Key Points

  • No hemodynamic monitoring device will improve patient outcome unless coupled to a treatment, which itself improves outcome.
  • Low venous oxygen saturations need not mean circulatory shock but do imply circulatory stress, as they may occur in the setting of hypoxemia, anemia, exercise, as well as circulatory shock.
  • There is no "normal" cardiac output, only one that is adequate or inadequate to meet the metabolic demands of the body. Thus, targeting a specific cardiac output value without reference to metabolic need, or oxygen-carrying capacity of the blood, is dangerous.
  • Cardiac output is estimated, not measured, by all devices routinely used in bedside monitoring (though we shall call it measured in this text).
  • Cardiac output estimates using arterial pulse pressure contour analysis cannot be interchanged among devices and all suffer to a greater or lesser extent by changes of peripheral vasomotor tone commonly seen in the critically ill.
  • Since metabolic demands can vary rapidly, continuous or frequent measures of cardiac output are preferred to single or widely spaced individual measures.
  • Integrating several physiologic variables in the assessment of the adequacy of the circulation usually gives a clearer picture than just looking at one variable.
  • Integrating cardiac output with other measures, like venous oxygen saturation, can be very helpful in defining the adequacy of blood flow.

Clinical Judgement of Hypoperfusion

Tissue hypoperfusion is a clinical syndrome, thus the presentation depends on which organ(s)/organ system(s) are being hypoperfused.

Table 1 Common Clinical Presentations of Tissue Hypoperfusion
Brain/CNS altered mental status, confusion
Pulmonary system dyspnea on exertion
Gastrointestinal tract slowed bowel function, abdominal discomfort, nausea/vomiting, anorexia
Renal system decrease urine output, increased serum creatinine
Hepatic system increased transaminases
Extremities/constitutional cool extremities, poor capillary refill, general fatigue/malaise
SvO2 in the case of normal CaO2 and VO2, SvO2 would decrease
Serum lactate level elevated serum lactate level

Note that some factors other than tissue hypoperfusion are able to cause the "hypoperfusion"-like presentation(s) which are similar as the one discussed above. For example, in the sepsis, the elevated levels of lactic acid might be caused by reasons other than hypoperfusion. In another case, the low urine output of a patient might be caused by the patient's ESRD per se rather than the low perfusion of kidneys. Also, if the patient had hypoglycemia or severe hyponatremia (which cause the neurons to edema), he or she would definitely lose the consciousness (differential diagnosis? or differential clinical judgement).

Tissue Oxygenation

Although mean artieral pressure (MAP) is a primary determinant of organ perfusion, normotension can coexist with circulatory shock.

Since metabolic demand of tissues varies by external (exercise) and internal (basal metabolism, digestion, fever) stresses, there is no "normal" cardiac ouput that the bedside caregiver can target and be assured of perfusion adequacy. Cardiac output is either adequate or inadequate to meet the metabolic demands of the body. Thus, although measures of cardiac output are important, their absolute values are relevant only in the extremes and when targeting specific clinical conditions, such as preoptimization therapy.

How then does one know that circulatory sufficiency is present or that circulatory shock exists? Clearly, since arterial pressure is the primary determinant of organ blood flow, systemic hypotension (i.e., mean arterial pressure <60 mm Hg) must result in tissue hypotension. Organ perfusion pressure can be approximated as mean arterial pressure (MAP) relative to tissue or outflow pressure. But, if intracranial pressure or intra-abdominal pressure increases, then estimating cerebral or splanchnic perfusion pressure using MAP alone will grossly overestimate organ perfusion pressure. In addition, baroreceptors in the carotid body and aortic arch increase vasomotor tone to keep cerebral perfusion constant if flow decreases, and the associated increased systemic sympathetic tone alters local vasomotor tone to redistribute blood flow away from more efficient O2 extracting tissues to sustain MAP and global O2 consumption (VO2) in the setting of inappropriately decreasing DO2. Thus, although systemic hypotension is a medical emergency and reflects severe circulatory shock, the absence of systemic hypotension does not ensure that all tissues are being adequately perfused.

  • Arterial pulse oximetry: SaO2
  • Venous Oximetry: ScvO2, SvO2
  • Tissue Oximetry
  • StO2 vascular occlusion test

Arterial Pulse Oximetry


Arterial blood O2 saturation (SaO2) can be estimated quite accurately at the bedside using pulse oximetry. Routinely, pulse oximeters are placed on a finger for convenience sake. However, if no pulse is sensed, then the readings are meaningless. Such finger pulselessness can be seen with peripheral vasoconstriction associated with hypothermia, circulatory shock, or vasospasm. Central pulse oximetry using transmission technology can be applied to the ear or bridge of the nose and reflectance oximetry can be applied to the forehead, all of which tend to retain pulsatility if central pulsatile flow is present. Similarly, during cardiopulmonary bypass when arterial flow is constant, pulse oximetry is inaccurate. The primary important functions of SaO2 are summarized below.

