Several key physiologic variables influence a patient’s ability to maintain the Pao₂ in systemic blood (i.e., arterial oxygenation). The alveolar partial pressure of O₂ (PAo₂) provides the driving pressure for the diffusion gradient that moves O₂ passively from exchanging airspaces into pulmonary capillary blood. The PAo₂ varies with the Fio₂, with the minute ventilation, and with the rate of O₂ extraction from the lungs by the pulmonary capillary blood. Minute ventilation, in turn, is affected by the ventilatory drive, airway resistance, lung and chest wall compliance, neuromuscular status, and a host of other factors.

Arterial oxygenation is also affected by ventilation/perfusion ([V with dot above]/[Q with dot above]) matching in the lungs, which determines the degree of contact between fresh gas in the airspaces and pulmonary capillary blood. Normally, hypoxic pulmonary vasoconstriction regulates the distribution of
pulmonary blood flow to match the distribution of ventilation.⁵ Ventilation/perfusion mismatch generates hypoxemia when airspaces do not receive sufficient O₂ in fresh gas to fully saturate the hemoglobin. Perfused regions that receive no ventilation at all send blood with the systemic mixed venous partial pressure of O₂ (P[v with bar above]O₂) to the left heart, generating a true shunt across the lung. Regions with relatively low levels of ventilation compared with perfusion (i.e., low [V with dot above]/[Q with dot above]) allow hemoglobin that is only partially saturated to reach the left heart. Low [V with dot above]/[Q with dot above] areas are caused by clinical conditions that decrease ventilation to an area with high perfusion or by factors that increase perfusion to an area with low ventilation. In any of these circumstances, admixture of blood containing unsaturated hemoglobin in the pulmonary veins and left atrium results in equilibration of O₂ among red cells, decreasing Pao₂ and hemoglobin saturation. Peripheral O₂ consumption exerts an indirect effect on arterial oxygenation by lowering the P[v with bar above]O₂. Usually, a low venous O₂ content only impacts PaO₂ if PAo₂ is marginal or if significant low [V with dot above]/[Q with dot above] units or shunting exist.

The body utilizes PaO₂ to regulate spontaneous ventilatory drive.⁶ Reduction of Pao₂ increases the afferent output from chemoreceptors in the carotid bodies and central nervous system (CNS) to the ventilatory center in the medulla, generating an increase in spontaneous ventilatory rate and depth. Also, insufficient delivery of tissue O₂ leads to anaerobic metabolism and lactic acidemia. A spontaneously breathing patient will hyperventilate to generate a respiratory alkalemia to compensate for the metabolic acidemia. Hyperventilation, therefore, is a major compensatory response to systemic hypoxemia that increases O₂ delivery to airspaces and perhaps expands the functional residual capacity (FRC).

In perioperative patients, hypoxemia is frequently caused by a global reduction of PAo₂ (see Table 8.1). If overall O₂ uptake from airspaces exceeds O₂ delivery by ventilation, the PAo₂ and Pao₂ will decrease. Hemoglobin in all red cells that reach the left heart will be only partially saturated with O₂. A globally decreased PAo₂ sufficient to cause hypoxemia almost always reflects a severe decrease in ventilation, because relatively high concentrations of inspired O₂ are frequently employed during intraoperative anesthetic management. During recovery, supplementation of inspired O₂ concentration usually offsets the impact of reduced minute ventilation, and thereby
decreases the value of pulse oximetry for monitoring ventilation.¹⁸ If hypoventilation is profound, or if ventilation ceases, Pao₂ rapidly declines at a rate that varies with age, body habitus, underlying illness, and the initial PAo₂.¹²

During anesthesia, hypoxemia related to a global reduction of PAo₂ frequently reflects suboptimal technique.¹⁹,²⁰,²¹ Loss of airway patency in spontaneously ventilating patients is a common cause. Partial airway obstruction increases airway resistance, leading to hypoventilation, and sometimes hypercapnia. However, partial obstruction alone does not usually reduce PAo₂ to dangerously low levels, especially with supplemental O₂ administration.²² Greater obstruction will cause a rapid decline in PAo₂; patients with obstructive sleep apnea (OSA) are particularly vulnerable.²³ Significant OSA occurs not only in obese patients but also in adults with average body habitus and in children.²⁴ In patients with OSA, airway dynamics are complex and unpredictable,²⁵ mandating a high level of vigilance for adverse ventilatory events.²⁶

