We report a method for calculating allowable error limits of laboratory data based on biological variation among individuals. In Japan, while the biological variation of biochemical data has been reported from several institutions, a description of the biological variation of coagulation tests has not yet been reported. In this study, we calculated the biological variations for 14 coagulation tests, 6 hematological tests, and 28 biochemical tests in the Laboratory Division, Kyushu University Hospital, and determined allowable error limits of intra-institutional variation and inter-institutional variation. The subjects chosen for the study were 20 healthy individuals consisting of 3 male and 3 female subjects in their 20's, 2 male and 2 female subjects in their 30's, 1 male and 4 female subjects in their 40's, and 3 male and 2 female subjects in their 50's. Measurements were performed before breakfast 12 times monthly between June 2001 and May 2002, and allowable error limits were calculated according to the method of Fraser CG. When compared with the coefficients of variation of control substances used in daily laboratory work, the allowable error limits of intra-institutional variation determined in this study appeared to be in the attainable range.

View full abstract