  • SaO2 is routinely used to identify hypoxemia. Hypoxemia is usually defined as an SaO2 of <90% (PaO2 of <60 mm Hg).
  • SaO2 is also used to identify the causes of hypoxemia. The most common causes of hypoxemia are ventilation-perfusion (V/Q) mismatch (we will discuss this topic in the clinical art of respiratory medicine) and shunt. With V/Q mismatch alveolar hypoxia ocurs in lung regions with increased flow relative to ventilation, such that the high blood flow rapidly depletes alveolar O2 before the next breath can refresh it. Accordingly, this process readily lends itself to improve oxygenation by increasing FiO2 and minimizing regional alveolar hypoxia. Collapsed or flooded lung units will not alter their alveolar O2 levels by this maneuver and are said to be refractory to increase in FiO2. Accordingly, by measuring the SaO2 response to slight increases in FiO2 one can separate V/Q mismatch (shuntlike states) from shunt (absolute intrapulmonary shunt, anatomical intracardiac shunts, alveolar flooding, atelectasis [collapse]). One merely measures SaO2 while switching from room air FiO2 of 0.21 to 2 to 4 L/min nasal cannula (FiO2 ~0.3). Importantly, atelectatic lung units (collapse alveoli) should be recruitable by lung expansion whereas flooded lung units and anatomical shunts should not. Thus, by performing sustained deep inspirations and having the patient sit up and take deep breaths one should be able to separate easily recruitable atelectasis from shunt caused by anatomy and alveolar flooding. Sitting up and taking deep breaths is a form of exercise that may increaes O2 extraction by the tissues, thus decreasing SvO2. The patient with atelectasis will increase alveolar ventilation increasing SaO2 despite the decrease in SvO2, whereas the patient with unrecruitable shunt will realize a fall in SaO2 as the shunted blood will carry the lower SvO2 to the arterial side. Summary: SaO2 is used to identify V/Q mismatch and shunt, atelectasis and anatomical/flooding shunt, respectively.
  • Detection of volume responsiveness. Recent interest in the clinical applications of heart-lung interactions has centered on the effect of positive-pressure ventilation on venous return and subsequently cardiac output. In those subjects who are volume responsive, arterial pulse pressure, as a measure of left ventricular (LV) stroke volume, phasically decreases in phase with expiration, the magnitude of which is proportional to their volume responsiveness. Since the pulse oximeter's plethysmographic waveform is a manifestation of the arterial pulse pressure, if pulse pressure varies from beat-to-beat so will the plethysmographic deflection, which can be quantified. Several groups have documented that the maximal variations in pulse oximeter's plethysmographic waveform during positive-pressure ventilation covaries with arterial pulse pressure variation and can be used in a similar fashion to identify those subjects who are volume responsive.

PS: Intrapulmonary shunt fraction is increased in the following situations:

  • When the small airways are occluded; e.g., asthma
  • When the alveoli are filled with fluid; e.g., pulmonary edema, pneumonia
  • When the alveoli collapse; e.g., atelectasis
  • When capillary flow is excessive; e.g., in nonembolized regions of the lung in pulmonary embolism

Venous Oximetery


SvO2 is the gold standard for assessing circulatory stress. A low SvO2 defines increased circulatory stress, which may or may not be pathological.

To the extent that SaO2 and hemoglobin concentration result in an adequate arterial O2 content (CaO2), then ScvO2 and SvO2 levels can be taken to reflect the adeuqacy of the circulation to meet the metabolic demands of the tissues. This is the truth. But attention must be paid that one needs to examine the determinants of DO2, global oxygen consumption (VO2), and effective O2 extraction by the tissues before using ScvO2, or SvO2 as markers of circulatory sufficiency.

Since VO2 must equal cardiac output times the difference in CaO2 and mixed venous O2 content (CvO2) (VO2 = CO x [CaO2 – CvO2]),if CaO2 remains relatively constant then CvOwill vary in proportion to cardiac output (CvO2 = CaO2 – VO2 / CO). Since the amount of O2 dissolved in the plasma is very small, the primary factor determining changes in CvO2 will be SvO2. Thus, SvO2 correlates well with the O2 supply-to-demand ratio.

Several relevant conditions may limit this simple application of SvO2 in assessing circulatory sufficiency and cardiac output (Table 32-1). If VO2 were to increase (as occurs with exercise), hemoglobin-carrying capacity to decrease (as occurs with anemia, hemoglobinopathies, and severe hemorrhage), or SaO2 to decrease (as occurs with hypoxic respiratory failure), then for the same cardiac output, SvO2 would also decrease. Similarly, if more blood flows through nonmetabolically extracting tissues as occurs with intravascular shunts, or mitochondrial dysfunction limits O2 uptake by tissues, then SvO2 will increase for a constant cardiac output and VO2 even though circulatory stress exists and may cause organ dysfunction.

