DO2/VO2 relationships

J. L. Vincent

Introduction
Most cellular activities require oxygen, primarily obtained from the degradation
of adenosine triphosphate (ATP) and other high-energy compounds. Oxygen
must, therefore, be present in the mitochondria in sufficient amounts to maintain
effective concentrations of ATP by the electron transport system. Cells must
perform various activities in order to survive, including membrane transport,
growth, cellular repair, and maintenance processes. They often also have faculta-
tive functions, such as contractility, electrolyte or protein transport, motility, or
various biosynthetic activities. If oxygen availability is limited, cellular oxygen
consumption may fall, and become supply-dependent. Facultative functions are
the first to be affected, leading to cellular and, ultimately, organ dysfunction. If the
situation becomes more serious, obligatory functions can no longer be main-
tained, and irreversible alterations may occur resulting in cell death. Maintaining
sufficient oxygen availability to the cell is thus fundamental for cell survival: the
hypoxic cell is doomed to become malfunctional and to die.

Oxygen delivery vs oxygen availability

The amount of oxygen available in the cell is determined by a number of central
and peripheral factors. The central factors depend on the adequacy of cardiorespi-
ratory function (cardiac index and PaO2) and the hemoglobin concentration,
according to the formulas given in Table 1. Peripheral factors depend on the
distribution of cardiac output to the various organs, and the regulation of the
microcirculation, which is determined by the autonomic control of vascular tone,
local microvascular responses, and the degree of affinity of the hemoglobin mole-
cule for oxygen.
Among the central factors, cardiac output is a more important determinant of
oxygen delivery (DO2) than the arterial oxygen content (Table 1), as a fall in
hemoglobin or SaO2 can be compensated by an increase in cardiac output, whereas
the opposite is not true. If cardiac output falls, SaO2 cannot rise above 100% and
hemoglobin concentration cannot increase acutely. Furthermore, an increase in
red blood cell mass does not efficiently increase DO2, because cardiac output
usually decreases as a result of the associated increase in blood viscosity. Hence,
252 J. L. Vincent

Table 1. The determinants of oxygen delivery, oxygen consumption, and oxygen extraction

Oxygen delivery (DO2) = CO x Hb x SaO2 x C x 10

Oxygen consumption (VO2) = CO x (CaO2 CvO2) x 10
(Neglecting the dissolved oxygen) = CO x Hb x (SaO2-SvO2) x C
Oxygen extraction (O2ER) = VO2/DO2 = (CaO2-CvO2)/CaO2
or neglecting the dissolved oxygen = (SaO2-SvO2)/SaO2
where CO represents the cardiac output, Hb the hemoglobin concentration, SaO2 and SvO2 the
arterial and the mixed venous oxygen saturations, respectively, and C the constant value repre-
senting the amount of oxygen bound to 1 g of Hb (this value is usually 1.34 or 1.39).

cardiac output is the most important factor in the constant adaptation of the body’s
oxygen needs in physiological conditions.
The peripheral factors can change substantially in inflammatory conditions
(including sepsis), when local control of the vascular tone may be altered, the
formation of microthrombi may shut down some capillaries, and edema may
develop. Changes in hemoglobin oxygen affinity can also influence the peripheral
delivery of oxygen.

Basic concepts: The Relationship between VO 2 and DO2

and the concept of VO2/DO2 Dependency
A number of animal experiments using different models [1–4] have shown that
oxygen uptake (VO2) remains independent of DO2 over a wide range of values,
because oxygen extraction (O2ER, which is the ratio of VO2 over DO2) can readily
adapt to the changes in DO2. When cardiac output is acutely reduced by acute
blood withdrawal, tamponade, anemia, or hypoxemia, O2ER increases (SvO2 de-
creases) and VO2 remains quite stable, until DO2 falls below a critically low
threshold (DO2crit), when VO2 starts to fall. An abrupt increase in blood lactate
concentrations then occurs, indicating the development of anaerobic metabolism
(Fig. 1). In the presence of sepsis mediators, as after the administration of endo-
toxin or live bacteria [5, 6], oxygen extraction capabilities are altered so that the
DO2crit is higher and the critical O2ER is typically lower than in control condi-
tions. In these conditions, VO2 can become dependent on DO2 even when DO2 is
normal or elevated. Altogether, these observations help to characterize the four
principal types of circulatory shock (Fig. 2).
Although such studies performed in anesthetized animals can hardly be repro-
duced in humans, an acute reduction in DO2 can be observed in the intensive care
unit (ICU) during withdrawal of life support [7]. In these dying patients, VO2
remained relatively constant until DO2 fell below very low values.
A number of studies have correlated the VO2/DO2 dependency phenomenon to
profound circulatory alterations. Bihari et al. [8] showed that an increase in VO2
during a prostacyclin infusion was a characteristic of non-survivors. A number of
DO2/VO2 relationships 253

Fig. 1. Relationship between oxygen

uptake (VO2) and oxygen delivery
(DO2) when DO2 is acutely reduced by
tamponade or hemorrhage in anesthe-
tized animals. Note that blood lactate
levels increase as soon as DO2 falls be-
low DO2crit.

