• Q Lu, L M Malbouisson, E Mourgeon, I Goldstein, P Coriat, and J J Rouby.
• Réanimation Chirurgicale Pierre Viars, Department of Anesthesiology, La Pitié-Salpêtrière Hospital, 47-83, Boulevard de l'Hôpital, 75013 Paris, France.
• Intensive Care Med. 2001 Sep 1;27(9):1504-10.

ObjectivesTo study whether PEEP-induced reopening of collapsed lung regions--defined as the decrease in nonaerated lung volume measured on a single or three computerized tomographic (CT) sections--is representative of the decrease in overall nonaerated lung volume.DesignReview of 39 CT scans obtained in consecutive patients with Acute Lung Injury.SettingsFourteen-bed surgical intensive care unit of a University Hospital.Measurements And ResultsPEEP-induced decrease in nonaerated lung volume was measured in 39 patients with ALI on a single juxtadiaphragmatic CT section, on three CT sections--apical, hilar, and juxtadiaphragmatic--and on contiguous apex-to-diaphragm CT sections. The percentage of decrease in nonaerated lung volume following PEEP, was compared between one, three and all CT sections using a linear regression analysis and Bland and Altman's method. The decrease in nonaerated lung volume measured on a single and three CT sections was significantly correlated with the decrease in nonaerated lung volume measured on all CT sections: R=0.83, P<0.0001 for one CT section and R=0.92, P<0.0001 for three CT sections. However, measurements performed on a single CT section were poorly representative of the overall lung: bias -6%, limits of agreement ranging between -37% and +25%. Measurements performed on three CT sections overestimated by 11% the overall decrease in nonaerated lung volume: bias -11%, limits of agreement ranging between -29% and +7%.ConclusionsPEEP-induced reopening of collapsed lung regions measured on a single or three CT sections sensibly differs from the reopening of collapsed lung regions measured on the overall lung. The inhomogeneous distribution of PEEP-induced reopening of collapsed lung regions along the cephalocaudal axis probably explains these discrepancies.

30 keywords...

hide…