The primary goal of mechanical ventilation is to supply oxygen (O₂) and to remove carbon dioxide (CO₂). Thus, 2 interrelated processes must occur: ventilation, the movement of air between the environment and the alveoli; and gas exchange, the transfer of O₂ and CO₂ between the alveolar gas and the mixed venous blood entering the lungs. Another goal includes prevention of anesthesia-related adverse effects, mainly atelectasis, as well as prevention of ventilator-induced lung injury (VILI).

General anesthesia induces significant changes to the basic lung function. The change from the upright position to the supine position reduces functional residual capacity (FRC) by 15% to 20%. Loss of muscle tone induces a change in the balance of elastic recoil and chest compliance leading to reduced lung volumes [5], increased shunting across the lung, and a further reduction in FRC [6]. The change in lung compliance and the reduction in FRC promote airway closure and induce compression atelectasis. Microatelectasis can be detected in up to 90% of patients undergoing general anesthesia and can involve up to 5% to 20% of the total gas exchange area [5]. Also, absorption atelectasis is common when high inspired O₂ concentrations (Fio₂) are applied [7]. It occurs when less gas enters the alveolus than is removed by uptake into the blood. In the perioperative period, compression and absorption atelectasis are the 2 major mechanisms for the occurrence of atelectasis.

General anesthesia is also associated with an increased alveolar dead space. The reduction in FRC moves all lung areas down on their pressure-volume curve. Therefore, in conditions of general anesthesia and mechanical ventilation, apical alveoli located on the linear part of this curve are best ventilated but less perfused for gravitational reasons. Increased alveolar pressure and reduced pulmonary artery pressure may further decrease apical perfusion. These changes in alveolar dead space reduce the efficacy of alveolar ventilation in eliminating CO₂ and increase the arterial-to-end-tidal CO₂ gradient.

Mechanical ventilation leads to ventilation-perfusion (V/Q) mismatch, more pronounced in the posterodorsal areas, and causes changes in the function of surfactant [8]. In addition, volatile anesthetics reduce hypoxic pulmonary vasoconstriction (HPV), also known as the Euler-Liljestrand reflex. Venous admixture increases and the alveolar-arterial O₂ pressure gradient increases. Under 2 minimum alveolar concentrations of isoflurane, HPV is reduced by almost 50% [5].

In summary, the right-to-left shunting increases from 1% to 5% observed in the healthy, upright, spontaneously ventilating adult to 15% to 20% in the patient who is supine, anesthetized, and ventilated. Therefore, oxygenation is moderately compromised during general anesthesia, despite increase in Fio₂ [5]. Also, atelectasis occurring during general anesthesia may promote injury to alveoli resulting from recurrent shear stress caused by opening and closing of the atelectatic areas (ie, atelectrauma).

In the intraoperative period, the traditional mode of ventilation has been a volume-control mode with moderately high tidal volumes (10–12 mL/kg), low respiratory rates (8–10 breaths/min), I/E ratio (1:2), slightly increased Fio₂ (30%–50%), and zero end-expiratory pressure. The reasons for using twice the physiologic tidal volumes stem from the concern that small tidal volumes can lead to loss of lung volume and hypoxemia caused by right-to-left shunting [9,10]. Volume-control ventilation (VCV) modes are preferred, because they provide fixed tidal volume, even during changes in compliance of chest wall and lungs that can occur during surgery (eg, caused by changes in patient’s position, external pressure on the thorax, and changing degrees of muscle relaxation).