Fig. 2. The four types of

acute circulatory failure.

investigators have also reported that patients with acute circulatory failure with
increased blood lactate concentrations demonstrate an increase in VO2 when DO2
is acutely increased by fluid infusion [9], blood transfusions or dobutamine ad-
ministration [10]. Such a phenomenon has not been observed in stable patients
with normal lactate concentrations [9–12].
Others have challenged these observations, arguing that the VO2 was usually
determined from the Fick principle rather than determined independently from
expired gas analysis. Hence, VO2 and DO2 were calculated from the same variables,
i.e., cardiac output, hemoglobin concentrations, and SaO2, resulting in mathemati-
cal coupling of data.
Indirect calorimetry also has its limitations and sources of error, and becomes
very imprecise when high FiO2 are delivered. Incidentally, many authors have
argued that VO2 is calculated using the Fick equation, but measured when obtained
by indirect calorimetry. This is clearly wrong: With both techniques, VO2 results
from a calculation of the product of flow (blood flow or gas flow) and oxygen
content differences (between arterial and venous blood or between inspired and
254 J. L. Vincent

Table 2. Calculation of oxygen uptake by indirect calorimetry

FiO2 x (1– FeO2 – FeCO2)

VO2 = x VE
(1 – FiO2 – FeO2)
where FeCO2 is the expired CO2 fraction, FiO2 and FeO2 the inspired and expired oxygen fraction,
respectively, and VE the expiratory flow rate

expired gases). In fact, the formula used to calculate VO2 by indirect calorimetry
is quite complex (Table 2).
In addition, this reasoning can itself be criticized. First, the effect of mathemati-
cal coupling of data does not seem to be major if the changes in DO2 are of sufficient
magntitude [13]. Second, this limitation cannot explain how the changes in VO2
can be observed in some individuals and not in others. It is important to note that
all studies using indirect calorimetry to determine VO2 included only stabilized
patients: this is largely due to the time needed to install the material used for VO2
determinations. The same applies to the studies arguing that changes in VO2 can
be observed only in patients with high lactate concentrations: these studies in-
cluded stabilized patients in whom signs of shock had already resolved. Admittedly,
the interpretation of elevated blood lactate concentrations is not always straight-
forward, as hyperlactatemia can be influenced by decreased lactate clearance. Also,
in sepsis, hyperlactatemia does not necessarily reflect anaerobic metabolism sec-
ondary to cellular hypoxia, but other mechanisms, like increased glycolysis or
abnormal pyruvate metabolism [14]. Hence, hyperlactatemia should complement
the clinical evaluation of circulatory shock, including arterial hypotension and
signs of altered tissue perfusion like altered sensorium, altered cutaneous perfu-
sion, and decreased urine output.
Altogether, these studies indicate that the VO2/DO2 dependency phenomenon
can be observed but only in patients who are clearly unstable, during shock
resuscitation; it is a hallmark of acute circulatory failure (shock) [15].
A more important limitation is that the global VO2/DO2 assessment is not
precise enough to be useful clinically and, more specifically, to guide therapy.
Furthermore, VO2/DO2 dependency may occur regionally, especially in the hepato-
splanchnic region [16] (Fig. 3). Comparisons of VO2 and DO2 are useless, because
obtaining these derived variables is hard to interpret and the plot of VO2 vs DO2 is
limited by the problem of mathematical coupling of data. However, evaluation of
the relationship between cardiac output and oxygen extraction may be very useful
to evaluate the adequacy of the cardiac output response [17]. Such a CI/O2ER
relationship has no problem of mathematical coupling of data (Fig. 4). Increased
lactate concentrations remain a reliable prognostic indicator, actually superior to
DO2 and VO2 values [18]; increasing DO2 to higher values when blood lactate levels
are normal has not been shown to be beneficial.
DO2/VO2 relationships 255

