Abstract
By means of two group experiments, performed in seven laboratories and using six different
modifications in the insulin radioimmunoassay technique, it is shown that the use
of standards diluted in human serum of low basal insulin concentration, instead of
buffer, improves appreciably the interlaboratory precision. The causes lies perhaps
in some serum factor, which could disturb some modifications of the radioimmunoassay
more than others. There was no improvement of precision when the same insulin standard
was used in all laboratories. Standard curves made up in different human fasting sera
have a scatter only within the range of pipetting and counting error.
Key words
Group Experiment - Insulin RIA - Data Processing - Quality Control - Insulin Recovery
1 Supported by the Deutsche Forschungsgemeinschaft (SFB 51, München, and SFB 87, Ulm).
