Abstract
The preparation of a stable and inexpensive equine haemoglobin solution for quality control of haemoglobin measurement is described. Twenty-one different batches of the preparation were analysed by 19 hospital laboratories during a 19-month period, and the results of 4457 determinations are evaluated. Systematic divergences between laboratories were found to contribute to the variability of haemoglobin measurement more than random variation. Interlaboratory variation was mainly due to different types of automated techniques, the frequency of the calibration of photometers, the use of different reagent solutions, and the qualification of laboratory staff. Intralaboratory variation was related to the frequency of the calibration of photometers and the qualification of the laboratory staff. The data reported in this study indicate the usefulness of the test preparation for the quality control of haemoglobin measurements.