Oxygen Content, Delivery, Extraction, and Utilization

Oxygen transport is a sequence: blood loads oxygen, the circulation distributes it, tissues extract it, and cells use it. Chapter 4 introduced that sequence. This chapter examines how its measurements fit together, where their assumptions fail, and why increasing a calculated transport value does not by itself establish a useful treatment response.

34.1 Oxygen Content Counts the Oxygen Carried in Blood

Arterial oxygen tension, PaO₂, describes the pressure associated with dissolved oxygen. Arterial oxygen content, CaO₂, describes the amount of oxygen carried in a volume of blood. Most of that oxygen is bound to hemoglobin under ordinary conditions. The smaller dissolved component remains important because oxygen tension helps determine loading, diffusion, and unloading.

A commonly used clinical approximation is:

CaO₂ is expressed in mL O₂/dL of blood. Hb is hemoglobin concentration in g/dL. SaO₂ is entered as a fraction, so 97% becomes 0.97. PaO₂ is in mm Hg. The coefficient 1.34 approximates the oxygen carried by one gram of fully oxygenated hemoglobin. The coefficient 0.0031 approximates dissolved oxygen in mL O₂/dL per mm Hg under usual physiological conditions.

These coefficients are conventions for estimation, not immutable constants for every temperature, hemoglobin mixture, or instrument. Shimizu and colleagues measured oxygen-binding capacity together with inactive hemoglobin species in blood samples. Their results illustrate why measured capacity and a theoretical capacity need not be identical. This chapter retains the 1.34 coefficient used throughout the book. 1

Consider illustrative Hb of 8 g/dL, SaO₂ of 97%, and PaO₂ of 90 mm Hg. The bound component is 1.34 × 8 × 0.97 = 10.3984 mL O₂/dL. The dissolved component is 0.0031 × 90 = 0.279 mL O₂/dL. Their sum is approximately 10.68 mL O₂/dL. Each 100 mL of arterial blood carries that amount in this model.

If PaO₂ rises to 300 mm Hg while hemoglobin remains 8 g/dL and saturation approaches 100%, the estimate becomes 1.34 × 8 + 0.0031 × 300 = 11.65 mL O₂/dL. A large rise in tension has produced a much smaller proportional rise in content. This arithmetic does not assess the other effects or appropriateness of additional oxygen.

34.2 Saturation Must Be Matched to the Measurement Method

SaO₂, SpO₂, and a blood gas's calculated saturation are related but different quantities. Pulse oximetry estimates arterial saturation from an optical pulsatile signal. A calculated blood gas saturation derives a value from oxygen tension and assumed dissociation behavior. Co-oximetry uses multiple wavelengths to distinguish measured hemoglobin species.

The denominator also matters. Functional saturation is oxyhemoglobin divided by oxyhemoglobin plus deoxygenated hemoglobin. Fractional oxyhemoglobin uses total measured hemoglobin, including relevant dyshemoglobin species. These values become similar when abnormal species are negligible, but can diverge when carboxyhemoglobin or methemoglobin is substantial.

Imagine 100 illustrative hemoglobin units comprising 90 oxyhemoglobin, 5 deoxygenated hemoglobin, and 5 carboxyhemoglobin units. Functional saturation is 90 ÷ 95 ≈ 94.7%. Fractional oxyhemoglobin is 90 ÷ 100 = 90%. Multiplying total hemoglobin by the functional value as though all species were normally available would overestimate oxygen carriage.

Buckley and colleagues compared pulse oximetry with co-oximetry in patients with carbon monoxide intoxication. The pulse reading failed to reflect the reduction in oxyhemoglobin accurately. This clinical study supports direct assessment of hemoglobin species when exposure or measurement discordance raises concern. 2

Carbon monoxide also alters oxygen unloading from the remaining binding sites, while methemoglobin cannot bind oxygen normally at its oxidized iron site. A preserved PaO₂ describes dissolved oxygen and does not exclude either problem. Applying the ordinary content equation requires an appropriate hemoglobin and saturation definition rather than an additional arbitrary correction.

