Introduction: Prediction of the clinical deterioration in hospitalized COVID-19 patients is an unmet goal.

Aim: To assess monitoring of lung fluid status of hospitalized COVID-19 patients as a tool to predict clinical respiratory deterioration and prognosis.

Methods: The present study population comprised 51 patients hospitalized in Hillel Yaffe Medical Center with COVID-19 infection. The lung fluid status of patients was monitored by repeat measurements the lung impedance (LI). The LI technique was found to be a very effective tool for monitoring and guiding the treatment of a heart failure patients. Decreasing LI reflects  lung fluid accumulation. Clinical and laboratory parameters, chest X-ray (CXR) and LI level were recorded during hospitalization.

Results: Of the 51 patients hospitalized for COVID-19 infection (37- men and 14 - women, 55.7±12.6 years old), 46 were discharged after successful treatment (Group 1) and 5 (9.8%) died during hospitalization (Group 2). The LI kinetics during hospitalization demonstrated a different pattern between groups (Figure 1, p< 0.01). In group 1 patients, a small LI decrease (-3.5±4.3%, p=0.7) during the first 4 days (median = 2.2 days, [Q1-3: 1-3.7 days]) of hospitalization was noted. Following this, LI increased progressively until discharge (+20.3±12.3%, p<0.01). Among group 2 patients, LI decreased progressively during hospitalization. Mechanical ventilation was initiated at the eighth day [median-8, Q1-3: 4-12 days] when LI decreased by 18.2±3.8% in comparison with the admission level (p<0.01). Deaths occurred at 12.4±2.7 days (median-12 days) after admission. Multivariate Cox regression analysis of clinical, laboratory and CXR variance has shown that the degree of LI decrease during hospitalization is the most reliable predictor of death (hazard ratio: 1.36 [1.04-1.79], p<0.04).

Conclusions: The combination of progressively decreasing LI after 4 days of hospitalization for COVID-19 infection and an LI decrease >15% is the most reliable predictor of death.