For evaluation of serological assays, a seropositive reference sample is often arbitrarily assigned a value (EIU) against which test samples are compared. An alternate method, in which seronegative samples are identified by iteration and in turn used to established mean and standard deviation for the quantification (in units of Standard Deviation Ratio (SDR)) of each tested sample, was investigated on 1100 serum samples routinely tested for IgG against Borrelia burgdorferi. Of the 76 specimens that were positive in the EIU system, our computerized SDR method classified 9 as borderline and 67 as positive, none as negative. Of the 917 EIU-negative specimens, the SDR method classified 4 as borderline and 913 as negative, none were SDR positive. We conclude that the SDR has numerous advantages: lower number of borderline specimens found, its computer automation and its more stable results. The method is only suited for larger assays.
Download Full PDF Version (Non-Commercial Use)