Effect of bias in hematofluorometer measurements of protoporphyrin in screening programs for lead poisoning.

Values for erythrocyte protoporphyrin (EP), measured in our laboratory after extraction with ethyl acetate-acetic acid, were compared with hematofluorometer measurements made in 21 other laboratories. We found that: (a) for samples of patients' blood, hematofluorometer results were 11 to 28% lower than the extraction-based values, depending on the concentration of EP and the mathematical model used; (b) hematofluorometers had mean errors of 0 to 3% for federal proficiency-testing samples; (c) there were no performance differences between fresh and shipped blood for the six laboratories that were analyzing both; (d) a hematofluorometer with a 20% low bias at an EP concentration of 500 micrograms per liter of whole blood (by the extraction method) will not detect about a third of the children whose EP concentration exceeds that cutoff value; and (e) at this same cutoff value for EP, the extraction tests detects about 45% of children whose blood lead exceeds 300 micrograms/L, whereas a 20% low-bias hematofluorometer detects only about 37%.