Monitoring hyperhydration and polyneuromyopathy in critically ill patients using Bioelectrical impedance

Abstract Background : Hyperhydration and muscle wasting can be profoundly detrimental in the critically ill. Properly assessing the fluid status and body composition is not trivial and is still a considerable challenge. Our study aims to assess the feasibility of using Bioelectrical Impedance Vector Analysis (BIVA) to monitor hydration status and skeletal muscle mass in the critically ill [including patients on extracorporeal membrane oxygenation (ECMO)], and to determine the effect on mortality of hyperhydration and muscle loss measured by BIVA compared with the recorded cumulative fluid balance (CFB). Another objective was to assess the prognostic ability of the phase angle (PA from BIVA) in comparison to that of other prognostic markers (APACHE II, serum presepsin). Methods : We set up a prospective, clinical blinded observational study comprising COVID-19 patients admitted to the ICU due to acute respiratory distress syndrome (ARDS) with an expected duration of artificial pulmonary ventilation (APV) of at least seven days. We recorded demographic data, daily cumulative fluid balance (CFB) and laboratory test results (albumin, prealbumin, presepsin, CRP) along with BIVA measurements. Measurements were performed 3 times during hospitalization. The first measurement was performed within 48 hours of admission, then one week after admission and the last one before discharge. Results: Using the 61 COVID-19 patients with ARDS on APV and/or ECMO with 3 consecutive BIVA measurements (each taken one week apart), we were able to demonstrate a positive correlation between hyperhydration (increase ECW/TBW and decrease PA) and mortality. The result was statistically significant (p = 0.005) when we excluded ECMO patients. Unlike CFB, which did not yield a statistically significant result in assessing hyperhydration, the ECW/TBW ratio and overhydration (OHY) from bioimpedance was more accurate (p = 0.0050 and p = 0.0402 in the non-ECMO group). There was no statistically significant decrease in skeletal muscle mass (SMM). Conclusions: BIVA can be used for monitoring hydration and nutritional status in critically ill patients and is more accurate than CFB, although the need for ECMO support limits its accuracy. Hyperhydration (ECW/TBW and OHY) and low PA are also effective in predicting mortality only in non-ECMO patients. Regression analysis confirms the correlation between hyperhydration (short BIVA vector length and higher ECW/TBW ratio) and increasing presepsin with concomitant decreasing albumin.