Prognostic Value of Extravascular Lung Water in Critically III Patients*

Sakka, Samir G. MD; DEAA
Klein, Magdalena
Reinhart, Konrad MD
Meier-Hellmann, Andreas MD

*From the Department of Anesthesiology and Intensive Care Medicine, Friedrich-Schiller-University of Jena, Germany.

Manuscript received September 12, 2001; revision accepted April 12, 2002.

Correspondence to: Samir Sakka, MD, Department of Anesthesiology and Intensive Care Medicine, Friedrich-Schiller-University of Jena, Bachstrasse 18, D-07740 Jena, Germany; e-mail: [email protected]

Chest 122(6):p 2080-2086, December 2002.

Objective:

Measurement of extravascular lung water (EVLW) as a clinical tool for the assessment of pulmonary function has been found to be more appropriate than oxygenation parameters or radiographic techniques. In this study, we analyzed the prognostic value of EVLW in critically ill patients.

Design:

Retrospective analysis.

Setting:

Operative ICU of a university hospital.

Measurements and results:

We retrospectively analyzed 373 critically ill patients (133 female and 240 male patients; age range, 10 to 89 years; mean ± SD age, 53 ± 19 years) who were treated in our ICU between 1996 and 2000. All these patients were hemodynamically monitored by the transpulmonary double-indicator (thermo-dye) dilution technique. Each patient received a femoral artery sheath through which a 4F flexible catheter with an integrated thermistor and fiberoptic was advanced into the infradiaphragmatic aorta. EVLW was calculated using a computer system. For each measurement, 15 to 17 mL of cooled 2% indocyanine green were injected central venously. In our results, maximum EVLW was significantly higher in nonsurvivors (n = 186) than in survivors (n = 187) [median, 14.3 mL/kg vs 10.2 mL/kg, respectively; p < 0.001]. In univariate logistic regression models, EVLW (r² = 0.024, p = 0.003) at baseline as well as simplified acute physiology score (SAPS) II (r² = 0.135, p < 0.0001) and APACHE (acute physiology and chronic health evaluation) II scores (r² = 0.050, p < 0.0001) were significant predictors of mortality. If SAPS II and APACHE II scores are combined, r² increases to 0.136, but the improvement over SAPS II alone is not significant. The addition of baseline EVLW further increases r² to 0.149 (p = 0.021 for the improvement), indicating that EVLW contributes independently to prognosis.

Conclusion:

EVLW correlated well with survival (ie, nonsurvivors had significantly higher EVLW values than survivors) and is an independent predictor of prognosis.