Article Text
Abstract
Since July 1969, portions of the same blood serum have been dispatched to clinical chemistry laboratories in the United Kingdom at 14-day intervals. The results of each serum survey were reported to each of the 390 participants within 11 days of their originally receiving the specimen.
During the first 18 months of the survey no overall improvement in the results was seen. Therefore a summary of each laboratory's ability consistently to produce results close to the mean of the method used was calculated and reported as a single figure, the variance index, and sent to all participants at regular intervals together with a histogram distribution of the variance indices of other participants. The subsequent improvement in the overall results is described.