ABSTRACT
Testing effectiveness of poliomyelitis vaccine. In 1954 the National Foundation for Infantile Paralysis contracted with the University of Michigan to conduct a nationwide test of the effectiveness of the Salk vaccine for poliomyelitis (or polio). It was called infantile paralysis in earlier days because of its crippling effect and its concentration among children. It was thought that an extensive study was necessary before the vaccine was put into general use. The first vaccine that seemed effective against polio was developed by Salk and the purpose of this study was to test its effectiveness. 1Wo reports are of interest: the summary report2 and an independent appraisal of the techniques used in the study:3
At that time, the average rate of occurrence of polio was about 50 per 100,000 children. Thus, it was necessary to use large samples of children to obtain a difference that could be attributed to the vaccine and not to sampling fluctuations. Assuming the vaccine is 50 percent effective, a sample of 100,000 vaccinated children would have about 25 cases; a sample of 100,000 children not vaccinated would have about 50 cases. This reduction would be strong evidence the vaccine is effective. This can be shown by assuming 25 and 50 are Poisson counts. Then the normal deviate is:
which is significant at the 99.8 percent level. (See Table 5, Chapter 25). Actually; slightly more than 200,000 children were used in the experimental group and about the same number were used in the placebo (no vaccine) group. See Table 1.
