Manipulative and inherent errors in anti-D quantitation using the AutoAnalyzer.

The limits of accuracy for the quantitation of anti-D using the AutoAnalyzer have been reduced from +/- 26% with manually prepared dilutions to +/- 14% using an automatic pipette, as expressed by the 95% confidence limits. The error inherent in the AutoAnalyzer was estimated to contribute +/- 10% to the overall error. Problems associated with the reproducibility of this method for anti-D quantitation have been investigated, namely, the effect of the age of a given test cell, the use of different test cells, loss of sensitivity of the machine over a given day, and the reproducibility of results obtained at different positions on the standard graph.