Except in patients with a 'diffuse' lung morphology whose radiographic and CT aspects are often concordant showing bilateral 'white lungs', the bedside frontal chest radiograph is misleading in the majority of patients with ARDS. In a series of 70 patients,bedside frontal chest radiography correctly identifiedlungmorphology in 41% of the patients only, the highest rate of error being observed in patients with 'focal' ARDS . Surprisingly, in some severely hypoxemic patients, the frontal bedside chest radiograph may remain grossly normal: only a few basal hyperden-sities can be identified and indirect signs suggesting a major reduction in the volume of lower lobes such as the visualization of the small fissura immediately above the right diaphragmatic cupola are present. This radiological presentation, which often confuses the clinician, generally corresponds to nonaerated and partially atelectatic lower lobes which stand on the posterior face ofthe diaphragm and caudally to the diaphragmatic cupola. In contrasting with the apparent preservation of lung aeration on the frontal chest radiograph, lung CT always shows a major lung volume loss predominating in lower lobes and explaining the severe impairment of arterial oxygenation . In fact, the aeration present on frontal chest radiography concerns essentially upper lobes, which, paradoxically, are also characterized by a substantial increase in lung tissue.
Because a substantial proportion of the lung parenchyma remains normally aerated at ZEEP in the majority of ARDS patients , regional lung compliances are unevenly distributed . Very often, upper lobes are more compliant than lower lobes and the initial increase in intrathoracic pressure distends cephalic lung regions before recruiting nonaerated pulmonary areas [20, 41]. Increasing the intrathoracic pressure by increasing PEEP often overinflates the aerated lung regions while, concurrently, nonaerated areas begin to be recruited [19,20,41,42]. These results have been experimentally observed in dogs with oleic acid-injured lungs where regional lung volumes were measured using the parenchymal marker techniques [9, 10]. In the minority of patients whose ARDS is characterized by a diffuse and bilateral loss of aeration, the risk of overinflation appears more limited [18,20,41]. The lack of normally aerated lung regions at ZEEP explains why PEEP greater than 10 cmH2O does not induce any detectable lung overinflation . The most severe forms of lung infection, pulmonary contusion, aspiration pneumonia fat embolism, amniotic embolism, and near drowing are characterized by a 'diffuse' loss of lung aeration whereas less severe forms of primary ARDS show a more 'focal' distribution of the aeration loss [20, 21]. Most of secondary ARDS cases are characterized by a 'focal' loss of aeration . As shown in a series of 69 patients with ARDS , primary and secondary ARDS patients do not differ as far as basal cardiorespiratory parameters, cardiorespiratory effects of PEEP, and survival are concerned. The response to PEEP is not influenced by the nature of lung injury primary or secondary as previously suggested  but rather by the lung morphology which depends on the severity of lung injury.
The rationale for selecting the 'right' PEEP level is based on the experimental and clinical finding that lung overinflation and alveolar recruitment occur simultaneously in many patients or animals with ARDS . The optimal PEEP for a given patient can be defined as the PEEP allowing optimization of arterial oxygenation without introducing a risk of oxygen toxicity and VILI . In the majority of ARDS patients in whom significant parts of the lungs remain normally aerated at ZEEP, the PEEP trial should range between 5 and 12 cmH2O in order to avoid overinflation of aerated lung regions that would inevitably result from the application of higher PEEP required to keep the lung fully aerated at end-expiration. In other words, most injured lungs cannot be entirely reaerated without introducing a risk of VILI. In the minority of patients without a single lung region normally aerated at ZEEP, high PEEP can be safely applied without overinflating other parts of the lung [19, 20, 41] and the concept of keeping the lung fully aerated maybe accepted . In both situations, the use of periodic sighs could be useful [38, 44, 45]. CT studies have provided evidence that at a given PEEP level, end-expiratory aeration is markedly dependent on the preceding inspiratory plateau pressure: the higher the inspiratory plateau pressure, the more the PEEP prevents end-expira-torylung derecruitment .
Was this article helpful?