Association between longitudinal changes in phase angle and mortality rate in adults critically ill with COVID-19: A retrospective cohort study.

Abstract

Background: Phase angle (PhA) obtained by bioelectrical impedance analysis (BIA) works as a predictor of clinical outcomes. Specific cutoff values for longitudinal changes and their relationship with clinical outcomes are still undetermined for patients with critical illness. Thus, the aim of this study was to analyze the association between longitudinal changes in PhA during intensive care unit (ICU) stay and all-cause 90-day mortality in patients critically ill with COVID-19.

Methods: This was a retrospective cohort study of adults critically ill with COVID-19 undergoing invasive mechanical ventilation with a length of stay >14 days. BIA was performed at ICU admission and at days 7 and 14 of ICU stay; PhA and hydration parameters were collected. Differences between survivors and nonsurvivors were assessed. Longitudinal changes were evaluated using repeated-measures analysis of variance. A receiver operating characteristics curve for PhA declined (%) during the first 14 days, and all-cause 90-day mortality was performed. Survival probability was reported using hazard ratios (HR).

Results: One-hundred nine patients were included. The change in the value of PhA was close to 17.1%. Nonsurvivors had a higher prevalence of individuals with a decrease in PhA >22.2% (area under the curve = 0.65) in the first 14 days in comparison with survivors (70% vs 34.8%, P < 0.01). PhA decrease >22.2% at 14 days was a significant predictor of all-cause 90-day mortality (HR = 2.2, 95% CI 1.71-3.6, P = 0.04).

Conclusion: Changes in PhA are associated with all-cause 90-day mortality. Future studies should be directed to interventions to prevent changes in this nutrition marker.