Motion, poor perfusion, optical limitations, and sampling errors can also produce disagreement. A high displayed saturation should be reconciled with the measurement method and clinical findings before it is used to conclude that oxygen transport is adequate.

34.3 Delivery Requires Both Content and Flow

Systemic oxygen delivery, DO₂, is the oxygen transported into the systemic circulation each minute. It is calculated by multiplying cardiac output by arterial oxygen content:

Cardiac output is in L/min, while content is in mL O₂/dL. One liter contains 10 deciliters, so the factor of 10 converts the blood-flow units. The result is mL O₂/min. Omitting that factor underestimates delivery tenfold.

Using the previous content of 10.6774 mL O₂/dL, an illustrative output of 4 L/min moves 40 dL of blood each minute. Delivery is 4 × 10.6774 × 10 ≈ 427 mL O₂/min. If output rises to 6 L/min at unchanged content, delivery becomes 6 × 10.6774 × 10 ≈ 641 mL O₂/min.

The equation does not establish that increasing output will improve tissue oxygen use. It describes the increased global transport produced if the assumed change occurs. Actual treatment may also change hemoglobin concentration, regional distribution, metabolic demand, and filling pressures.

Indexing introduces another unit. Replacing cardiac output with cardiac index gives an oxygen delivery index expressed in mL O₂/min/m². For a body surface area of 1.8 m², the first example corresponds to 427 ÷ 1.8 ≈ 237 mL O₂/min/m². It must not be compared directly with an unindexed value as if the two numbers used the same scale.

Global transport is also different from effective regional delivery. A high-output circulation can still distribute flow unevenly. An organ with restricted inflow or poor capillary perfusion may receive inadequate oxygen despite a satisfactory total. The global calculation establishes the amount entering the arterial circulation, while further evidence establishes where it goes.

34.4 The Fick Relationship Accounts for Oxygen Removed from Blood

Mixed venous blood carries oxygen remaining after systemic extraction. Its oxygen content, CvO₂, can be estimated using mixed venous saturation and tension with a compatible content equation. In a sufficiently stable circulation, oxygen uptake equals flow multiplied by the arterial-to-mixed-venous content difference:

VO₂ is in mL O₂/min when flow is in L/min and both contents are in mL O₂/dL. The arterial and venous samples should describe the same physiological period. An arbitrary peripheral venous sample is not mixed venous blood.

Continue the illustrative Hb of 8 g/dL and arterial content of 10.6774 mL O₂/dL. Suppose mixed venous saturation is 55% and mixed venous PO₂ is 35 mm Hg. Venous content is 1.34 × 8 × 0.55 + 0.0031 × 35 = 6.0045 mL O₂/dL.

The difference is 10.6774 − 6.0045 = 4.6729 mL O₂/dL. At output of 4 L/min, uptake is 4 × 4.6729 × 10 ≈ 187 mL O₂/min. This is a calculated uptake under the stated assumptions, not an independent measurement of mitochondrial activity.

The ordinary systemic calculation also assumes that the sampled arterial and venous streams represent the relevant outflow and return. Intracardiac shunts or extracorporeal oxygen exchange require explicit attention to which stream is sampled and which flow belongs in the balance. A pulmonary flow estimate cannot be substituted automatically for systemic flow when they differ.

The relationship can instead calculate output from measured VO₂ and the content difference. If VO₂ is assumed, its error enters the output estimate. If output is measured and used to calculate VO₂, the resulting uptake shares inputs with calculated DO₂.

Phang and colleagues examined this shared-input problem in ARDS. Mathematical coupling could produce apparent dependence of calculated consumption on calculated delivery. An increase in both values therefore does not prove that the tissues increased oxygen use. Independent respiratory gas measurements can help, provided their own leak, oxygen-fraction, calibration, and steady-state requirements are met. 3

34.5 Extraction Links Delivery, Consumption, and Venous Saturation

The oxygen extraction ratio is the fraction of delivered oxygen removed across the systemic circulation:

In the preceding example, uptake is about 187 mL O₂/min and delivery about 427 mL O₂/min. Using the unrounded contents, extraction is 4.6729 ÷ 10.6774 ≈ 0.438, or 43.8%. This describes the whole-body balance represented by those samples. It is not a treatment target or a measurement of extraction in each organ.

