Sample commutability in external quality assessment surveys (EQAS) for thyroid-related antibodies--state of the art.

External quality assessment surveys for thyroid-related antibodies have been offered by INSTAND for 20 years. During this time, some problems have remained, especially those between the similarity of samples sent and routine patient samples. Here the questions of "matrix effects" and "commutability of results" are topics discussed at most EQA-meetings. This short communication deals with the effects of lyophilization on results for thyroid-associated antibodies in an EQA-survey carried out by over 230 participants in October 2005 by INSTAND in Düsseldorf, Germany. The results show that there are small but statistically significant differences (at the level p < 0.05) in results from liquid and lyophilized samples from the same source, the tendency to higher results in the lyophilized samples for anti-TPO and anti-Tg and to lower results for TRAb. The precision in measurement was significantly better for anti-TPO and TRAb in the lyophilized samples, there being no significant difference for anti-Tg. The differences in results between liquid and lyophilized samples were minimal when compared with the numerical results for anti-TPO and anti-Tg, despite the fact that all kits were calibrated with the NIBSC reference materials, although the latter are now both 40 years old.