Association between serum phosphate levels and outcome among critically ill newborns: a retrospective cohort study.

BACKGROUND: This study aimed to explore the relationship between serum phosphate levels and clinical outcomes in critically ill neonates. METHODS: A retrospective cohort study was conducted using the Pediatric Intensive Care (PIC) database from (2010- 2018.) Multivariable regression models, generalized additive models, and threshold effect analysis were adopted to evaluate the relationship between serum phosphate levels within 24 h of ICU admission (Pi24) and neonatal outcomes. RESULTS: A total of 1,902 critically ill neonates were enrolled. Pi24 was stratified into three tertile groups (Q1-Q3). Baseline characteristics and Kaplan-Meier curves showed that higher Pi24 correlated with higher 28-day mortality, peaking at 11% in the Q3 group. Multivariate Cox proportional hazards regression analysis revealed that, Q3 neonates had a 76% increased risk of 28-day mortality compared with Q1(HR = 1.76, 95%CI:1.08∼2.88, P = 0.025), Multivariate logistic regression indicated an 80% higher odds of in-hospital death in Q3 vs. Q1(OR = 1.80, 95%CI: 1.1-2.95, P = 0.019). Within Pi24 ranged from 1.76 to 4.37 mmol/L, each 1.0 mmol/L elevation in serum phosphate level increased in-hospital death risk by 45.6% (OR = 1.456, 95% CI: 1.039-2.039, P = 0.029). Multivariate linear regression analysis revealed that Q1 neonates had significantly longer hospital LOS than Q3 (β = -4.68, 95% CI: -7.84∼-1.53, P = 0.004). Threshold effect analysis further demonstrated that when serum phosphate was < 2.5 mmol/L, each 1.0 mmol/L reduction prolonged hospital LOS by 6.89 days (β = -6.892, 95% CI: -10.801∼-2.983, P = 0.001). CONCLUSION: Elevated Pi24 levels were associated with increased risks of 28-day mortality and in-hospital death in critically ill neonates, while lower Pi24 levels independently predict prolonged hospital LOS Pi24 may represent a valuable prognostic biomarker for this population.