If dissolved oxygen is neglected and hemoglobin concentration is the same in arterial and venous samples, the common hemoglobin term cancels:

With SaO₂ of 0.97 and SvO₂ of 0.55, the estimate is (0.97 − 0.55) ÷ 0.97 ≈ 43.3%. Its small difference from the full content calculation reflects the neglected dissolved terms. Those terms become more relevant with very high oxygen tensions or low hemoglobin.

The Fick relationship can also estimate the saturation that would accompany an assumed uptake:

Suppose uptake is 200 mL O₂/min, output is 4 L/min, Hb is 8 g/dL, and SaO₂ is 0.97. The saturation decrement is 200 ÷ (4 × 10 × 1.34 × 8) ≈ 0.466. Estimated SvO₂ is 0.97 − 0.466 ≈ 0.504, or 50.4%. Both dissolved components are neglected in this separate example.

A lower venous saturation can therefore result from lower flow, lower hemoglobin, lower arterial saturation, or higher uptake. The equation explains their interaction but cannot identify which changed without additional measurements.

34.6 Critical Delivery Is a Physiological Transition Rather Than One Universal Number

When delivery falls, tissues can initially remove a larger fraction of the oxygen arriving. Venous content falls while oxygen uptake remains approximately stable. This compensation preserves oxidative metabolism over a range of reduced transport.

Eventually, extraction cannot increase enough to preserve uptake. Further reductions in delivery then reduce oxygen consumption. This is supply dependence. The transition is often described as critical oxygen delivery, but it depends on the conditions under which the relationship is measured.

Cain experimentally reduced oxygen availability through anemia or hypoxemia in anesthetized dogs. Once oxygen supply became limiting, uptake fell with delivery. The experiment provides a mechanistic demonstration of the transition. Its numerical critical value is not a threshold for an awake human, a septic organ, or a specific ICU intervention. 4

Demand changes the transition. Fever, seizures, shivering, and respiratory muscle loading can increase the amount of oxygen required. Regional flow distribution and diffusion also influence how much of the delivered oxygen is usable. The same global delivery can therefore be adequate under one condition and inadequate under another.

A global curve combines many organs. A vulnerable region can become supply limited while uptake elsewhere compensates enough to preserve the whole-body value. Conversely, a fall in global uptake can reflect reduced demand after treatment rather than inadequate supply.

The response to a delivery intervention adds evidence only when measured independently and interpreted in context. Increased uptake after an inotrope might represent improved availability, increased cardiac work, another rise in metabolic demand, or shared calculation error. Supply dependence is an explanation requiring evidence, rather than a label established by any upward pair of numbers.

34.7 Anemia Changes the Circulatory Task

A lower hemoglobin concentration reduces the bound oxygen carried in each deciliter. To preserve delivery at the same arterial saturation, the circulation must move more blood, tissues must extract a larger fraction, demand must fall, or some combination must occur.

At an illustrative SaO₂ of 98% and PaO₂ of 100 mm Hg, Hb of 12 g/dL gives 1.34 × 12 × 0.98 + 0.31 ≈ 16.07 mL O₂/dL. Hb of 8 g/dL gives 1.34 × 8 × 0.98 + 0.31 ≈ 10.82 mL O₂/dL. Saturation has not changed, but oxygen per blood volume has fallen substantially.

An output of 5 L/min at the higher content gives approximately 5 × 16.0684 × 10 ≈ 803 mL O₂/min. Maintaining that delivery at content of 10.82 would require approximately 803 ÷ (10.82 × 10) ≈ 7.4 L/min. This arithmetic demonstrates the additional flow requirement. It is not a request to achieve that output in a patient.