The traditional goal of ventilation is to maintain mild hypocapnia (ie, end-tidal CO₂ [ETCO₂] around 30 to 35 mm Hg). This goal is probably intended to achieve an apneic threshold, defined as the highest arterial CO₂ tension at which a subject remains apneic, which is approximately 4 or 5 mm Hg less than resting arterial CO₂ tension achieved during spontaneous ventilation. However, the benefits of maintaining mild intraoperative hypocapnia remain controversial. Recent evidence suggests that mild hypercapnia (ie, ETCO₂ values around 40–50 mm Hg) can improve tissue oxygenation through improved tissue perfusion and oxygenation resulting from increased cardiac output and vasodilatation as well as increased O₂ off-loading from the shift of the oxyhemoglobin dissociation curve to the right [11,13]. In addition, mild hypercapnia and mild respiratory acidosis attenuate or dampen the inflammation caused by ischemia-reperfusion, established bacterial pneumonia [14], and endotoxin-induced lung injury. However, hypercapnia and acidosis may increase intracranial pressure as well as increase pulmonary vascular resistance, and thereby increase right ventricular workload [15]. Therefore, it should be avoided in select patient populations.

One of the major concerns of mechanical ventilation is development of VILI, which is caused by volutrauma and barotrauma resulting in an enhanced systemic inflammatory response with worsening oxygenation. One hour of mechanical ventilation alone without surgery in healthy patients did not lead to an increase in measured inflammatory parameters [16]. Choi and colleagues [17] compared the use of tidal volumes of 12 mL/kg without positive end-expiratory pressure (PEEP) with tidal volumes of 6 mL/kg with PEEP of 10 cm H₂O in patients undergoing major abdominal surgery. Bronchiolar lavage was performed before and after 5 h of mechanical ventilation. Lavage fluid from the high tidal volume group showed a pattern of leakage of plasma into the alveoli consistent with alveolar lung injury. This study suggests that, even in patients with no lung disease, the use of large tidal volumes without PEEP causes systemic inflammation and lung injury. The severity of this injury seems to be directly related to the duration of mechanical ventilation and usually remains subclinical.

An observational study found a 25% incidence of ALI/ARDS within 5 days of mechanical ventilation [18]. The development of lung injury was associated with the initial ventilator settings and correlated with an odds ratio of 1.6 for peak airway pressures greater than 30 cm H₂O and an odds ratio of 1.3 for each milliliter of tidal volume exceeding 6 mL/kg predicted body weight. Perioperative ALI becomes clinically important when injurious ventilation patterns are used in patients who have other concomitant lung injuries, such as pulmonary resection, cardiopulmonary bypass, or transfusion related lung injury (see later discussion).

Edema, secretions, and infiltrations all contribute to worsening pulmonary dynamics measured as a decreased static and dynamic compliance, which increase peak airway pressure, regardless of the mode of ventilation. However, peak inspiratory pressure (PIP) reflects the proximal airway pressure and not the alveolar pressure. Therefore, in the setting of reduced compliance or reduced thoracic expansion, high pressures may not induce lung injury [19]. A better approximation of end-tidal alveolar pressure is obtained by using the plateau pressure in a VCV mode. A plateau pressure of greater than 35 cm H₂O has also been implicated in the development of VILI [20].

Overall, volume rather than pressure seems to be the culprit, coining the term volutrauma instead of the traditional expression of barotrauma. Taken together, all studies in healthy patients, as well as in patients at risk, show that ventilation with large tidal volumes for a longer time period may represent a primary hit or an additional hit in cases of extrapulmonary injury [17,21–25]. In patients with pulmonary injury, even shorter periods of harmful ventilation can cause exaggerated injuries to the lung, possibly triggering a systemic inflammatory response.

VILI encompasses 3 different entities: overinflation of alveoli, high-permeability type pulmonary edema (presents as a protein-rich expression of lung fluid), and lung inflammation, also called biotrauma. During ALI, regional compliance heterogeneity creates a functional baby lung. Normal or large tidal volumes delivered to a reduced number of ventilated alveoli can lead to further distention of lung areas that are already overdistended. The alveolar endothelial membrane is altered by mechanical distortion, increased transmural pressure, surfactant inactivation, and the immigration of inflammatory cells. Secondary changes include emphysemalike lesions, lung cysts, and bronchiectasis in nondependant and caudal lung regions [26].

