A proposal for the abandonment of activity units for most plasma enzymes.

Although Vernet-Nyssen does not agree, in principle, that different methods of measurement should give different results, in practice, however, such differences are often observed, either as a consequence of specimen preparation or as a result of analytical technique. In the previous Letter (Clin Chem 28: 544, 1982) we communicated our results based on external quality control schemes involving lyophilized samples only. As we have no experience with liquid controls, we cannot comment on these. Vernet-Nyssen agrees with us that pH plays a major role in reproducibility and accuracy of TIBC determinations, and furthermore, calls to our attention the fact that lyophilized sera have a slower binding kinetics than native human serum. As a consequence, she suggests a 10-mm incubation step to achieve complete saturation. Such a step would not be appropriate for quality-control purposes, as it will draw a special attention to quality control material but not other specimens. However, such a step would be appropriate if all samples were treated similarly. But is it a valid modification to change a routine methodology to accommodate external quality control material? If liquid controls are found satisfactory for internal quality control schemes, the problem of handling samples from external quality control schemes would remain.