Fig. 3. Regional VO2/DO2 relationship in the splanchnic circulation in patients with severe sepsis.
Group I: patients with gradient between mixed venous and hepatic venous oxygen saturation
lower than or equal to 10%. Group II: patients with gradient between mixed venous and hepatic
venous oxygen saturation higher than 10%. Data are presented as mean ± SEM. (From [16] with
permission)

Fig. 4. Cardiac index/O2ER diagram during a short term dobutamine infusion indicating VO2/DO2
dependency in patients with increased lactate levels but not in those with normal lactate levels
(data from [10]).
256 J. L. Vincent

Clinical implications

The Supranormal DO2 Approach

William Shoemaker and his colleagues proposed that DO2 should be maintained
at supranormal values (at least 600 ml/min.M²) in all patients at risk of complica-
tions, to ensure sufficient oxygen availability to the cells [19]. This proposal was
based on the observation that survivors from sepsis or trauma usually generate
higher DO2 than non-survivors [20]. Although this approach may have merits in
some populations [21, 22], it is limited by two important aspects. One is that
patients with higher DO2 are more likely to survive, simply because they have a
better physiological reserve, allowing them to generate a higher cardiac output.
The second is that increasing DO2 to supranormal values in all patients ‘at risk’
may be beneficial to some, still underresuscitated, but harmful to others, already
well resuscitated, who would thus receive too much fluid and adrenergic agents
like dobutamine.
This concept is an oversimplification of a complex phenomenon. When applied
to a mixed group of critically ill patients, such strategies have been shown to be
ineffective [23] and may even be harmful, especially if high doses of dobutamine
are administered [24].

The Titrated Approach

It is more meaningful to have a titrated approach, individualized according to
results of a careful clinical evaluation and some paraclinical tests including meas-
urements of cardiac index, SvO2, blood lactate concentrations, and perhaps re-
gional PCO2. This requires a complete understanding of the pathophysiologic
alterations
As mentioned above, the relationship between CI and SvO2 does not have the
problem of mathematical coupling of data associated with the evaluation of the
relationship between VO2 and DO2 when both are obtained from the same values
of cardiac output, hemoglobin concentrations, SaO2, and SvO2. The study of such
variables also avoids cumbersome calculations, as cardiac index is a primary
variable and O2ER is very simply calculated (Table 1). In most cases, the relation-
ship between CI and SvO2 or even central venous oxygen saturation (ScvO2) alone
may suffice. There are, however, two reasons why the relationship between CI and
O2ER would be better (Fig. 4.). One is that the relationship between CI and SvO2 is
curvilinear, rendering the data interpretation more difficult. The second, is that
even when hypoxemia is avoided, SaO2 can still vary between about 90 and 99% in
the acutely ill patient, i.e., a 10% variation in the variable. Nevertheless, SvO2, or
maybe even ScvO2 alone, may be used in an algorithm for resuscitation. Rivers et
al. [25] showed that monitoring ScvO2 could result in a significantly lower mortality
rate in patients with severe sepsis and septic shock. Likewise, Polonen et al.
[26]found, in cardiac surgery patients, that maintaining SvO2 at normal or high
levels shortens hospital stay and lowers the degree of organ dysfunction at time of
discharge from hospital. Nevertheless, lactate concentrations remain valuable in
DO2/VO2 relationships 257

shock states. Although one may argue that lactate concentrations reflect other
cellular abnormalities than anerobic metabolism secondary to hypoxia, persist-
ently raised lactate levels should represent an alarm signal. Hence, in addition to
clinical evaluation, repeated measurements of SvO2 and blood lactate may be
helpful.

Conclusion
Maintenance of adequate DO2 is essential to preserve organ function, as a low DO2
is a straightforward path to organ failure and death, and treatment must be
titrated to the individual based on the integration of several factors including
clinical examination and available oxygenation and hemodynamic parameters.
The relationship between VO2/DO2 remains an important concept, even though
its simple application to guide therapy may be too simplistic. The relationship
between cardiac index and O2ER (or its simplification SvO2) can be helpful.

