Pediatric Glucose-6-Phosphate Dehydrogenase Reference Intervals Derived from Archived Phenotype and Genotype Test Results.

Glucose-6-phosphate dehydrogenase (G6PD) deficiency is the most prevalent global enzymopathy with risks of hemolytic anemia and hyperbilirubinemia. G6PD activity peaks at birth and decreases during the first year of life. Therefore, partitioned reference intervals (RI) are important for diagnosing G6PD deficiencies in pediatric patients. Given the paucity of pediatric RIs and difficulties in direct sample collection from healthy children, indirect RI methods were utilized to evaluate RIs for G6PD activity in this population. We established accurate RIs and percentage of normal activity thresholds (PNAT) for G6PD activity in children using refineR and genotype data. G6PD activity (U/g Hb) from patients <18 years old tested in our laboratory were used to derive RIs and PNATs. refineR with 1-parameter Box-Cox transformation was used to determine RIs and medians for PNATs. Bootstrapping was used to calculate confidence intervals for the estimated RIs. Retrospective analysis of G6PD activity results and correlating G6PD genotypes, were used to verify the estimated RIs and PNATs. Age-partitioned 2.5th, 50th, and 97.5th percentiles (point estimate, U/g Hb) for children were established from a dataset of 32 576 results <8 days (16.4, 19.9, 24.1), 8-<31 days (13.6, 18.2, 23.9), 1-<7 months (11.5, 16.2, 21.5), 7-<13 months (10.3, 13.7, 17.9), and 1-<18 years (9.7, 12.9, 16.9). Alignment of RIs and PNATs with genotype classification (n = 105) showed 96% and 93% agreement, respectively, in unaffected, heterozygous, and deficient individuals across age groups. Indirect RIs for G6PD in children from a noncurated real-world dataset showed strong alignment with genotype classification.