[Regression analysis of the relation between antitoxic activity of sera (Jensen method) and antibody titers in the...

Antidiphtherial and antitetanus antitoxic immunity was determined in the sera of 1588 children (20 experiments) and 611 children and adults (12 experiments) by two methods--on animals and in the passive hemagglutination test (PHT). Factual data were grouped in accordance with a conditioned scale of the antitoxin values in IU composed of the values average for each interval. The series of two-fold dilutions represented a scale of titres of the same sera in the PHT. Mathematical analysis showed that there was a direct linear regressive relationship between the indices of antitoxic immunity obtained by different methods. As a rule, regression coefficients were below a unit, i. e. the rate of increase of the values by the HDT titre scale was somewhat retarded from the growth by the IU titre, this pointing to the difference in their value. The free member of the regression equation (a) characterizing this choice of the sera varied widely. Between the regression coefficients calculated for individual experiments statistically significant differences were revealed in almost 50% of cases. Transformation of quantitative indices of one test into quantitative characteristics obtained by the other method, with the aid of regression equation calculated outside the definite experiment, proved to be unfounded.