References
1. Cain SM (1977) Oxygen delivery and uptake in dogs during anemic and hypoxic hypoxia. J
Appl Physiol 42:228–234
2. Nelson DP, Beyer C, Samsel RW, Wood LDH, Schumacker PT (1987) Pathological supply
dependence of O2 uptake during bacteremia in dogs. J Appl Physiol 63:1487–1492
3. Van der Linden P, Gilbert E, Engelman E, Schmartz D, Vincent JL (1991) Effects of anesthetic
agents on systemic critical O2 delivery. J Appl Physiol 71:83–93
4. Zhang H, Spapen H, Benlabed M, Vincent JL (1993) Systemic oxygen extraction can be
improved during repeated episodes of cardiac tamponade. J Crit Care 8:93–99
5. Samsel RW, Nelson DP, Sanders WM, Wood LDH, Schumacker PT (1988) Effect of endotoxin
on systemic and skeletal muscle O2 extraction. J Appl Physiol 65:1377–1382
6. Zhang H, Vincent JL (1993) Oxygen extraction is altered by endotoxin during tamponade-in-
duced stagnant hypoxia in the dog. Circ Shock 40:168–176
7. Ronco JJ, Fenwick JC, Tweeddale MG, et al (1993) Identification of the critical oxygen delivery
for anaerobic metabolism in critically ill septic and nonseptic humans. JAMA 270:1724–1730
8. Bihari D, Smithies M, Gimson A, Tinker J (1987) The effects of vasodilation with prostacyclin
on oxygen delivery and uptake in critically ill patients. N Engl J Med 317:397–403
9. Haupt MT, Gilbert EM, Carlson RW (1985) Fluid loading increases oxygen consumption in
septic patients with lactic acidosis. Am Rev Respir Dis 131:912–916
10. Vincent JL, Roman A, De Backer D, Kahn RJ (1990) Oxygen uptake/supply dependency:
Effects of short-term dobutamine infusion. Am Rev Respir Dis 142:2–8
11. Bakker J, Vincent JL (1991) The oxygen supply dependency phenomenon is associated with
increased blood lactate levels. J Crit Care 6:152–159
12. Gilbert EM, Haupt MT, Mandanas RY, Huaringa AJ, Carlson RW (1986) The effect of fluid
loading, blood transfusion and catecholamine infusion on oxygen delivery and consumption
in patients with sepsis. Am Rev Respir Dis 134:873–878
13. Stratton HH, Feustel PJ, Newell JC (1987) Regression of calculated variables in the presence
of shared measurement error. J Appl Physiol 62:2083–2093
14. Gore DC, Jahoor F, Hibbert JM, DeMaria EJ (1996) Lactic acidosis during sepsis is related to
increased pyruvate production, not deficits in tissue oxygen availability. Ann Surg 224:97–102
258 J. L. Vincent

15. Friedman G, De Backer D, Shahla M, Vincent JL (1998) Oxygen supply dependency can
characterize septic shock. Intensive Care Med 24:118–123
16. De Backer D, Creteur J, Noordally O, Smail N, Gulbis B, Vincent JL (1998) Does hepa-
tosplanchnic VO2/DO2 dependency exist in critically ill patients. Am J Respir Crit Care Med
157:1219–1225
17. Silance PG, Simon C, Vincent JL (1994) The relation between cardiac index and oxygen
extraction in acutely ill patients. Chest 105:1190–1197
18. Bakker J, Coffernils M, Leon M, Gris P, Vincent JL (1991) Blood lactate levels are superior to
oxygen derived variables in predicting outcome in human septic shock. Chest 99:956–962
19. Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee TS (1988) Prospective trial of supra-
normal values of survivors as therapeutic goals in high-risk surgical patients. Chest
94:1176–1186
20. Shoemaker WC, Montgomery ES, Kaplan E, Elwyn DH (1973) Physiologic patterns in surviv-
ing and nonsurviving shock patients. Use of sequential cardiorespiratory variables in defining
criteria for therapeutic goals and early warning of death. Arch Surg 106:630–636
21. Yu M, Levy MM, Smith P, Takiguchi SA, Miyasaki A, Myers SA (1993) Effect of maximizing
oxygen delivery on morbidity and mortality rates in critically ill patients: A prospective,
randomized, controlled study. Crit Care Med 21:830–838
22. Lobo SM, Salgado PF, Castillo VG, et al (2000) Effects of maximizing oxygen delivery on
morbidity and mortality in high-risk surgical patients. Crit Care Med 28:3396–3404
23. Gattinoni L, Brazzi L, Pelosi P, et al (1995) A trial of goal-oriented hemodynamic therapy in
critically ill patients. N Engl J Med 333:1025–1032
24. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D (1994) Elevation of systemic
oxygen delivery in the treatment of critically ill patients. N Engl J Med 330:1717–1722
25. Rivers E, Nguyen B, Havstad S, et al (2001) Early goal-directed therapy in the treatment of
severe sepsis and septic shock. N Engl J Med 345:1368–1377
26. Polonen P, Ruokonen E, Hippelainen M, Poyhonen M, Takala J (2000) A prospective,
randomized study of goal-oriented hemodynamic therapy in cardiac surgical patients. Anesth
Analg 90:1052–1059