Pediatric sepsis remains a leading cause of admission and mortality in pediatric intensive care units (PICUs) worldwide. The systemic dysregulation of the host inflammatory response triggers the synthesis and release of acute-phase proteins, among which serum ferritin has emerged as an important prognostic biomarker1. Markedly elevated serum ferritin concentrations reflect severe pro-inflammatory states and macrophage activation, and have been strongly correlated with multi-organ dysfunction syndrome (MODS) and an increased risk of mortality in critically ill septic children2,3. Despite established associations in Western cohorts, routine risk stratification using ferritin remains inconsistent in resource-limited settings. Moreover, recent evidence from a tertiary care center in India indicated that admission serum ferritin levels did not reliably predict organ dysfunction or mortality in children presenting with sepsis due to endemic tropical infections, underscoring the necessity for localized validation4. Therefore, a critical gap exists in establishing reliable, region-specific cut-offs for hyperferritinemia to aid early clinical decision-making. This study aimed to evaluate the prognostic role of admission serum ferritin levels in predicting mortality and clinical outcomes among pediatric patients over one month of age admitted with severe sepsis to a tertiary care PICU.

This hospital-based cross-sectional analytical study was conducted in the 12-bedded pediatric intensive care unit (PICU) of a tertiary hospital in North India, over a period of one year. The study protocol was approved by the Institutional Ethics Committee, and written informed consent was obtained from the parents or legal guardians of all participants prior to enrollment. Children aged between >1 month and 18 years admitted to the PICU with a diagnosis of sepsis were consecutively enrolled. Sepsis was defined as the presence of systemic inflammatory response syndrome (SIRS) secondary to suspected or proven infection, requiring at least two SIRS criteria (with one necessarily being an abnormal core body temperature or leukocyte count)5. Patients were excluded if they met any of the following criteria: presence of pre-existing chronic organ dysfunction (liver, kidney, lung, or heart); chronic conditions affecting iron metabolism or erythropoiesis (e.g., thalassemia, sideroblastic anemia, hereditary hemochromatosis); non-infective chronic inflammatory diseases; and confirmed or suspected malignancies or autoimmune disorders. Relevant demographic, anthropometric, and clinical parameters were documented at enrollment, and infectious etiologies were confirmed via peripheral blood cultures and standard laboratory protocols. At admission, prior to initiating targeted therapeutic interventions, an additional 5 mL venous blood sample was drawn to quantify serum ferritin alongside protocol-driven intensive care investigations (including C-reactive protein, procalcitonin, comprehensive metabolic panels, and arterial blood gas analysis). Serum ferritin concentrations were assayed quantitatively by chemiluminescence using a Beckman Coulter auto-analyzer, with hyperferritinemia defined as levels exceeding 300 ng/mL. All participants were followed prospectively to evaluate the primary outcome of PICU mortality. Statistical analysis Data was analyzed using SPSS Statistics version 22.0 (IBM Corp). Categorical variables were summarized as frequencies and percentages and compared using the Pearson exact test. Continuous variables were assessed for distribution and summarized as means with standard deviations (SD) or medians with interquartile ranges (IQR), being compared via Student’s t-test or the Mann-Whitney U test as appropriate. Receiver operating characteristic (ROC) curve analysis was executed to evaluate the predictive accuracy of admission serum ferritin for mortality, calculating the area under the curve (AUC) and establishing an optimal discriminatory cut-off via the Youden Index. A two-tailed p-value <0.05 was considered statistically significant

A total of 70 children admitted to the PICU with sepsis were enrolled during the study period. The median age of the cohort was 6 years (IQR: 1–11 years), comprising 38 boys (54.29%) and 32 girls (45.71%). The primary outcome measure, overall PICU mortality, was documented at 31.43% (n=22). Infectious etiology was confirmed via positive peripheral blood cultures in 68 patients (97.14%). Baseline demographic, socioeconomic, and clinical severity characteristics stratified by PICU outcome are consolidated in Table 1. Fatal outcomes were significantly associated with rural residence (p=0.042). Markers of severe multi-organ dysfunction were profoundly more prevalent among non-survivors, demonstrating a universal requirement for