In the intensive care unit (ICU) setting, ventilation with lung volumes of 6 to 8 mL/kg ideal body weight (IBW), as well as peak inspiratory plateau pressures of less than 30 cm H₂O, is now considered the standard of care [27,28]. The ARDSnet study has demonstrated a reduction in mortality from 40% to 31% when a population of patients with ARDS/ALI was ventilated with half the tidal volume (6 vs 12 mL/kg, IBW) [29]. This approach reduces volutrauma caused by high tidal volumes and barotrauma secondary to shear stress related to high airway pressures. Furthermore, biotrauma, a combination of lung and distant organ injury caused by systemic inflammatory response, is reduced [30]. Lower intraoperative tidal volumes of 6.7 mL/kg versus 8.3 mL/kg IBW and avoidance of excessive fluid infusion resulted in reduced respiratory failure after pneumonectomy [31]. Similarly, protective ventilation strategies reduce lung injury and time on the ventilator after esophageal cancer surgery [32].

Intraoperative mechanical ventilation should be similar to that in the ICU setting because of the difficulty in diagnosing ALI/ARDS and the potential for multiple hits caused by intraoperative volutrauma [33]. Therefore, lung protective patterns of intraoperative mechanical ventilation using physiologic tidal volumes and appropriate PEEP should be considered.

Concerns of atelectasis from general anesthesia as well as the use of low tidal volumes have generated a considerable interest in maneuvers designed to prevent and recruit atelectatic lung regions. PEEP prevents alveolar collapse and maintains end-expiratory lung volumes recruited during inspiration. However, the appropriate level of PEEP is controversial [34]. Attempts to optimize alveolar recruitment by increasing PEEP may be either poorly effective [35] or deleterious because of overinflation of more compliant lung regions [36]. This overinflation counteracts the beneficial effects from low tidal volumes and limited airway pressure ventilation [37].

A recent meta-analysis concluded that, in patients with ALI/ARDS, lower tidal volumes reduced hospital mortality regardless of low or high levels of PEEP. However, higher PEEP levels required 50% fewer interventions for rescue therapy in patients with severe hypoxemia, and reduced mortality in this population [38,40]. The appropriate PEEP values in patients with noncompliant thoraces or increased abdominal pressure are not known, because this patient population was excluded in most of the studies.

Overall, selection of PEEP level should take into consideration lung morphology and regional distribution [41]. Higher levels of PEEP may be appropriate when the loss of aeration is diffuse and involves all lung regions, commonly referred to as white lung [41]. In addition, higher PEEP levels can be beneficial without adding harm from overdistention, provided plateau pressures were kept to less than 30 cm H₂O [34]. Moderate PEEP levels of around 10 cm H₂O should maintain a balance between aeration and overinflation when the loss of aeration is localized and focally distributed (in which the lung behaves as several compartments).

Application of PEEP alone may not always reverse atelectasis and improve arterial oxygenation [42]. The application of 5 cm H₂O of PEEP without prior recruitment was similar to application of no PEEP [43]. In contrast, PEEP of 10 cm H₂O without recruitment consistently reopened some collapsed lung tissue [44]. In 1963, Bendixen and colleagues [45] suggested that periodic deep breaths prevent progressive atelectasis and intrapulmonary shunting. To reexpand the atelectatic lung, it may be necessary to use a vital capacity maneuver with higher inflation pressures, referred to as a recruitment or Lachmann maneuver [46]. Inspiratory pressures of 30 cm H₂O are required to reexpand half of the anesthesia-induced atelectatic lung, but PIPs of up to 40 cm H₂O may be needed to fully reverse anesthesia-induced collapse of healthy lungs, and even higher pressures may be required if the patient is grossly obese [47]. The duration of recruitment maneuver should generally be at least 7 to 8 seconds [48]. Most of the reexpanded lung tissue should remain inflated for about 40 minutes [49]. During a recruitment maneuver, arterial blood pressure should be closely monitored because of its potential to reduce preload and induce hypotension. In a pig model, recruitment maneuvers, applied every 6 hours, had no negative effect on alveolocapillary membrane integrity as measured by extravascular lung water, pulmonary clearance, and light microscopy [50].

