Lung structure and function in different stages of severe adult respiratory distress syndrome

To assess the clinical consequences of duration of adult respiratory distress syndrome (ARDS) on lung structure and function.

Retrospective analysis.

A university hospital referral center for extracorporeal support.

A total of 84 patients with severe ARDS (Murray score > 2.5) recruited from 48 intensive care units (1979 to 1992), who suffered ARDS and underwent mechanical ventilation for up to 1 week (37 patients with early ARDS), between 1 and 2 weeks (24 patients with intermediate ARDS), or more than 2 weeks (23 patients with late ARDS) and subsequently underwent extracorporeal support.

Before beginning extracorporeal support, we measured gas exchange, pulmonary mechanics, hemodynamics, oxygen transport and delivery, incidence of barotrauma (presence of one or more thoracic tubes for pneumothorax drainage), and organ dysfunctions. In a subgroup of 16 patients, we studied lung structure by computed tomographic scan, scoring the densities and quantifying the emphysemalike lesions (bullae).

Late ARDS showed lower respiratory compliance, higher dead space, higher PaCO2, lower venous admixture, and lower positive end-expiratory pressure requirement compared with early ARDS (P < .01). The incidence of pneumothorax (48.8% of the entire population) was significantly (P < .01) higher in late ARDS (87%) vs intermediate ARDS (46%) and early ARDS (30%). The mortality of patients with pneumothorax (66%) was significantly (P < .01) higher compared with patients without pneumothorax (46%). The number of bullae per lung was significantly higher in late ARDS vs intermediate and early ARDS (mean [SD], 8 [4], 4.3 [5], and 1.9 [3.9], respectively; P < .01), and they were preferentially distributed in the dependent lung regions. The number of bullae per lung was significantly higher in patients with pneumothorax vs those without pneumothorax (mean [SD], 13.6 [9.8] vs 1.4 [2.1]; P = .007). The mean (SD) duration of ARDS in patients with pneumothorax was 15.3 (10.0) days vs 7.0 (6.6) days in those without pneumothorax (P = .0001). No differences within the three groups were found in computed tomographic scan densities, hemodynamics, and number of organ system dysfunctions.

The lung structure and function changes markedly with ARDS duration, and the late stages may be described as restrictive lung disease with superimposed emphysemalike lesions. Presence of pneumothorax affects survival and appears to be related to the lung structural changes occurring with time.