Inadequate positive-pressure ventilation is another common cause of decreased PAo₂ secondary to hypoventilation. Loss of upper airway patency impairs the ability to ventilate with positive pressure. Obstruction of this degree can be caused by soft tissue, space-occupying lesions, foreign bodies, gastric contents, or airway edema secondary to trauma.²⁷,²⁸ Laryngospasm or bronchospasm can also impede ventilation.²⁹ The presence of an airway device, such as an endotracheal tube, does not automatically eliminate airway obstruction as a cause of hypoxemia. Kinking, luminal occlusion by clots or inspissated secretions, inadvertent extubation, or misplacement of a laryngeal mask or oral airway can all cause upper airway obstruction.

Coughing, straining, or chest wall rigidity will impede effective positive-pressure ventilation, as will a poor face mask seal or a loss of pressure in the anesthesia circuit due to leakage. Excessive leakage past an endotracheal cuff, inadequate sealing of a laryngeal mask, or inadvertent tracheal placement of a gastric tube can reduce PAo₂. During difficult intubation or tracheostomy, PAo₂ can fall dramatically, unless ventilation is intermittently supplied. Preintubation hyperventilation with 100% O₂ and attention to the duration of apnea help avoid this problem. Esophageal intubation stops effective positive-pressure ventilation and causes a precipitous decline of PAo₂. In a patient who is dependent on mechanical ventilation, other ominous problems that reduce PAo₂ are failure to activate the mechanical ventilator, failure to set a sufficient rate or tidal volume, failure to supply handbag ventilation, ventilator or bellows malfunction, and unrecognized disconnection.

Spontaneously breathing patients frequently suffer hypoventilation and decreased PAo₂ when medications depress the ventilatory drive. Opioids are potent ventilatory suppressants, as are volatile anesthetics, sedative agents, induction agents, and some antiemetics.³⁰,³¹,³²,³³,³⁴,³⁵,³⁶ The depressant effects of these different agents are synergistic, and CNS depression often generates some upper airway obstruction as well. In patients with OSA, the depressant effects of opioids and sedatives likely accentuate the frequency and depth of baseline desaturations that occur in these patients during sleep.²³,³⁷,³⁸ Ventilatory suppression and hypoxemia is a leading cause of morbidity during monitored anesthetic care and deep sedation.³⁹ Local anesthetic toxicity secondary to inadvertent intravascular injection or uptake can cause profound ventilatory depression and airway obstruction after the administration of regional anesthetics. If local anesthetics reach the intracranial cerebrospinal fluid, ventilatory depression is often immediate and complete. Depression from the central spread of neuraxial opioids is often more insidious in onset.⁴⁰

Impaired ventilatory mechanics can also lead to hypoventilation sufficient to cause hypoxemia. An obvious example is neuromuscular paralysis. However, the loss of intercostal and diaphragmatic muscle strength caused by the spread of spinal or epidural anesthetics to higher levels can seriously impair ventilation, as can reduced lung or chest wall compliance secondary to pneumothorax, increased lung water, or extremes of position during surgery or recovery.⁴¹ Disruption of skeletal or muscular integrity in traumatic conditions, such as flail chest or diaphragmatic rupture, also cause hypoxemia. Children with active or recent upper respiratory infection are prone to breath-holding, severe straining, and arterial desaturations below 90% during induction and recovery. This problem is increasingly likely after intubation and/or airway surgery, or if they have reactive airway disease or secondhand smoke exposure.⁴²

Hypoxemia from global reduction of PAo₂ may result from delivery of a hypoxic mixture of inspired gases (e.g., an inordinately high concentration of nitrous oxide without sufficient O₂) during anesthesia. With the exception of a gas pipe misconnection during maintenance or
renovation, safeguards on contemporary anesthesia machines, such as interlink systems, pin indexed mounts, low pressure shutoffs, and O₂ analyzers in the anesthesia circuit, make delivery of a hypoxic mixture almost impossible. Redundant O₂ sources such as transport and machine backup cylinders provide additional safety.