inotropic support (100% vs. 41.67%; p<0.001). Furthermore, acute respiratory distress syndrome (ARDS) was present in 95.45% of non-survivors compared to 25.0% of survivors (p<0.001). Consequently, non-survivors required significantly prolonged durations of mechanical ventilation (11.68 ± 2.77 days vs. 5.63 ± 3.36 days; p<0.001) and extended overall PICU stays (14.59 ± 2.97 days vs. 6.92 ± 3.49 days; p<0.001). Table 1 Demographic, clinical, and treatment characteristics stratified by PICU outcome Characteristics Total (N=70) Survivors (n=48) Non-Survivors (n=22) p-value Age distribution, n (%) <1 year 17 (24.29) 10 (20.83) 7 (31.82) 0.099 1–5 years 13 (18.57) 10 (20.83) 3 (13.64) 6–10 years 21 (30.00) 18 (37.50) 3 (13.64) 11–18 years 19 (27.14) 10 (20.83) 9 (40.91) Sex (Boys), n (%) 38 (54.29) 28 (58.33) 10 (45.45) 0.315 Residence (Rural), n (%) 29 (41.43) 16 (33.33) 13 (59.09) 0.042 Socioeconomic status, n (%) Lower 39 (55.71) 28 (58.33) 11 (50.00) 0.721 Middle 18 (25.71) 11 (22.92) 7 (31.82) Upper 13 (18.57) 9 (18.75) 4 (18.18) Inotrope requirement, n (%) 42 (60.00) 20 (41.67) 22 (100.0) <0.001 CVS dysfunction present, n (%) 54 (77.14) 32 (66.67) 22 (100.0) 0.002 ARDS present, n (%) 33 (47.14) 12 (25.00) 21 (95.45) <0.001 Mechanical ventilation, days* 8.52±4.32 5.63±3.36 11.68±2.77 <0.001 PICU stay, days* 9.33±4.88 6.92±3.49 14.59±2.97 <0.001 CVS: Cardiovascular; ARDS: Acute respiratory distress syndrome; PICU: Pediatric intensive care unit. *Continuous variables expressed as mean ± standard deviation. Admission laboratory investigations and systemic inflammatory biomarkers are compared in Table 2. Non-survivors exhibited significantly greater initial hematological derangements, characterized by a lower mean hemoglobin concentration (p<0.001), pronounced leukocytosis (p<0.001), innate neutrophilia (p=0.042), and marked lymphopenia (p<0.001) compared to survivors. Crucially, admission serum ferritin concentrations were profoundly elevated among non-survivors, presenting a mean value of 3317.73 ± 1025.74 ng/mL versus 578.85 ± 409.63 ng/mL among survivors (p<0.001). While secondary inflammatory reactants, including C-reactive protein (206.73 ± 54.60 mg/L vs. 47.99 ± 37.69 mg/L; p<0.001) and procalcitonin (28.31 ± 21.00 ng/mL vs. 5.29 ± 5.52 ng/mL; p<0.001), were also significantly higher in fatal cases, ferritin demonstrated superior separation. Table 2 Comparison of baseline laboratory parameters and inflammatory biomarkers Biomarker / Laboratory Parameter Survivors (n=48) Non-Survivors (n=22) p-value Hemoglobin, g/dL 10.31±1.14 8.85±1.01 <0.001 Total Leukocyte Count, cells/mm³ 16457.50±4938.56 30214.91±3756.09 <0.001 Neutrophil count, % 19.28±30.10 35.64±39.99 0.042 Lymphocyte count, % 21.50±6.47 9.59±3.54 <0.001 C-reactive protein, mg/L 47.99±37.69 206.73±54.60 <0.001 Procalcitonin, ng/mL 5.29±5.52 28.31±21.00 <0.001 Serum Ferritin, ng/mL 578.85±409.63 3317.73±1025.74 <0.001 *All values expressed as mean ± standard deviation. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic and discriminatory ability of admission serum ferritin concentrations in predicting PICU mortality among pediatric patients presenting with severe sepsis (Table 3). The analysis demonstrated an outstanding area under the curve (AUC) of 0.998 (95% CI: 0.991–1.000; p<0.001), indicating near-perfect discriminatory performance. Using the Youden Index, an optimal admission serum ferritin cut-off value of 1542.0 ng/mL was established. At this threshold, hyperferritinemia predicted sepsis-related mortality with a sensitivity of 100.0% and a specificity of 97.92% (Fig 1). Table 3 Diagnostic performance of admission serum ferritin for predicting PICU mortality Parameter Value Area Under the Curve (AUC) 0.998 95% Confidence Interval (CI) 0.991–1.000 Optimal Cut-off Value 1542.0 ng/mL Sensitivity (%) 100 Specificity (%) 97.92 Positive Predictive Value (PPV, %) 95.65 Negative Predictive Value (NPV, %) 100 p-value < 0.001

admission serum ferritin concentrations are powerfully associated with multiorgan dysfunction, prolonged organ support, and mortality among critically ill children presenting with severe sepsis6-8. The median age of the cohort indicates that the burden of severe pediatric sepsis is distributed widely across infancy, early childhood, and adolescence rather than being isolated to a specific developmental window. This broad age representation and the mild male predominance align closely with multiple published PICU cohorts in the region9,10. The overall mortality proportion of 31.43% observed in this study falls at the higher end of previously documented ferritin-based sepsis studies (which typically report mortality between 26% and 33%)8,10,11. This high mortality is likely attributable to the extreme severity of illness at presentation, the universal