The current chapter gives an example of how, with the help of advanced statistics, pretty convincing quantitative conclusions can be drawn about drug risk and safety measurements.
For the purpose confidence intervals of ratios of normal variables are needed. Using the quadratic method, these confidence intervals can be approximated. In order to compute the variance of the ratio of two paired terms the underneath equation is required. In an example assessing the chance of intolerable dizziness of different treatment modalities for angina pectoris, the benefit risk ratio was estimated from