The acutely injured lung comprises a heterogeneous environment of aerated and nonaerated lung (Fig. 1) 8, the nonaerated lung consisting of collapsed or consolidated alveoli. Positive pressure ventilation generates tensions at the boundaries between aerated and nonaerated lung, and repeated high-pressure inflations may cause damaging shearing forces at these junctional interfaces 9. Another stress induced by positive pressure ventilation is the cyclic opening and closing of alveoli, in the presence of inadequate positive end-expiratory pressure (PEEP) to maintain alveolar patency through the respiratory cycle 10. These mechanical stresses may have a number of effects, including epithelial and endothelial damage, cellular inflammatory damage, and release of cytokines 511.

Pressure-limited ventilatory strategies have been introduced to limit these ventilator-induced stresses 1213, but they do not address the primary problem of inhomogeneity of the aeration of the lung. In fact, reduced tidal volumes are probably responsible for increasing alveolar derecruitment 14. From a pathophysiologic perspective, attempts to open the nonaerated lung units seem appropriate, bearing in mind that only collapsed but not consolidated alveoli are likely to respond 15. Recruitment appears to be a continuous process that occurs throughout the pressure–volume curve and not all lung units are recruitable at safe pressures 16. In general, lung units can be kept open by airway pressures that are lower than those required to open them 16, leading to the concept of recruitment using periodic higher pressure maneuvers with moderate levels of PEEP to maintain alveolar patency. The 'open' lung is ventilated on the expiratory limb of the pressure–volume curve, rather than the underinflated lung on the inspiratory portion of the curve (Fig. 2).

In animal models of acute lung injury lung recruitment maneuvers have been demonstrated to improve oxygenation and to open nonaerated lung 417. Recruitment maneuvers may have differential effects depending on the mechanism of lung injury 18. Because of the increased atelectasis, they appear to be more effective in situations in which a low PEEP is being used, and the benefit is far less in a high PEEP model 418. It was recently demonstrated that recruitment strategies may prevent microvascular leak and right ventricular dysfunction in rats without pre-existing lung injury undergoing pressure limited ventilation 19.

The findings of clinical studies of recruitment maneuvers in patients with ARDS have been variable. This may relate to heterogeneity of the patients studied in terms of their underlying lung disease, duration of ARDS, and method of recruitment 2021. Several studies have demonstrated a beneficial effect on oxygenation, which is sustained in the presence of adequate PEEP 222324. Patients ventilated in the supine position benefit more than when in the prone position, which is probably related to the presence of more dependent, collapsed lung 2125. Similarly, the oxygenation benefit of recruitment maneuvers in patients ventilated with a high PEEP strategy is only modest 21. Several other clinical studies have demonstrated minimal or no beneficial effect of recruitment maneuvers 2627. A study of a moderate sustained inflation (35 cmH₂O for 30 s) in patients on a relatively high PEEP ventilation protocol demonstrated only a small and variable improvement in oxygenation, which was not sustained 26.

Another potential role for lung recruitment maneuvers is in the evaluation of the appropriate PEEP and tidal volume combination for a patient, and to gauge responsiveness to PEEP 20. A decremental PEEP trial following a recruitment maneuver can identify the PEEP level required to prevent derecruitment 28.

Recruitment maneuvers clearly improve oxygenation in some patients with ARDS. However, it remains unknown whether this is associated with a reduction in ventilator-induced lung injury, as has been demonstrated in animal models. Few randomized controlled trials incorporating lung volume recruitment maneuvers have been published. The study conducted by Amato and coworkers 29 demonstrated a mortality benefit in the arm treated with pressure limitation and an open lung approach that included recruitment maneuvers. It is difficult to determine the beneficial effect of the recruitment component given the other significant differences in ventilatory strategy. A US National Institutes of Health funded study comparing pressure limited ventilation using a high PEEP strategy (including recruitment maneuvers) with a low PEEP strategy was discontinued early because of a lack of benefit 30. A large Canadian study incorporating recruitment maneuvers into a lung protective strategy is nearing completion.

Many recent innovations in mechanical ventilation provide their benefit largely through recruitment of derecruited lung units, including high frequency oscillation, partial liquid ventilation, and prone positioning 31. In this section of the review, lung volume recruitment maneuvers are described that can be applied to the patient on conventional modalities of ventilation.

