Pediatric Reference and Optimal Curves for Hemoglobin

Area of Science:

• Pediatric Hematology
• Clinical Biochemistry
• Biostatistics

Background:

• Reference intervals (RIs) are crucial for interpreting pediatric laboratory results.
• Traditional RIs require age and sex partitioning, leading to limited sample sizes and imprecise estimates.
• Hemoglobin level modeling as a continuous function of age addresses these limitations.

Purpose of the Study:

• To establish sex-specific hemoglobin reference curves (RCs) for children.
• To estimate hemoglobin optimal curves (OCs) using health-based criteria.
• To complement World Health Organization (WHO) hemoglobin thresholds and inform reporting standards.

Main Methods:

• Cross-sectional study of healthy Canadian children aged 2 weeks to 10 years.
• Nonparametric quantile regression with restricted cubic splines to estimate RCs and OCs.
• Development of a web-based platform for visualizing curves and calculating limits.

Main Results:

• Hemoglobin RCs and OCs were estimated for 4597 and 3426 children, respectively.
• Optimal curves (OCs) showed sex-specific differences compared to reference curves (RCs), particularly in younger children.
• World Health Organization (WHO) thresholds were often higher than the lower limits of the OCs, especially in younger age groups.

Conclusions:

• Modeled hemoglobin RCs overcome limitations of traditional pediatric RIs.
• Optimal curves (OCs) provide additional clinical context beyond RIs.
• Findings suggest potential misalignments between new curves and existing WHO thresholds, particularly for younger children.