Do recruitment maneuvers simply improve oxygenation?

Recruitment maneuvers have been the subject of intense investigation. Their role in the acute care setting is debated given the lack of information on their influence on clinical outcomes. Oxygenation improvement is often a striking effect, together with changes of respiratory mechanics. However, hemodynamic compromise is frequently associated with the maneuver, sometimes even barotrauma. Another possible downside is bacterial translocation secondary to lung overdistention, as suggested by experimental and initial clinical data. When a recruitment maneuver is performed, the patho-physiological consequences of lung recruitment should guide clinicians more than oxygenation improvement alone.

Th e investigation reported by Constantin and colleagues in a recent issue of Critical Care [1] dwells on a philosophy summarized by Lachmann years ago in an aphorism still current among intensive care practitioners: 'Open up the lung and keep it open' [2]. Along the learning curve of the ventilator settings of patients with acute lung injury, the pivotal role of positive endexpiratory pressure (PEEP) soon became clear. It took some time to understand that PEEP maintains the lung open once, and if, previously opened. Since then several investigators have focused on the recruitment maneuver, proposing diff erent strategies with the common rationale of needing to 'open up' the lung. Th e study by Constantin and colleagues is one such investigation: their randomized controlled trial aimed to evaluate improvement of oxygenation after a recruitment maneuver performed soon after intubation in the ICU. Th e setting is new and of interest, although the primary end-point (oxygenation improvement) is somewhat expected in the early stages of acute respiratory distress syndrome, when lung atelectasis is likely to occur [3].
Th e absence of patients with chronic obstructive pulmo nary disease or cardiogenic shock, and the exclusion of patients with encephalopathy, coma, acute brain injury, the need for cardiac resuscitation, or recent thoracic surgery do not allow generalization of the results in the ICU setting. Nevertheless, the trial clearly confi rms the culprit role of recruitment maneuvers in the acute care settings.
However, the results of the trial and the discussion of the results raise a number of questions, one of which is: should we care about oxygenation during recruitment maneuvers? Yes, of course. In fact, apart from being a great relief to clinicians, oxygenation improvement usually implies resolution of atelectasis, better distribution of ventilation, and delivery of a given tidal volume to a greater lung volume, hence lower lung strain, all factors that may contribute to attenuate ventilator-induced lung injury. However, oxygenation per se is possibly the least relevant variable. Arterial partial pressure of oxygen (PaO 2 ) is only one factor (not the most important for quantity) contributing to oxygen delivery. Is a PaO 2 of 80 mmHg a real concern in a sedated paralyzed patient, with normal oxygen carrier capability (hemoglobin) and peripheral delivery of oxygen (cardiac output, distri bution of fl ow)? Not from a metabolic point of view. In this light, the increased CO 2 wash-out that followed the recruitment maneuver in Constantin and colleagues' investigation looks more interesting. Unless exceedingly low, what matters is not the absolute value of PaO 2 , but rather the full price to obtain it: low or high inspiratory fraction of oxygen, 5 or 15 cmH 2 O of PEEP, supine or prone position? Th us are recruitment maneuvers. Th ey undoubtedly improve oxygenation in many circumstances, but clearly have a price: frequently hemodynamic compromise, although transient and short lasting; sometimes desaturation; sometimes new air leaks [4].
In their investigation, Constantin and colleagues shed some light on an interesting potential downside of recruit ment maneuvers: microbiological safety. Data show that mechanical ventilation strategies promote dissemination of intratracheally instilled bacteria [5] or increase susceptibility to development of bacteremia [6]. Constantin and colleagues investigated the relationship between tracheal intubation and blood versus endotracheal aspirate bacteriological fi ndings. In their study this was a secondary endpoint, and the number of positive samples was too low to draw any conclusion. Yet the results are of note: several patients with negative blood cultures at baseline became positive for the same bacterial species found in the tracheal aspirate as soon as 5 minutes after intubation. Whether the species found in the blood were lung borne is not proven. However, the time course of blood culture is striking. Th e fact that a similar number of patients in the control and recruitment maneuver group became positive after intubation apparently favors the safety of the recruitment maneuver itself, as Constantin and colleagues discuss. Exacerbation of hypoxia at the time of intubation [7] or the interaction between hypoxia and mechanical ventilation [5], maybe even peri-intubation ventilation, cannot be excluded. Irrespective of the mechanism, this is an intriguing concept that deserves further insight, given its potential impact on outcome.
Whatever the contribution of recruitment maneuvers to the development of ventilator-induced lung injury, if any, data on their long-term eff ects on primary or secondary outcomes in patients with acute lung injury are lacking. What is given is that recruitment maneuvers do not simply improve oxygenation.