Admission Serum Ferritin Levels as a Predictor of Severe Disease in Hospitalized Pediatric Patients in Mexico.

Abstract

Biomarkers that predict the severity of a disease on admission would be a valuable tool for pediatricians. Ferritin has been proposed as a predictive biomarker for various conditions; however, multiple thresholds have been described depending on the diagnosis. A threshold for severe pediatric disease regardless of the final diagnosis is lacking. We performed a single-center retrospective observational study of patients admitted to the emergency department (ED) of a pediatric reference hospital in Mexico. Serum ferritin levels were requested to all admitted patients <16 years old and included in the study. Risk prediction models for various clinical outcomes were built using receiver operating characteristic (ROC) curves. Admission ferritin levels were significantly higher in patients who required vasopressors (p < 0.001), blood transfusion (p < 0.001), and admission to the pediatric intensive care unit (PICU; p < 0.001). Patients with admission ferritin >500 ng/mL had a longer hospital stay (p = 0.037). Admission ferritin was a predictor of all-cause mortality (area under the curve [AUC] = 0.67, 95% confidence interval [CI]: 0.54-0.80, p = 0.009; sensitivity 55%, specificity 76.3%). A cut-off point of 385 ng/mL was associated with mortality (OR 3.21, 95% CI: 1.29-7.97). We provide evidence that ferritin may be used as a predictive biomarker in the pediatric ED. In our study, ferritin levels on admission predicted all-cause mortality. Multicenter studies are required to validate our findings in a larger sample.