Extravascular lung water and acute respiratory distress syndrome - Oxygenation and outcome

Abstract

We studied eleven consecutive patients to assess the influence of extravascular lung water on clinical outcome. All patients were mechanically ventilated using a standardized protocol. Inspired oxygen concentration was adjusted to an initial target P(a)O2 of greater then 8.0 kPa (60.8 mmHg). All patients received inhaled nitric oxide (NO) at a concentration of 20 ppm. Extravascular lung water index (EVLWI) was measured by a dual indicator technique (COLD Z-021 monitoring system, Pulsion, Munich, Germany). Patients were managed with fluids and inotropes according to a standard protocol. Median age was 45 (range 27-60 years), mean APACHE II score on admission 31 (range 17-36), duration of mechanical ventilation 15 (range 6-28) days, mean admission Murray lung injury score 2.5 (range 2-3) and admission EVLWI 20.8 (range 8.7 to 54.7) ml.kg-1. The only variables independently predictive of P(a)O2/FiO2 ratio were serum albumin (B = 1.7 ± 1.61) and EVLWI (B = 2.1 ± 0.47), r2 = 0.33, P < 0.001. In severe ARDS, (P(a)O2/FiO2 < 150 mmHg), mean EVLWI was 24.4 (22.4 to 26.4, 95% confidence intervals) ml.kg-1 compared with 15.1 (12.2 to 18.0) ml.kg-1 during moderate ARDS (P < 0.001). Serum albumin likewise differed, 29.4 (27.6 to 31.2) vs 35.1 (31.8 to 38.4) g.l-1, P < 0.005. PAOP was higher during periods of poor oxygenation, 12.7 (11.9 to 13.5) vs 9.3 (7.9 to 10.7) mmHg, P < 0.001. The four survivors had greater initial EVLWI than non-survivors, 31 (24.1 to 37.9) vs 20.7 (16.0 to 25.4) ml.kg-1, P = 0.034 and showed a greater reduction in lung water, 15.2 (9.3 to 21.1) vs 5.4 (2.1 to 8.7) ml.kg-1, P = 0.013.