Selecting outcome parameters in studies aimed at improving rational use of antibiotics - practical considerations

When designing studies of the optimal use of antimicrobial drugs, it is important to define at an early stage, which outcome parameters to use. It must be clear to which category a parameter to be measured belongs: the structure, the process or a clinical outcome. Attention must be paid to the measurement scales as well as the statistical tests to be used. The outcome measure must be sensitive, specific and reliable. Furthermore, the timeframe chosen for the performance of a study should be validated. Whether process parameters or clinical outcomes are to be preferred depends on the settings and the intended purpose. The essential point is the appropriate choice of outcome measure for the study. Examples drawn from the field of quality of use of antimicrobial drugs are discussed.