ObjectivesTo investigate the serum IL-6 levels and their rate of change in predicting the mortality of critically ill patients with COVID-19 in Vietnam.DesignA single-centre, cross-sectional study.SettingAn Intensive Care Centre for the Treatment of Critically Ill Patients with COVID-19 in Ho Chi Minh City, Vietnam.ParticipantsWe included patients aged 18 years or older who were critically ill with COVID-19 and presented to the study centre from 30 July 2021 to 15 October 2021. We excluded patients who did not have serum IL-6 measurements between admission and the end of the first day.Primary outcome measuresThe primary outcome was hospital all-cause mortality.ResultsOf 90 patients, 41.1% were men, the median age was 60.5 years (Q1–Q3: 52.0–71.0), and 76.7% of patients died in the hospital. Elevated IL-6 levels were observed on admission (41.79 pg/mL; Q1–Q3: 20.68–106.27) and on the third day after admission (72.00 pg/mL; Q1–Q3: 26.98–186.50), along with a significant rate of change in IL-6 during that period (839.5%; SD: 2753.2). While admission IL-6 level (areas under the receiver operator characteristic curve (AUROC): 0.610 (95% CI: 0.459 to 0.761); cut-off value ≥15.8 pg/mL) and rate of change in IL-6 on the third day of admission (AUROC: 0.586 (95% CI: 0.420 to 0.751); cut-off value ≥−58.7%) demonstrated poor discriminatory ability in predicting hospital mortality, the third day IL-6 rate of change from admission ≥−58.7% (adjusted OR: 12.812; 95% CI: 2.104 to 78.005) emerged as an independent predictor of hospital mortality.ConclusionsThis study focused on a highly selected cohort of critically ill COVID-19 patients with a high IL-6 level and mortality rate. Despite the poor discriminatory value of admission IL-6 levels, the rate of change in IL-6 proved valuable in predicting mortality. To identify critically ill COVID-19 patients with the highest risk for mortality, monitoring the serial serum IL-6 measurements and observing the rate of change in serum IL-6 levels over time are needed.
