B-328 Analytical Performance of a Screening Test for Lead Using Dried Blood Spots and ICP-MS

Abstract

Abstract Introduction There is no safe blood lead levels (BLL) in children and exposure can cause irreversible neurological harm. In 2021, the CDC decreased the blood lead reference value (BLRV) to 3.5 ug/dL from 5.0 ug/dL, based on the 97.5th percentile of BLL among US children under 5 years of age. Despite this, lead poisoning disproportionately affects Black children, those below poverty levels, and living in houses built before 1978. The American Academy of Pediatrics recommends targeted screening for BLL in children aged 12–24 months living in high prevalence areas. Children enrolled in Medicaid also require screening by 24 months of age. Only one-third of children who require screening receive it and lead screening declined by >50% during the COVD-19 pandemic. Quality improvement strategies have included electronic health record alerts and point of care testing. To improve screening rates in our pediatric population, we developed a screening method utilizing dried blood spot (DBS) cards. This study aimed to validate the accuracy and precision of this screening method with emphasis on the performance around the BLRV cut-offs, which would be used to determine the need to follow up with a confirmatory venous lead test. Methods The method was calibrated using certified reference material traceable to NIST SRM 3128 (VHG labs, Manchester, NH, USA) in whole blood. The calibrators, controls and patient samples were spotted onto Whatman 903 cards. A 6 mm filter paper disc in extraction buffer was vortexed, allowed to sit at room temperature for 30 min and centrifuged. The samples were analyzed on the Thermo Fisher iCAP RQ and TQ (Whaltman, MA) in kinetic energy discrimination mode. The assay has a linear range of 1–100 µg/dL, and the recovery ranged from 92%–102%. We evaluated this method for reproducibility, and accuracy across the measurement range. In this study, we focused on the reproducibility and accuracy of DBS at lead concentrations around the BLRV cut-offs. We used in 2 samples spiked with NIST CRM 3128 (Inorganic Ventures, Christiansburg, VA, USA) analyzed 10 times over 3 days and 13 leftover patient samples run over at least 5 days. The total allowable error was 2 µg/dL or 10%. Results In two samples spiked to BLL of 3.5 and 3.9 µg/dL, interday precision (n = 10) ranged from 3.4% to 17.4% (SD 0.1–0.7) in samples run for 3 days. Removal of an apparent outlier decreased the %CV ranged observed to <8.5%. The average bias for each day ranged from 0.1 to 0.6 µg/dL. In 13 patient samples with concentrations around the BLRV ranging from 1.4 to 5.8 µg/dL, the coefficient of variation (CV) ranged from 0.2% to 32% with the largest SD at 0.8 µg/dL. The average bias for this sample set was -0.3 (11.7%). Conclusion The assay’s analytical performance is sufficient for screening purposes, but the clinical performance and the impact of introducing this test to enhance screening rates during 1 and 2 years well visits need to be evaluated further.