presence of cardiovascular compromise among fatal cases, and the high prevalence of septic shock characteristic of admissions to a tertiary referral center12. An important socio-demographic finding was the statistically significant association between rural residence and fatal outcomes. While not uniformly explored in prior biomarker evaluations, this disparity likely reflects systemic pre-hospital challenges, including delayed initial presentation, transport and referral delays, and consequently, more advanced physiological derangement upon arrival at the intensive care unit13. Such factors underscore the necessity of interpreting critical care biomarkers within the context of local referral frameworks. Circulatory collapse and severe respiratory failure were the dominant determinants of mortality in this cohort. The universal requirement for vasopressor or inotropic support and the overwhelming prevalence of ARDS among non-survivors reflect a profound hyperinflammatory phenotype12,13. These clinical manifestations drove significantly prolonged durations of mechanical ventilation and extended PICU stays among non-survivors, reinforcing that severe hyperferritinemia parallels sustained multiorgan failure11,14. Initial hematological parameters further illustrated the severity of systemic dysregulation. Non-survivors exhibited significantly greater degrees of anemia, marked innate leukocytosis, neutrophilia, and profound lymphopenia compared to survivors. This pattern mirrors systemic cytokine-mediated marrow suppression, amplified innate immune activation, and subsequent adaptive immune exhaustion15. While established acute-phase reactants such as C-reactive protein (CRP) and procalcitonin were also substantially elevated in non-survivors, their overall prognostic precision was more variable. CRP functions as a non-specific reactant to broad inflammatory stimuli without reflecting the precise magnitude of macrophage activation, whereas procalcitonin remains highly sensitive to bacterial etiologies but fluctuates depending on the timing of measurement15. Conversely, admission serum ferritin exhibited good prognostic separation, yielding an area under the curve of 0.998. The optimal discriminatory cut-off of 1542 ng/mL established in this study is notably higher than several earlier proposed thresholds (typically ranging from 300 to 500 ng/mL), indicating that extreme hyperferritinemia drives the predictive power in this population6,8,10. Biologically, ferritin is not merely an iron storage protein but an active acute-phase reactant released during intense macrophage activation, oxidative stress, and endothelial injury15. Hyperferritinemia thus identifies a highly lethal hyperinflammatory phenotype. It is critical to acknowledge, however, that these findings contrast with specific regional cohorts; for instance, Williams et al. reported that admission ferritin failed to predict organ dysfunction or mortality in children presenting with sepsis secondary to endemic tropical infections, emphasizing that biomarker performance can vary depending on the underlying infectious etiology4. The AUC observed in our study exceeds most published literature and was likely influenced by cohort homogeneity, the single-center design, the specific timing of admission sampling, and the clustering of profound shock among non-survivors within a modest sample size. The primary strengths of this study include its prospective design within a tertiary PICU setting, ensuring the enrollment of patients with objectively defined sepsis. Restricting biomarker quantification strictly to the time of admission enhances its clinical utility as an early risk-stratification tool prior to extensive therapeutic confounding. Despite these strengths, important limitations must be noted. The study was conducted at a single center with a relatively modest sample size, which limits broader generalizability and predisposes statistical models to an overestimation of diagnostic metrics like the AUC. The cross-sectional design restricts dynamic evaluation, as serial ferritin measurements were not performed to track trajectory over the clinical course. Additionally, despite predefined exclusion criteria, subclinical baseline confounders affecting iron metabolism cannot be entirely ruled out

Elevated admission serum ferritin concentrations serve as a highly reliable, biologically plausible prognostic biomarker strongly associated with multiorgan failure, prolonged mechanical ventilation, and mortality in pediatric severe sepsis. Its rapid turnaround time, wide availability, and cost-effectiveness make it an exceptionally valuable early risk-stratification tool, particularly in resource-constrained critical care settings where advanced multi-variable scoring systems are frequently delayed.

