Background Coronavirus disease 2019 (COVID-19) has a wide spectrum of severity, and early identification of high-risk patients remains a major clinical challenge, particularly in resource-limited settings. Serum ferritin, an acute-phase reactant, has been proposed as a potential biomarker of disease severity. This study aimed to evaluate the association between serum ferritin levels and COVID-19 disease severity and to assess its diagnostic performance as a biomarker for identifying severe-to-critical cases among COVID-19 patients in Bangladesh. Methods This hospital-based cross-sectional study was conducted between May and December 2021 at Bangabandhu Sheikh Mujib Medical University. A total of 46 reverse transcription polymerase chain reaction (RT-PCR)-confirmed adult COVID-19 patients were included via consecutive sampling, with equal distribution between the mild-to-moderate and severe-to-critical groups. Clinical data and laboratory parameters, including serum ferritin, were recorded at admission. Statistical analysis was performed via SPSS version 26. Group comparisons, logistic regression, and receiver operating characteristic (ROC) curve analyses were conducted. Results The mean age of the participants was 53.35 ± 15.9 years, and 23 (50%) participants were in each severity group. Serum ferritin remained significantly associated with disease severity in both the unadjusted (OR = 1.3, 95% CI: 1.03-1.33, p = 0.015) and age-adjusted (OR = 1.1, 95% CI: 1.01-1.23, p = 0.018) models. Neutrophilia, lymphopenia, and elevated D-dimer and C-reactive protein (CRP) levels were also significantly associated with severe disease (p < 0.001). ROC analysis revealed an area under the curve (AUC) of 0.76, with a cutoff value of 752 ng/mL, yielding 69.6% sensitivity and 73.9% specificity (p = 0.003). Conclusion Serum ferritin is significantly associated with COVID-19 severity and may be a useful, low-cost biomarker for identifying severe disease. It may serve as a potential indicator for clinical assessment. However, further large-scale multicenter studies are needed to validate its prognostic utility.
Afroz et al. (2026) studied this question.