Recruitment maneuvers should not be used routinely, and should be restricted for the rapid reversal of loss of aeration from disconnections from the ventilator, for example, to perform tracheal suctioning [51]. The primary aim of lung recruitment should be to achieve arterial saturation greater than or equal to 90% at Fio₂ less than or equal to 0.6. In cases of severe hypoxemia, adjunctive interventions (eg, prone positioning, inhaled nitric oxide, or extracorporeal oxygenation) may be considered.

Ideally, lung recruitability should be assessed by dynamic lung imaging techniques, which would allow the determination of the best physiologic PEEP. In addition, the use of transpulmonary pressure would allow the assessment of the detrimental effects of low tidal volumes. Until this becomes widely available, optimal ventilator settings would include tidal volumes of 6 mL/kg IBW and the maximum PEEP based on the upper limit of airway pressure. In the perioperative setting, maintaining a PEEP of 5 cm H₂O from preoxygenation to the tracheal extubation should reduce atelectasis [52]. Intermittent recruitment maneuvers may be necessary, which should be followed by sufficient PEEP to maintain alveolar unit expansion [49]. Vital capacity maneuvers without subsequently applied PEEP proved to be ineffective for increasing lung volumes and increasing gas exchange [53].

In summary, both PEEP and recruitment maneuvers complement each other and are considered components of the open lung concept. However, recruitment maneuvers during general anesthesia cannot at this time be generally recommended for routine use, and may be best applied in certain circumstances such as ventilation with high Fio₂ and after cardiopulmonary bypass.

Optimal respiratory rate implies selecting the best compromise between 2 opposing goals: CO₂ elimination and avoidance of the development of intrinsic (auto) PEEP. With the use of lower tidal volumes, higher respiratory rates may be necessary to maintain adequate CO₂ levels. Higher respiratory rates shorten expiratory times, which may result in intrinsic PEEP. In general, respiratory rates can be increased to 20 to 30 breaths/min without generating significant intrinsic PEEP in healthy patients. However, if higher respiratory rates are used, it may be necessary to monitor inspiratory and expiratory flows. If the end-expiratory flow remains zero, respiratory rates can be increased without the risk of intrinsic PEEP formation [54].

Intraoperative Fio₂ is frequently increased to compensate for the gas exchange impairment related to anesthesia. However, use of excessive O₂ concentrations is potentially deleterious for the lungs and other organs. Application of 100% O₂ for 24 hours has been reported to cause tracheobronchial irritation and pulmonary toxicity leading to irreversible fibrosis [55]. In the intraoperative period, even short-term high Fio₂ may result in absorption atelectasis, which can exaggerate the effects of other stressors to the lung such as stretch injury secondary to inappropriately high tidal volumes or airway pressures [56,57]. High Fio₂ has been linked to poor regulation of blood glucose levels [58] and increased systemic vascular tone [59].

Fio₂ of 0.8 during surgery and until 2 hours after major colorectal surgery correlated with reduced surgical site infection (SSI) rate and postoperative complications [60]. The incidence of postoperative nausea and vomiting (PONV) may also be reduced by higher Fio₂ [61]. However, these early results have not been confirmed in subsequent studies involving other surgeries [62,63]. A recent meta-analysis did not support a beneficial effect of higher Fio₂ on PONV [64]. Another meta-analysis found that the use of higher Fio₂ resulted in an absolute SSI risk reduction of 3% and relative risk reduction of 25% [65]. However, other measures for improving tissue oxygenation, such as temperature control [66], fluid and sympathetic tone management [67], and avoidance of hypocapnia, may prove to be equally important in reducing infections and improving wound healing [68]. When using high Fio₂, lung protective ventilation strategies should be implemented to avoid complications of mechanical ventilation and O₂ toxicity [57,69]. Therefore, a low-O₂ strategy may preserve lung function and improve arterial oxygenation better than high-O₂ therapy during elective surgery [70]. The clinical relevance of these observations currently awaits further confirmation.