Acute isovolemic anemia can reduce viscosity and systemic resistance and provoke cardiovascular compensation. Weiskopf and colleagues examined severe acute anemia under controlled human conditions and demonstrated increased heart rate, stroke volume, and cardiac index. Their selected resting population and controlled volume state cannot establish a safe hemoglobin for bleeding, shock, coronary ischemia, or increased metabolic demand. 5

Compensation has consequences. Increasing heart rate and output can increase myocardial work. Coronary disease, ventricular dysfunction, impaired vascular regulation, and hypoxemia can reduce reserve. Acute blood loss also removes circulating volume, so it cannot be interpreted as isovolemic hemodilution.

Transfusion changes content, volume, viscosity, and sometimes the subsequent flow response. The increase in hemoglobin therefore does not guarantee a proportional increase in tissue oxygen use. Physiology identifies the potential gain; outcome evidence determines when that gain justifies treatment in a defined population.

34.8 Central and Mixed Venous Blood Represent Different Mixtures

Mixed venous saturation, SvO₂, is measured in pulmonary arterial blood after venous return has mixed. Central venous saturation, ScvO₂, is usually sampled from the superior vena cava or a nearby central location. These samples represent different proportions of regional drainage, including lower-body and coronary venous return.

A central sample therefore cannot be inserted into every mixed-venous calculation without changing its meaning. Chawla and colleagues compared paired samples in critically ill patients and found that the saturations were not equivalent. Substitution also produced unacceptable errors in calculated uptake. Their findings do not support a fixed correction that makes central and mixed values interchangeable. 6

Even a true mixed value is a flow-weighted mixture. A high-flow, weakly extracting region can contribute more blood to the sample than a severely hypoperfused region. A threatened organ may contribute little to the final mixture precisely because little blood is leaving it.

A low venous saturation can signal a delivery–demand mismatch requiring investigation. A high saturation can accompany high delivery relative to demand, reduced metabolic activity, abnormal distribution, or impaired extraction. It cannot independently distinguish adequate tissue oxygenation from those alternatives.

Textoris and colleagues reported an association between higher maximum ScvO₂ values and mortality during septic shock. This observational result does not show that the saturation caused harm or that lowering it would help. Nor does a high value diagnose sepsis or prove mitochondrial failure. 7

Sampling location, timing, catheter position, recent interventions, and analytical method influence trends. A change is most useful when comparable measurements fit a coherent change in flow, hemoglobin, arterial oxygenation, demand, or regional perfusion. The saturation is evidence within that explanation, rather than the explanation itself.

34.9 Utilization Requires Delivery, Diffusion, and Functioning Cellular Machinery

Oxygen reaching arterial blood still must enter the relevant microvessels, diffuse to cells, and reach functioning mitochondria. Total transport can be adequate while one of those later steps limits oxidative energy production.

Uneven microvascular perfusion can leave some cells distant from useful oxygen supply even when nearby vessels carry substantial flow. Edema can increase diffusion distances or change the mechanical conditions around small vessels. The importance of either mechanism depends on the organ, regional oxygen pressures, and cellular demand.

Cellular utilization is a separate question. Inhibiting the respiratory chain can reduce oxygen-dependent ATP production despite oxygen availability. Sepsis can also involve altered metabolic demand and mitochondrial function, but a low uptake or high venous saturation does not identify the cellular mechanism by itself.

Brealey and colleagues examined skeletal muscle biopsies from septic patients and found relationships between mitochondrial biochemical findings, ATP concentrations, severity, and outcome. Their human observations support bioenergetic dysfunction as a plausible contributor. They do not demonstrate the state of every organ or show that a specific mitochondrial treatment improves survival. 8

Lactate adds information but does not directly measure failed oxygen utilization. James and colleagues linked epinephrine-stimulated aerobic lactate production to pump activity in isolated rat skeletal muscle. This experimental mechanism does not quantify its contribution in human sepsis; reduced removal and genuine hypoperfusion can coexist. 9

Consequently, a rising lactate with a high ScvO₂ is a pattern requiring explanation, not proof that delivery is normal and mitochondria have failed. Regional ischemia, impaired clearance, high glycolytic activity, altered circulation, and cellular dysfunction remain possible. Direct examination and organ-specific assessment remain necessary.

