Novel pediatric within-subject biological variation estimates derived from laboratory data using an indirect method

Abstract Objectives Biological variation (BV) is critical for interpreting lab results and setting analytical performance specifications. However, within-subject BV (CV I ) data are limited in pediatric populations. This study aimed to estimate pediatric CV I values for selected lab tests using an indirect method. Methods In this retrospective study at Umraniye Education and Research Hospital (Istanbul), 69,984 results from 11,535 pediatric patients (0–19 years) over 20 months were analyzed for CBC and selected biochemical analytes (glucose, urea, creatinine, ALT, AST) within reference intervals. Patients with 5–15 results were included, with outliers (Q test) and trends (regression) excluded. Person-specific CVs were calculated and then further improved by removing values outside the 2.5–97.5 percentile of the Gaussian distribution, combined to obtain CVpooled, and adjusted by subtracting analytical variation (CV A ) to estimate CV I , which was subsequently compared with CALIPER BV and EFLM-BVD data using the <0.5 or >1.5 ratio threshold. Results Overall, pediatric CV I s estimated by the indirect method were generally high. For CBC parameters, Hgb, Hct, and MCHC ratios were within the threshold (1.0±0.5), whereas others exceeded it; for biochemical parameters, all ratios compared to CALIPER exceeded the threshold. Conclusions This study provides novel pediatric CV I estimates using an indirect method. Elevated CV I values may reflect physiological variability and pre-analytical factors in children. Indirect approaches are useful alternatives for BV estimation where direct sampling is challenging.