Improvement of the Kleihauer-Betke Test by Automated Detection of Fetal Erythrocytes in Maternal Blood

Abstract

Background: Reliable detection and quantification of fetal red cells in maternal blood is important in routine obstetric practice. The manual Kleihauer-Betke test (KBT) is widely used, but it is imprecise and subjective. This study investigated whether automated readout of the KBT could improve sensitivity and accuracy. Methods: Glass slides containing dilutions of fetal red cells in adult blood were prepared and stained by using acid elution. Standard manual evaluation of the KBT was performed for all slides by one investigator. In addition, automated microscopy and image analysis of the same slides were performed, whereby detected fetal cells were reviewed by two independent investigators. Results: Ten replicate measurements of fetal cell numbers showed high reproducibility and very small interobserver and intraobserver variabilities. Typical coefficients of variation were 3-4% for concentrations ranging from 0.001% to 0.1%. The automated KBT showed strong correlation between theoretical and detected concentrations of fetal cells (r2 = 0.999). In the range from 0.0001% to 0.001%, the standard KBT underestimated the fetal cell percentage, whereas the automated KBT was very precise. The correlation between methods was good (r2 = 0.999). Conclusion: Automated readout of the KBT improved accuracy of fetal cell detection in the range from 0.0001% to 1% fetomaternal hemorrhage, particularly when larger numbers of cells were analyzed. © 2003 Wiley-Liss, Inc.