34.10 Raising Delivery Above an Arbitrary Goal Does Not Establish Benefit

Correcting a limiting determinant can improve oxygen availability. That reasoning does not imply that pushing delivery progressively higher will produce progressively better outcomes. The intervention itself changes physiology and can introduce harm.

Increasing cardiac output with an inotrope can increase global transport, while also increasing myocardial demand or provoking rhythm and vascular effects. Additional fluid may increase flow in a responsive circulation but raise filling pressures and worsen edema. More oxygen can increase content when saturation is low, while producing less content gain once binding sites are nearly full.

Hayes and colleagues randomized critically ill patients to a strategy intended to raise delivery and consumption with dobutamine. Delivery increased, but the strategy did not improve outcome and hospital mortality was higher in the intervention group. The result concerns that studied strategy; it does not mean that an inotrope can never help a patient with inadequate cardiac output. 10

Gattinoni and colleagues compared strategies targeting a normal cardiac index, a supranormal index, or venous oxygen saturation. The broader trial did not show improved morbidity or mortality from pursuing the higher-index or saturation goals. It tested treatment strategies, rather than invalidating oxygen transport as a physiological relationship. 11

The distinction is between identifying a mechanism and selecting an outcome-supported intervention. A low delivery estimate prompts assessment of what is limiting transport. It does not by itself establish a fluid indication, an inotrope requirement, or a target above the patient's physiological need.

Likewise, a higher venous saturation after treatment may reflect improved supply, reduced demand, or altered distribution. The favorable interpretation becomes stronger when perfusion, organ function, and other observations improve together. Pursuing the number after that balance changes can defeat the original purpose.

34.11 Transfusion Trials Test the Clinical Consequence of Increasing Content

Transfusion increases the number of oxygen-carrying cells, but the content equation alone cannot determine when transfusion is beneficial. Clinical trials compare the balance of potential oxygen benefit, transfusion exposure, and adverse effects in defined populations.

In TRICC, Hébert and colleagues compared restrictive and liberal transfusion strategies in critically ill patients after initial treatment. Overall 30-day mortality was not significantly different, and the restrictive group received fewer transfusions. The study does not establish that any hemoglobin concentration is acceptable during uncontrolled hemorrhage or acute coronary ischemia. 12

In TRISS, Holst and colleagues compared thresholds of 7 versus 9 g/dL in septic shock. Ninety-day mortality and reported ischemic events were similar, while the lower-threshold strategy used fewer red-cell units. This challenges the assumption that raising hemoglobin routinely to increase calculated delivery must improve septic-shock outcomes. The trial result remains tied to its population and protocol. 13

Acute myocardial infarction requires distinct evidence. In MINT, Carson and colleagues compared restrictive and liberal strategies in patients with infarction and anemia. The liberal strategy did not achieve a statistically significant reduction in the primary 30-day death-or-infarction outcome, but the findings could not exclude harm from the restrictive strategy. Failure to show superiority is not proof that the strategies are equivalent. 14

These studies teach why a threshold cannot be inferred from oxygen content alone or transferred indiscriminately between populations. Bleeding, ischemia, symptoms, physiological reserve, and the current clinical evidence influence treatment decisions. The thresholds named here describe trial comparisons rather than a prescription for the illustrative patients in this chapter.

The same reasoning applies to the post-transfusion response. Higher hemoglobin confirms a change in one transport determinant. Whether tissue function improves requires assessment of circulation, pulmonary consequences, perfusion, and the underlying illness. A larger calculated DO₂ is an intermediate observation, not the final clinical outcome.

34.12 Integrating an Oxygen Transport Pattern

Consider an illustrative patient with Hb of 8 g/dL, arterial saturation of 97%, PaO₂ of 90 mm Hg, and cardiac output of 4 L/min. Mixed venous saturation is 55% and its PO₂ is 35 mm Hg. These are the same compatible measurements used in the earlier calculations.

Arterial content is approximately 10.68 mL O₂/dL, global delivery approximately 427 mL O₂/min, calculated uptake approximately 187 mL O₂/min, and extraction approximately 43.8%. The arterial saturation looks reassuring, while the content and flow together explain why a substantial fraction of delivered oxygen has been removed.