1. Garcia PCR, Longhi F, Branco RG, et al. Ferritin levels in children with severe sepsis and septic shock. Acta Paediatr. 2007;96(12):1829–1831. https://doi.org/10.1111/j.1651-2227.2007.00564.x 2. Horvat CM, Fabio A, Nagin DS, et al. Mortality risk in pediatric sepsis based on C-reactive protein and ferritin levels. Pediatric Critical Care Medicine. 2022;23:968–979. https://doi.org/10.1097/pcc.0000000000003074 3. Taylor MD, Allada V, Moritz ML, et al. Use of C-reactive protein and ferritin biomarkers in daily pediatric practice. Pediatrics in Review. 2020;41:172–183. https://doi.org/10.1542/pir.2018-0101 4. Williams V, Menon N, Bhatia P, et al. Serum ferritin predicts neither organ dysfunction nor mortality in pediatric sepsis due to tropical infections. Frontiers in Pediatrics. 2020;8:607673. https://doi.org/10.3389/fped.2020.607673 5. Schlapbach LJ, Watson RS, Sorce LR, et al. International consensus criteria for pediatric sepsis and septic shock. JAMA. 2024;331:665. https://doi.org/10.1001/jama.2024.0179 6. Anand U, Kumbhar SG. Role of serum ferritin as a prognostic marker in pediatric sepsis. MedPulse Int J Pediatr. 2023;25:7–11. 7. Sharma J, Sharma R. Serum ferritin: a prognostic marker in patients with sepsis in pediatric age group: a prospective cohort study. Int J Med Health Res. 2018;4:86–89. 8. Tonial CT, Costa CAD, Andrades GRH, et al. Prediction of poor outcomes for septic children according to ferritin levels in a middle-income setting. Pediatr Crit Care Med. 2020;21:e259–e266. 9. Sarkar M, Roychowdhury S, Uz Zaman MA, Raut S, Bhakta S, Nandy M. Can serum ferritin be employed as prognostic marker of pediatric septic shock and severe sepsis? J Pediatr Crit Care. 2021;8:20–26. 10. Shaikh GN, Ramamoorthy JG, Parameswaran N, Senthilkumar GP. Serum ferritin for predicting outcome in children with severe sepsis in the pediatric intensive care unit. Indian Pediatr. 2022;59:939–942. 11. Lal M, Goel M, Shrivastava N, Pal PK, Datta M. Serum ferritin levels as a prognostic marker for predicting outcomes in children with severe sepsis and their correlation with pediatric sequential organ failure assessment score. Cureus. 2025;17:e84436. 12. Irvan A, Mariko R, Jurnalis YD, et al. Association of serum ferritin levels with Phoenix score, length of stay and mortality in children with sepsis at Dr. M Djamil Hospital Padang. Int J Res Rev. 2025;12:382–389. 13. Tawari N, Sharma S, Castellino N, Sharma S, Kabra N. Serum ferritin level, a prognostic marker of morbidity and mortality in pediatric intensive care unit in correlation with PRISM III score. Indian J Pediatr. 2023;90:1158. 14. Sucianto A, Pudjiastuti P, Kawuryan DL. The role of ferritin serum level as predictor sepsis mortality on children in Dr. Moewardi Hospital of Surakarta. J Matern Child Health. 2023;8:210–216. 15. Valerie IC, Prabandari AASM, Wati DK. Ferritin in pediatric critical illness: a scoping review. Clin Exp Pediatr. 2023;66:98–109.