Plasma Vitamin C, 25-Hydroxyvitamin D, and the Lactate-to-Albumin Ratio at PICU Admission in Survivors and Non-Survivors of Phoenix-Defined Pediatric Sepsis and Septic Shock: An Exploratory Prospective Cohort Study

Background: Vitamin C, 25-hydroxyvitamin D (25[OH]D), and the lactate-to-albumin ratio (LAR) have been proposed as prognostic markers in sepsis. However, pediatric data are limited, and these markers have not been evaluated together in a cohort classified according to the Phoenix criteria. We quantified these markers at pediatric intensive care unit (PICU) admission and assessed their unadjusted association with mortality. Methods: This single-center prospective cohort study followed the STROBE statement. We enrolled consecutive children aged 1 month to 17 years with suspected infection and a Phoenix Sepsis Score of at least 2; septic shock required at least one cardiovascular point. Blood was sampled within 1 h of admission. We analyzed ascorbate by ultra-high-performance liquid chromatography coupled with mass spectrometry in light-protected heparin plasma. The endpoint was pre-discharge PICU mortality. We expressed differences as Hodges–Lehmann estimates with exact 95% confidence intervals (CIs), and discrimination as the area under the receiver operating characteristic curve (AUC). The five principal biomarker comparisons were subjected to Benjamini–Hochberg false discovery rate (FDR) control. We fitted no multivariable model (11 events). Results: Of the 37 children enrolled, 33 (89.2%) had septic shock, and the mortality rate was 29.7%. Non-survivors had lower plasma vitamin C concentrations (51.1 vs. 73.8 umol/L; difference −17.0, 95% CI −28.4 to −5.7, p = 0.028, q = 0.141, AUC 0.73, 95% CI 0.55–0.90), and mortality declined across ascending vitamin C tertiles, although the gradient was not statistically significant (p = 0.058). No patient had hypovitaminosis C (<23 umol/L). We did not detect an association between 25(OH)D, lactate, albumin, or the LAR and mortality (all p ≥ 0.18; AUC 0.50–0.64), and the LAR was collinear with lactate (rho = 0.95). The weight-for-age z-score correlated with vitamin C (rho = 0.39, p = 0.033). Conclusions: In this small cohort, non-survivors had lower plasma vitamin C concentrations within the replete range in unadjusted analysis; this difference did not persist after FDR control or in the subgroup with evaluable nutritional status, and nutritional confounding cannot be excluded. No association with mortality was detected for 25(OH)D, lactate, albumin, or the LAR. These findings are hypothesis-generating, do not establish prognostic value, and require multicenter confirmation.