The measurements do not reveal whether this uptake matches the patient's needs. A sedated, normothermic patient and a patient with shivering or seizures can require different amounts. They also do not show whether a vulnerable organ receives an adequate share of flow.

Investigate the determinants separately. Establish whether anemia reflects dilution, chronic disease, hemolysis, or blood loss. Assess cardiac function, loading, rhythm, regional perfusion, and the reliability of output measurement. Review metabolic demand and possible measurement discordance before attributing a venous value to one cause.

An intervention can then be assessed through its effects. Improved arterial oxygenation, more useful flow, less excessive demand, or correction of a specific oxygen-carriage problem may improve the balance. A change that raises delivery while worsening congestion, myocardial stress, or regional perfusion requires a different interpretation.

The transport equations provide a consistent accounting system. Their value lies in making each quantity explicit and connecting it to the patient's physiology. They complement direct evidence of tissue function and the outcome evidence for treatment rather than replacing either.

References

1. Shimizu S, Enoki Y, Kohzuki H, Ohga Y, Sakata S. Determination of Hüfner's factor and inactive hemoglobins in human, canine, and murine blood. Jpn J Physiol. 1986;36(5):1047–1051. doi:10.2170/jjphysiol.36.1047

2. Buckley RG, Aks SE, Eshom JL, Rydman R, Schaider J, Shayne P. The pulse oximetry gap in carbon monoxide intoxication. Ann Emerg Med. 1994;24(2):252–255. doi:10.1016/S0196-0644(94)70137-7

3. Phang PT, Cunningham KF, Ronco JJ, Wiggs BR, Russell JA. Mathematical coupling explains dependence of oxygen consumption on oxygen delivery in ARDS. Am J Respir Crit Care Med. 1994;150(2):318–323. doi:10.1164/ajrccm.150.2.8049809

4. Cain SM. Oxygen delivery and uptake in dogs during anemic and hypoxic hypoxia. J Appl Physiol Respir Environ Exerc Physiol. 1977;42(2):228–234. doi:10.1152/jappl.1977.42.2.228

5. Weiskopf RB, Viele MK, Feiner J, et al. Human cardiovascular and metabolic response to acute, severe isovolemic anemia. JAMA. 1998;279(3):217–221. doi:10.1001/jama.279.3.217

6. Chawla LS, Zia H, Gutierrez G, Katz NM, Seneff MG, Shah M. Lack of equivalence between central and mixed venous oxygen saturation. Chest. 2004;126(6):1891–1896. doi:10.1378/chest.126.6.1891

7. Textoris J, Fouché L, Wiramus S, et al. High central venous oxygen saturation in the latter stages of septic shock is associated with increased mortality. Crit Care. 2011;15:R176. doi:10.1186/cc10325

8. Brealey D, Brand M, Hargreaves I, et al. Association between mitochondrial dysfunction and severity and outcome of septic shock. Lancet. 2002;360(9328):219–223. doi:10.1016/S0140-6736(02)09459-X

9. James JH, Wagner KR, King JK, et al. Stimulation of both aerobic glycolysis and Na+-K+-ATPase activity in skeletal muscle by epinephrine or amylin. Am J Physiol. 1999;277:E176–E186. Original experimental study.

10. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330(24):1717–1722. doi:10.1056/NEJM199406163302404

11. Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented hemodynamic therapy in critically ill patients. N Engl J Med. 1995;333(16):1025–1032. doi:10.1056/NEJM199510193331601

12. Hébert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med. 1999;340(6):409–417. doi:10.1056/NEJM199902113400601

13. Holst LB, Haase N, Wetterslev J, et al. Lower versus higher hemoglobin threshold for transfusion in septic shock. N Engl J Med. 2014;371(15):1381–1391. doi:10.1056/NEJMoa1406617

14. Carson JL, Brooks MM, Hébert PC, et al. Restrictive or liberal transfusion strategy in myocardial infarction and anemia. N Engl J Med. 2023;389(26):2446–2456. doi:10.1056/NEJMoa2307983