Admission biomarkers associated with pediatric intensive care unit admission in children with multisystem inflammatory syndrome: an exploratory retrospective cohort study
Abstract
Multisystem inflammatory syndrome in children (MIS-C) may rapidly progress to respiratory or hemodynamic compromise requiring pediatric intensive care unit (PICU) admission. We investigated the associations between routinely available admission biomarkers and PICU admission and explored cohort-specific candidate thresholds. This retrospective single-center study included 107 consecutive children with MIS-C admitted to a tertiary pediatric referral center between March 2020 and December 2022. The first laboratory values obtained at presentation to our center, before intravenous immunoglobulin, corticosteroids, or other MIS-C-specific immunomodulatory treatment administered at our center, were analyzed. Exploratory receiver operating characteristic (ROC) analyses using Youden’s J index were performed for procalcitonin, C-reactive protein (CRP), ferritin, albumin, and platelet count. Candidate thresholds were derived in this cohort and require independent validation before clinical application. Sixteen patients (15.0%) were admitted to the PICU, and two (1.9%) died. All PICU patients required oxygen therapy, nine (56.3%) received inotropic support, and two (12.5%) required invasive mechanical ventilation. PICU patients had higher admission procalcitonin, cardiac troponin I (cTnI), CRP, ferritin, lactate dehydrogenase (LDH), and aspartate aminotransferase (AST) levels and lower albumin concentrations, platelet counts, and absolute lymphocyte counts than ward patients. Exploratory ROC analyses identified candidate thresholds of procalcitonin ≥ 33 ng/mL, CRP ≥ 180 mg/L, ferritin ≥ 899 ng/mL, albumin ≤ 2.98 g/dL, and platelet count ≤ 123 × 10³/µL; corresponding area under the ROC curve (AUC) values were 0.83, 0.78, 0.77, 0.79, and 0.77, respectively. Routinely available admission biomarkers were associated with PICU admission in children with MIS-C. The ROC-derived thresholds are cohort-specific and require independent validation before clinical application. Prospective multicenter studies using standardized assays and contemporary populations are needed to confirm their reproducibility. Not applicable.