PURPOSE: The difficulties of fluid therapy in patients with septic shock are to maintain sufficient vascular volume while preventing pulmonary edema formation. Thus, it is important to find a biomarker that can reliably predict pulmonary edema formation after fluid loading. We evaluated the association of plasma protein levels with the increase in extravascular lung water index (ΔEVLWI) after fluid loading. METHODS: This was an observational study in which we retrospectively reviewed medical records of septic patients in whom hemodynamic variables were measured by transpulmonary thermodilution technique before and after fluid loading. Plasma protein levels were measured before fluid loading. Patients were divided into 2 groups according to the changes in EVLWI (ΔEVLWI ≥ 10%) after fluid loading. Diagnostic performance of plasma proteins in predicting pulmonary edema formation was assessed. RESULTS: A total of 62 patients were included, and 27 of them showed a ΔEVLWI 10% or higher after fluid loading. Plasma albumin and transferrin were significantly lower in this group than in the group with ΔEVLWI less than 10% (21.7 ± 5.8 vs 25.3 ± 5.0 g/L for albumin, P < .05; 107.9 ± 50.1 vs 136.8 ± 44.2 mg/dL for transferrin, P < .05). Areas under the curve of albumin and transferrin were 0.68 (cardiac index, 0.54-0.83) and 0.72 (cardiac index, 0.59-0.86), respectively. At a cutoff value of 87.9 mg/dL, transferrin had a sensitivity of 0.91 in predicting ΔEVLWI 10% or higher. CONCLUSIONS: Plasma transferrin and albumin levels were associated with ΔEVLWI 10% or higher after fluid loading. The high sensitivity of both biomarkers indicated that patients with normal values were less likely to develop pulmonary edema after fluid loading.
PURPOSE: The difficulties of fluid therapy in patients with septic shock are to maintain sufficient vascular volume while preventing pulmonary edema formation. Thus, it is important to find a biomarker that can reliably predict pulmonary edema formation after fluid loading. We evaluated the association of plasma protein levels with the increase in extravascular lung water index (ΔEVLWI) after fluid loading. METHODS: This was an observational study in which we retrospectively reviewed medical records of septicpatients in whom hemodynamic variables were measured by transpulmonary thermodilution technique before and after fluid loading. Plasma protein levels were measured before fluid loading. Patients were divided into 2 groups according to the changes in EVLWI (ΔEVLWI ≥ 10%) after fluid loading. Diagnostic performance of plasma proteins in predicting pulmonary edema formation was assessed. RESULTS: A total of 62 patients were included, and 27 of them showed a ΔEVLWI 10% or higher after fluid loading. Plasma albumin and transferrin were significantly lower in this group than in the group with ΔEVLWI less than 10% (21.7 ± 5.8 vs 25.3 ± 5.0 g/L for albumin, P < .05; 107.9 ± 50.1 vs 136.8 ± 44.2 mg/dL for transferrin, P < .05). Areas under the curve of albumin and transferrin were 0.68 (cardiac index, 0.54-0.83) and 0.72 (cardiac index, 0.59-0.86), respectively. At a cutoff value of 87.9 mg/dL, transferrin had a sensitivity of 0.91 in predicting ΔEVLWI 10% or higher. CONCLUSIONS: Plasma transferrin and albumin levels were associated with ΔEVLWI 10% or higher after fluid loading. The high sensitivity of both biomarkers indicated that patients with normal values were less likely to develop pulmonary edema after fluid loading.