Influence of laboratory test volume and geographic location on maternal alpha-fetoprotein results. Information from the College of American Pathologists Maternal alpha-fetoprotein Survey.

Abstract

Using 1989 College of American Pathologists Maternal alpha-fetoprotein (AFP) Survey data, we found virtually no association between a laboratory's AFP test volume and the reliability of reported multiples of the median uncorrected for maternal weight, race, or diabetes. We also found no differences in AFP gestational age-specific medians across geographic regions of the United States. We did find that clinically important adjustments for maternal weight, race, and diabetes were made more frequently by laboratories with higher clinical AFP test volumes. Consequently, we believe that a minimum AFP test volume for satisfactory performance of AFP testing, as proposed by several professional organizations, cannot be justified on inability of laboratories with low AFP test volumes to quantify AFP in clinical specimens or calculated uncorrected multiples of the medians accurately. However, high-volume laboratories enhance clinical utility of the reported AFP results by making clinically important adjustments of their reported multiples of the medians, and they tend to provide more accurate clinical interpretations. Therefore, we recommend that efforts to improve the quality of AFP testing nationally should focus on improving the ability of laboratories to provide appropriate adjustments to and clinical interpretation of their AFP results, in addition to more traditional efforts directed toward quality control of the analytical aspects of the AFP assay.