ObjectiveTo establish the value of the67Gallium (Ga) pulmonary leak index, a marker of increased permeability edema of the lungs, in assessing the severity and course of the adult respiratory distress syndrome (ARDS).DesignProspective observational study.SettingMedical intensive care unit of a university hospital.PatientsSeventeen consecutive, mechanically-ventilated ARDS patients. Eleven patients (recovering from ARDS) improved, as defined by the ability to taper the level of positive end-expiratory pressure (PEEP) to 0 cm H2O at a median of 7 days after admission. Ten patients survived. Six patients did not recover and died a median of 3.5 days after admission.InterventionsNone.Measurements and Main ResultsThe pulmonary leak index (i.e., upper limit of normal 14.1 times 10 sup -3/min) was measured within 72 hrs after admission, and repeated within 48 hrs at the time of recovery in recovering patients. At admission and recovery, respiratory variables were recorded and a lung injury score was calculated. At admission, the pulmonary leak index was increased in each patient to 32.3 (range 15.6 to 52.4) and 28.7 (range 26.0 to 40.8) times 10 sup -3/min in recovering and nonrecovering patients, respectively (NS). Groups did not differ with respect to the oxygenation ratio, the level of PEEP, radiographic abnormalities, and the lung injury score. At recovery, the pulmonary leak index in recovering patients had decreased in each patient and had normalized in four patients, averaging 15.2 (range 5.6 to 25.9) times 10 sup -3/min, concomitantly with an increased oxygenation ratio, less radiographic abnormalities, and a decreased lung injury score (p < .01 vs. admission). For pooled recovering and nonrecovering patient data, the pulmonary leak index directly related to the lung injury score (p < .005).ConclusionsThe67Ga pulmonary leak index may be a clinically useful tool to quantify the severity and course of ARDS, since clinical recovery from the syndrome may be characterized by amelioration of increased microvascular permeability.(Crit Care Med 1996; 24:1467-1472)