Predicted effects of proposed tightened acceptance criterion for Pb proficiency testing (PT) on PT sample pass rates for Pb point-of-care testing

Background There is current advocacy for change in Pb proficiency testing (PT) acceptance criterion from ± 4 μg/dl ([Pb] < 40 μg/dl; criterion a) to ± 2 μg/dl ([Pb] < 20 μg/dl, criterion b). We examined the effect of this proposed change on PT sample pass rates for point-of-care testing (POCT) as predicted by imprecision of POCT PT sample results. Methods Inter-site standard deviations (s) of POCT PT results were tabulated as a function of [Pb] and characterized as a linear function of [Pb] (r² > 0.8). Given s, predicted minimum, random-error-only PT failure rates (Fp) as a function of [Pb] were computed as the fraction of a normal distribution of results ([Pb] ± s) that would fall outside of boundaries of acceptance criterion a or b. Results For [Pb] = 2–20 μg/dl, current observed PT sample failure rates using criterion a range from 3 to 6%, which are greater than the predicted minimum failure rates based on s alone (Fp(a) = 0–6%). In contrast, predicted minimum failure rates based on s using criterion b are greatly increased (Fp(b) = 5–35%). Conclusions Given the degree of inter-site imprecision among POCT Pb PT results, adoption of criterion b for PT acceptance will dramatically increase Pb PT sample failure rates for POCT due to random-error alone.