Rarely, an excessive alveolar partial pressure of another gas may cause hypoxemia. Rapid outpouring of nitrous oxide into alveoli can theoretically displace alveolar O₂ and lower PAo₂. The risk of this “diffusion hypoxia” is greatest immediately after discontinuation of nitrous oxide, especially if a patient is breathing ambient air or hypoventilating. The problem is easily countered by ventilation for a short period with 100% O₂. Volume displacement of O₂ can also occur with severe hypercapnia, again, if a patient is breathing ambient air. Serious respiratory acidemia usually precedes hypoxemia.

Increasing the O₂ content in the FRC with supplemental O₂ helps to safeguard against hypoxemia caused by airway obstruction or hypoventilation. However, supplemental O₂ does not address the underlying cause of hypoxemia and may limit the use of peripheral pulse oximetry as an early predictor of insidious hypoventilation.¹⁸,²²,⁴³ Whether this possibility outweighs the benefits of supplemental O₂ is a matter of individual judgment.

Some surgical patients are more likely to develop hypoxemia from decreased lung volume.⁴⁶ Heavy smoking, obesity, sleep apnea, severe asthma, and chronic obstructive pulmonary disease (COPD) seem to predict an increased risk of postoperative ventilatory events,¹,⁴⁷,⁴⁸,⁴⁹,⁵⁰ but preoperative pulmonary function testing has limited predictive value.⁵¹ The incidence of [V with dot above]/[Q with dot above] mismatching increases with age, because reduced elastic recoil in lung tissue leads to less radial traction on airways and airspaces. Elderly patients likely will experience an intermittent closure of dependent airways at end-expiration. Patients with COPD experience more severe closure that is exacerbated by small reductions in FRC. Obese patients may
suffer large decreases of an already compromised FRC, secondary to the weight of the thoracic fat pad and increased intra-abdominal pressure.⁵²,⁵³,⁵⁴ Prone, lithotomy, jackknife, or Trendelenburg positions are particularly disadvantageous, especially in obese patients. Retraction, packing, peritoneal gas insufflation, and manipulation of organs reduce FRC during upper abdominal and thoracic surgical procedures.²⁴,⁵⁵,⁵⁶,⁵⁷,⁵⁸,⁵⁹ Chest wall or abdominal compression from surgical assistants leaning on the patient or restrictive bandages will impede chest cavity expansion and promote loss of FRC.⁵⁶ Pneumothorax or hemothorax also reduce lung volume⁴¹ and can cause catastrophic hypoxemia in neonates.

Increased lung water or pulmonary edema from overhydration, ventricular dysfunction, or increased capillary permeability interferes with both [V with dot above]/[Q with dot above] matching and O₂ diffusion.⁶⁰ Strong inspiratory efforts against an obstructed airway decrease FRC and promote negative-pressure (postobstructive) pulmonary edema, a frequent cause of transient hypoxemia in healthy patients after even brief periods of laryngospasm or upper airway obstruction. The onset of acute respiratory distress syndrome (ARDS) or transfusion-related acute lung injury will seriously impact [V with dot above]/[Q with dot above] matching.⁶¹,⁶²,⁶³

Extreme [V with dot above]/[Q with dot above] mismatching and hypoxemia occur if a large airspace is completely denied ventilation, as when a bronchus is occluded by a mucus plug or a clot. Right upper lobe atelectasis secondary to a partial right mainstem intubation is a commonly overlooked etiology for loss of lung volume and hypoxemia. Atelectasis can be caused by mechanical occlusion of the upper lobe bronchus or by high gas flow rates across the orifice, leading to decreased pressure in the right upper lobe by the Bernoulli effect. Complete intubation of a mainstem bronchus, inadvertently or during one-lung anesthesia, precludes ventilation to the contralateral lung, generating a large shunt and sometimes profound hypoxemia.⁶⁴ During thoracic procedures, the weight of unsupported mediastinal contents and abdominal pressure on the diaphragm also reduce volume in the dependent, ventilated lung, while gravity and lymphatic obstruction promote interstitial fluid accumulation, thereby accentuating the [V with dot above]/[Q with dot above] mismatch. This “down lung syndrome” can appear as unilateral pulmonary edema on a chest radiograph.

