The use of centrally prepared reagents in an external quality experimental trial

The authors provide the results of a short-term experimental trial in external quality assessment in 42 clinical laboratories conducted by the Mexican Ministry of Health. Assay kits for glucose, urea, and creatinine were prepared by the Ministry. The results may prove useful to organizers of external quality assessments in third-world countries who may opt for this strategy to improve performance. The laboratories performed the tests on reconstituted lyophilized control serum, also prepared by the Ministry. All three assays were performed manually using colorimetric methods. On the basis of their intralaboratory precision (coefficients of variation less than 8%, 8-12%, and greater than 12% for high, medium, and low precision, respectively), 12 laboratories demonstrated high precision for all three tests. Eight laboratories showed medium and low precision for different tests, while the other 22 fell in between. The results showed that the strategy of using centrally prepared reagents to improve interlaboratory agreement did not work well for urea and creatinine, but met expectations for glucose. The laboratories achieved an interlaboratory coefficient of variation of 10% for glucose in this first trial.