Bioelectrical impedance analysis values as markers to predict severity in critically ill patients

Abstract Purpose We investigated bioelectrical impedance analysis (BIA)-derived parameters in critically ill patients to evaluate any differences between survivors and nonsurvivors. Methods We calculated severity scores for 241 critically ill surgical patients (161 male and 80 female; mean age, 62.9 years) using three severity scoring systems (Acute Physiology and Chronic Health Evaluation II, Sequential Organ Failure Assessment, and Simplified Acute Physiology Score III). Body composition was measured using a portable BIA device for segmental BIA. Results Among the BIA values, impedance (odds ratio [OR], 0.99; P < 0.001), reactance (OR 0.90; P < 0.001), and phase angle (PhA) (OR, 0.53; P < 0.001) were highly statistically significant for predicting mortality in univariate and multivariate logistic regression analysis. Comparison of area under the curve (AUC) between severity scoring systems and BIA values showed statistically significant differences between reactance and PhA with all three severity scoring systems. Covariate-adjusted receiver operating characteristic curve analysis showed that compared with severity scoring, all three BIA values (impedance, reactance, and PhA) had higher AUC values. Conclusions PhA, impedance, and reactance determined by BIA in critically ill patients were associated with mortality outcomes and revealed stronger predictive power for mortality than severity scoring systems commonly used in an intensive care unit.