Tracheal intubation eliminates the resistance to gas flow past the vocal cords that can help maintain lung volume during spontaneous exhalation. An intubated, spontaneously breathing patient cannot generate a significant positive pressure in the airways. Allowing this patient to ventilate at ambient pressure for prolonged periods can cause a gradual reduction in FRC and a worsening of [V with dot above]/[Q with dot above] matching. Although healthy, slender patients will often tolerate short periods without positive pressure, it is prudent to allow such patients to exhale against a slight CPAP. Also, loss of gas volume during tracheal suctioning can cause serious hypoxemia from loss of lung volume and physical removal of O₂ from the FRC.⁶⁵,⁶⁶

Decreases in FRC or regional lung volume that occur during surgery persist and even worsen postoperatively. In the recovery phase, conservative measures oriented toward restoration of lung volume often improve oxygenation. When possible, patients should recover in a semisitting Fowlers position to reduce abdominal pressure on the diaphragm. Provision of sufficient analgesia reduces pain associated with deep ventilation and improves maintenance of dependent lung volume, especially with upper abdominal or chest wall incisions.⁶⁷ Deep breaths, cough, chest physiotherapy, and incentive spirometry may help maintain the FRC, mobilize secretions, and accustom a patient to incisional discomfort. However, the efficacy of these interventions is being debated.⁶⁸,⁶⁹ CPAP delivered by facemask helps to offset more significant or persistent reduction of FRC. If hypoxemia is severe, or acceptance of mask CPAP is poor, tracheal intubation may be required. The requirement for positive-pressure ventilation is assessed independently, considering the work of breathing, PACO₂, and arterial pH (pHa). Usually, 5 to 10 cm H₂O of CPAP or PEEP improves Pao₂. Airway pressure >10 cm H₂O can impede venous return, cause hypotension, and may increase the risk of barotrauma or increased intracranial pressure.⁴¹ However, patients with ARDS or pulmonary contusion may require higher pressures to improve oxygenation. If these measures do not improve Pao₂, the cause of hypoxemia should be reevaluated.

Poor distribution of pulmonary blood flow also causes [V with dot above]/[Q with dot above] mismatching and hypoxemia. Flow distribution is primarily determined by hydrodynamic factors (pulmonary arterial and venous pressures, pulmonary vascular resistance), which are, in turn, affected by gravity, cardiac dynamics, vascular competence, airway pressure, and lung volume. Patient positioning affects oxygenation if gravity forces blood flow to areas with reduced ventilation; for example, placing a poorly ventilated lung in a dependent position can reduce Pao₂. Intraoperative or postoperative changes in pulmonary artery pressure, airway pressure, and lung volume have complex effects on blood flow distribution that can adversely affect [V with dot above]/[Q with dot above] matching. Inhalational anesthetics, vasodilators, and sympathomimetic agents directly affect vascular tone and hypoxic pulmonary vasoconstriction, partially explaining larger alveolar/arterial O₂ gradients during and after general anesthesia. Patients with cirrhosis of the liver exhibit disordered blood flow distribution and [V with dot above]/[Q with dot above] mismatching caused by circulating humoral substances resulting from abnormal hepatic metabolism. Circulating endotoxin also impairs hypoxic pulmonary vasoconstriction, contributing to hypoxemia in septic patients.

Few interventions are practical to improve [V with dot above]/[Q with dot above] matching by managing pulmonary blood flow. When possible, avoid placing atelectatic or diseased lung tissue in a dependent position. Placing poorly ventilated parenchyma in a nondependent, “up” position may improve [V with dot above]/[Q with dot above] matching, but could promote drainage of purulent material from diseased lung segments into unaffected areas. Avoiding β-mimetic or vasodilatory medications may improve PaO₂, but the benefits from using the medication almost always outweigh the drawback of impaired hypoxic pulmonary vasoconstriction. Maintaining pulmonary artery and airway pressures within acceptable ranges likely optimizes any yield from hemodynamic interventions aimed at improving [V with dot above]/[Q with dot above] matching.