PURPOSE: Patients with coronavirus disease 2019 (COVID-19)-related acute respiratory distress syndrome (ARDS) who lack clinical improvement are frequently treated with high-dose corticosteroids (HDS). Since HDS is used to reduce hyperinflammation in these patients, levels of (pro-)inflammatory biomarkers after commencing HDS treatment could be useful in predicting mortality. This study aims to evaluate biomarker levels after commencing HDS over time, along with their capacity to predict mortality. PATIENTS AND METHODS: This retrospective cohort study included patients with COVID-19 ARDS treated with HDS in the intensive care unit (ICU) at an academic hospital in the Netherlands between March 2020-March 2022. Inflammatory biomarkers (ie, C-reactive protein (CRP), D-dimer, ferritin, leukocyte count, interleukin-6 (IL-6), lactate dehydrogenase (LDH), neutrophil-to-lymphocyte ratio (NLR), and procalcitonin (PCT)) were assessed daily from start of HDS (ie baseline) until day 7. Associations between biomarker levels and all-cause-hospital-mortality were evaluated each day using logistic regression, with cut-offs identified by optimizing sensitivity (Se) and specificity (Sp). RESULTS: Of the 122 patients included, 53 (43.4%) died during hospitalization. HDS was initiated for a median 7 days (IQR=1-11) after ICU admission. At baseline, a moderately high predictive capacity for mortality was observed at a ferritin level >
1281 µg/L (Se=62%/Sp=64%), leukocyte count >
13.7 × 109/L (Se=42%/Sp=79%), and NLR >
12.1 (Se=61%/Sp=77%). During follow-up, CRP >
50 mg/L on day 6 (Se=50%/Sp=75%) and >
42 mg/L on day 7 (Se=50%/Sp=75%), ferritin >
1082 µg/L on day 6 (Se 63%/Sp=71%) and >
1852 µg/L on day 7 (Se=31%/Sp=79%), IL-6 >
67 mg/L on day 7 (Se=56%/Sp=79%) and LDH >
396U/L on day 6 (Se=38%/Sp=83%) and >
373 U/L on day 7 (Se=47%/Sp=72%) showed moderate capacity to predict mortality. NLR was consistently associated with mortality for all days, except day 1 (Se=36-68%/Sp=72-92%). CONCLUSION: In COVID-19 ARDS patients receiving HDS, several clinically available inflammatory biomarkers moderately predicted all-cause-hospital-mortality after the start of HDS, particularly on days 6 and 7. NLR demonstrated the most consistent association with mortality over time. The use of these markers requires validation in larger cohorts.