Table 32-1 Limitations to the Use of SvO2 to Trend Circulatory Sufficiency
Independent events that decrease SvO2 independent of cardiac output
Event Process
Exercise Increased VO2
Anemia Decreased O2-carrying capacity
Hypoxemia Decreased arterial O2 content
Independent events that increase SvO2 independent of cardiac output
Event Process
Sepsis Microvascular shunting
End-stage hepatic failure Macrovascular shunting
Carbon monoxide poisoning Mitochondrial respiratory chain inhibition


SvO2 and ScvO2 covary in the extremes but may change in opposite directions as conditions change.

ScvO2 threshold values to define circulatory stress are only relevant if low (a high ScvO2 is nondiagnostic).

ScvO2 does not sample true mixed venous blood and most vena cacal blood flow is laminar, thus if the tip is in one of these laminar flow sites it will preferentially report a highly localized venous drainage site O2 saturation. Clearly, the potential exists for spurious estimates of SvO2. Most central venous catheters are inserted from internal jugular or subclavian venous sites with their distal tip residing in the superior vena cava, usually about 5 cm above the right atrium. Thus, even if measuring a mixed venous sample of blood at that site, ScvO2 reflects upper body venous blood while ignoring venous drainage from the lower body. Accordingly, ScvO2 is usually higher than SvO2 by 2% to 3% in a sedated resting patient because cerebral O2 consumption is minimal and always sustained above other organs.

Tissue Oximetry

Tissue O2 saturation (StO2) varies little until severe tissue hypoperfusion occurs.

StO2 coupled to a VOT allows one to diagnose circulatory stress before hypotension develops.

The most currently used technique to measure peripheral tissue O2 saturation (StO2) is near-infrared spectroscopy (NIRS). NIRS is a noninvasive technique based in the differential absorption properties of oxygenated and deoxygenated hemoglobin to assess the muscle oxygenation. Although there is a good correlation between the absolute StO2 value and some other cardiovascular indexes, the capacity of the baseline StO2 values to identify impending cardiovascular insufficiency is limited (sensitivity, 78%; specificity, 39%).

screen-shot-2016-10-24-at-12-52-46-pmHowever, the addition of a dynamic vascular occlusion test (VOT) that induces a controlled local ischemic challenge with subsequent release has been shown to markedly improve and expand the predictive ability of StO2 to identify tissue hypoperfusion. The VOT StO2 response derives from the functional hemodynamic monitoring concept, in which the response of a system to a predetermined stress is the monitored variable. The rate of DeO2 is a function of local metabolic rate and blood flow distribution. If metabolic rate is increased by muscle contraction, the DeO2 slope increases, whereas in the setting of altered blood flow distribution the rate of global O2 delivery is decreased. Sepsis decreases the DeO2. The ReO2 slope is dependent on how low StO2 is at the time of release, being less steep if StO2 is above 40% than if the recovery starts at 30%, suggesting that the magnitude of the ischemic signal determines maximal local vasodilation. This dynamic technique has been used to assess circulatory sufficiency in patients with trauma, sepsis and during weaning from mechanical ventilation.

Cardiac Output

There is no "normal" cardiac output (QT).

QT is either adequate or inadequate

Other measures besides QT define adequacy

Shock reflects an inadequate DO2 to meet the body's metabolic demand and cardiac output is a primary determinant of DO2. Indeed, except for extreme hypoxemia and anemia, most of the increase in DO2 that occurs with resuscitation and normal biological adaptation is due to increasing cardiac output. Since cardiac output should vary to match metabolic demands, there is no "normal" cardiac output. Cardiac output is merely adequate or inadequate to meet the metabolic demands of the body. Measures other than cardiac output need to be made to ascertain if the measured cardiac output values are adequate to meet metabolic demands. The two most common catheter-related methods of estimating cardiac output are indicator dilution and arterial pulse contour analysis.

Indicator Dilution

The principle of indicator dilution cardiac output measures is that if a small amount of a measurable substance (indicator) is ejected upstream of a sampling site and then thoroughly mixed with the passing blood then measured continuously downstream, the area under the time-concentration curve will be inversely proportional to flow based on the Stewart-Hamilton equation. The greater the indicator level, the slower the flow, and the lower the indicator level, the higher the flow. The most commonly used indicator is temperature (hot or cold) because it is readily available and indwelling thermistors can be made to be highly accurate.

Arterial Pulse Contour Analysis

screen-shot-2016-10-24-at-8-44-47-pmThe primary determinants of the arterial pulse pressure are LV stroke volume and central arterial compliance (pulse pressure ≈ SV / C). Compliance is a function of size, age, sex, and physiological inputs, like sympathetic tone, hypoglycemia, temperature, and autonomic responsiveness of the vasculature. Hamilton and Remington explored this interaction over 50 years ago developing the overall approach used by most of the companies who attempt to report cardiac output from the arterial pulse. The main advantage of these arterial pressure-based cardiac output monitoring systems over indicator dilution measurements is their less invasive nature.

However, since all these devices presume a fixed relation between pressure propagation alnog the vascular three and LV stroke volume, if vascular elastance (reciprocal of compliance) changes, then these assumptions may become invalid. Thus, a major weakness of any pulse contour device is the potential for artificial drift in reported values if major changes in arterial compliance occur.