An Analysis of the Wilson-Worcester Method for Determining the Median Effective Dose of Pertussis Vaccine

Publisher Website
Google Scholar
Abstract
Results of "intracerebral" protection tests of pertussis vaccine were analyzed by the Wilson-Worcester and the Probit methods. The statistics were in complete agreement. The scatter diagram of the points, obtained by plotting the respective median effective doses, could be fitted by a single straight line which passed through the origin. The results were also analyzed by the Reed-Muench method. A similar graphic comparison of this ED 50 with the Probit one gave a scattering of points. One line could not be fitted to all points, thereby showing that some of the Reed-Muench statistics were biased. Tests that were rejected by the Probit method as being inconclusive were likewise rejected by the Wilson-Worcester method. The latter method appears to be adequate for use in the assay of pertussis vaccine. It provides an exact statistic with a standard error and a formula for determining the homogeneity of response to 2 vaccines; and the time consumed for calculation is about 0.1 of that used by the Probit method and about the same as by the Reed-Muench method. Advantageous information furnished from analysis by the Wilson-Worcester method in contrast to that by the Reed- Muench method is emphasized by an example.

This publication has 13 references indexed in Scilit: