Human serum albumin: prediction model and reference values for preterm and term neonates

Area of Science:

• Neonatal pharmacology
• Clinical chemistry
• Pharmacokinetics

Background:

• Human serum albumin (HSA) levels influence drug distribution in neonates.
• Understanding longitudinal HSA trends is crucial for neonatal care.
• A predictive model for HSA concentrations in neonates is needed.

Purpose of the Study:

• To describe longitudinal, real-world human serum albumin (HSA) trends in neonates.
• To develop a multivariable prediction model for HSA concentrations in a neonatal cohort.
• To provide population-specific HSA centiles for clinical application.

Main Methods:

• Retrospective analysis of neonates in a neonatal intensive care unit (postnatal age ≤28 days).
• Linear mixed models used to explore covariate effects on HSA.
• A multivariable prediction model developed using backward selection.

Main Results:

• 848 neonates included; median HSA concentration was 32.3 g/L.
• HSA concentrations increased with postnatal age and gestational age.
• A high-performance (R²=76.3%) multivariable HSA prediction model was developed, with provided HSA centiles.

Conclusions:

• Developed population-specific HSA centiles and an accurate neonatal HSA prediction model.
• The model incorporates maturational and non-maturational covariates for precise HSA prediction.
• Results enhance clinical care and pharmacokinetic analyses for HSA-bound drugs in neonates.