonary dead space.
Traditionally, the PE CO2 has been measured over several minutes by collecting the expired gas (into a Douglas

Abstract
Acute lung injury and acute respiratory distress syndrome are characterized by a non-cardiogenic pulmonary edema responsible for a signifi cant impairment of gas exchange. The pulmonary dead space increase, which is due primarily to an alteration in pulmonary blood fl ow distribution, is largely responsible for carbon dioxide retention. Previous studies, computing the pulmonary dead space by measuring the expired carbon dioxide and the Enghoff equation, found that the dead space fraction was signifi cantly higher in the non-survivors; it was even an independent risk of death. The computation of the dead space not by measuring the expired carbon dioxide but by applying a rearranged alveolar gas equation that takes into account only the weight, age, height, and temperature of the patient could lead to widespread clinical diff usion of this measurement at the bedside. bag) and analyzing it with a blood gas analyzer [11]. Nowadays, indirect calorimeters or volumetric capnography allows accurate PE CO2 measurements at the bedside [11][12][13]. Unfortunately, these methods can be cumber some and thus may limit the widespread measure ment of dead space in clinical practice.
In the present study, Siddiki and colleagues [1] 'estimated' the physiologic dead space not by taking into account the PE CO2 but by applying a rearranged alveolar gas equation. Th e dead space was equal to 1 − [(0.86 × V  CO2 estimated)/(minute ventilation × Pa CO2 ]. Th e V  CO2 estimated (the estimated CO 2 production) is simply measured by the Harris-Benedict equation, which takes into consideration the weight, height, age, and body temperature of the patient. Th e Harris-Benedict equation is commonly used to predict the resting energy expenditure (REE) [14]. Th ere are some important limitations to the study by Siddiki and colleagues: (a) the absence of any comparison of the physiologic dead space measurements obtained with the rearranged alveolar gas equation and with the Enghoff equation and (b) the use of the Harris-Benedict equation, which has been reported to be weakly related to the REE in severely ill patients [15]. However, even with these limitations, these data suggest that the pulmonary dead space can be measured in ALI/ARDS patients non-invasively, rapidly, and without any expensive equipment. Th us, in addition to taking daily measurements of oxygenation, compliance, and transpulmonary pressure, clinicians should include the pulmonary dead space for better clinical management.
Abbreviations ALI, acute lung injury; ARDS, acute respiratory distress syndrome; CO 2 , carbon dioxide; Pa CO2, arterial partial pressure of carbon dioxide; PA CO2 , alveolar partial pressure of carbon dioxide; PE CO2, expired partial pressure of carbon dioxide; REE, resting energy expenditure; V · CO2 estimated, estimated carbon dioxide production.

Competing interests
The authors declare that they have no competing interests.