Animal and clinical studies have described diverse methods for recruiting the lung. A sustained high-pressure inflation uses pressures from 35 to 50 cmH₂O for a duration of 20–40 s 222729. Pressure may need to be individualized, with higher airway pressures required to generate an equivalent transpulmonary pressure in the patient with increased intra-abdominal pressure. Bladder pressure measurements can be used to identify these patients. A sustained inflation is usually achieved by changing to a CPAP mode and setting the pressure to the desired level. It is important to ensure that the pressure support level is set to zero to avoid additional pressure increases. Paralysis is usually not required for sustained inflations, but additional short-acting sedation may be useful. The patient should be closely monitored during this short period for hypotension and hypoxemia. Intermittent sighs have been demonstrated to achieve recruitment, using three consecutive sighs set at 45 cmH₂O pressure 23. An 'extended sigh' has been described, involving a stepwise increase in PEEP and decrease in tidal volume over 2 min to a CPAP level of 30 cmH₂O for 30 s 32. Other methods include an intermittent increase in PEEP for two breaths every minute 24 and increasing peak inspiratory pressure by increments of 10 cmH₂O to levels greater than 60 cmH₂O for brief periods 33. Increasing the ventilatory pressures to a peak pressure of 50 cmH₂O for 30–120 s may provide equivalent recruitment effects 343536. The effect of recruitment may not be sustained unless adequate PEEP is applied to prevent derecruitment 212228.

The effect of recruitment maneuvers can be monitored at the bedside using gas exchange indices or physiological parameters such as lung compliance. Imaging techniques, including chest radiography or computed tomography, may also be useful. Bedside evaluation of recruitment was discussed in detail in a recent review 37. From a practical perspective, improved oxygenation with a reduction in partial carbon dioxide tension indicates lung recruitment. Pressure effects may redirect blood flow and improve oxygenation in the absence of recruitment, but this would not be associated with a reduced partial carbon dioxide tension.

Despite the increasing body of literature on recruitment, few studies have compared the various methods in terms of efficacy and adverse effects. Sustained high pressure may cause transient hypotension, and may be less well tolerated than methods using higher pressure ventilation. Sustained or intermittent increases in peak pressure carry a risk for barotrauma. The choice of recruitment maneuver may depend on the baseline ventilatory mode; a spontaneously breathing patient may not tolerate a sustained high-pressure inflation, and a transient increase in PEEP and peak pressure may be more appropriate in this situation. There is some evidence that the type of lung injury (pulmonary versus extrapulmonary) may affect tolerance to and efficacy of various recruitment modalities 21. The frequency with which recruitment maneuvers must be applied is also unknown. This probably depends on the underlying disease, the level of PEEP, and procedures such as endotracheal suctioning 35. Other than the study conducted by Amato and coworkers 29, no outcome data exist suggesting that there is a mortality benefit from recruitment maneuvers.

Although most studies have evaluated recruitment maneuvers within the context of ARDS, this intervention may be of value in patients with atelectasis related to general anesthesia 38, during postoperative ventilation 39, following suctioning 35, or in other conditions that produce hypoxemia including heart failure. Response to recruiting interventions does not occur in all patients with ARDS 4041, and several studies have identified characteristics that may predict a response, in terms of oxygenation or improved lung mechanics.

The duration of ARDS appears to be an important factor, with a higher response rate noted in patients early in their disease course (e.g. < 72 hours) than later 41. This probably relates to the change in disease from an exudative to a fibroproliferative process. Similarly, the underlying pulmonary process may have an impact on responsiveness to recruitment attempts. Patients with extrapulmonary ARDS (e.g. secondary to sepsis) have a higher response rate than those with pulmonary ARDS (e.g. pneumonia) 1523. Patients with pneumonia may have a limited amount of recruitable lung tissue, and the higher pressure may overinflate normal lung rather than aerating the consolidated tissue 16. The effect of recruitment maneuvers may be limited by the ability of the chest wall to expand. Patients with poor chest wall compliance were less likely to benefit from recruitment maneuvers than those with compliant chest walls 41. Patients with ARDS who are ventilated with high tidal volumes or high levels of PEEP are less apt to derecruitment and may not exhibit a response to recruiting interventions 1424. Because prone positioning recruits lung volume and reduces the anteroposterior intrathoracic pressure gradient, volume recruitment maneuvers may be less necessary. However, in the prone position the pressure required to achieve recruitment is lower and the effect is more sustained 2125.

The inspired oxygen fraction may affect lung recruitment, because of absoption atelectasis in situations where inspired oxygen fraction approaches 1.0. The recruitment effect may be rapidly lost in patients ventilated on 100% oxygen 42.

The time course of response to recruitment maneuvers remains unclear. Lung mechanics in ARDS vary with time 43, and it remains unknown whether the recruitment response varies throughout the day or is related to changes in patient position or spontaneous ventilatory effort. Although a response is more likely early in the course of disease, these studies have only been performed at a single time period. Although the studies cited above have given some insight into identifying patients who may respond to recruitment maneuvers, this does not address the question of whether this intervention is beneficial in terms of reducing lung injury or mortality in this group.