EVALUATING RESULTS OF CLINICAL-TRIALS IN TREATMENT OF COMPLICATED URINARY-TRACT INFECTION

Clinical trials of new drugs should produce results that are not due to chance, inherent differences in groups, or handling of the groups by the investigator. These sources of error may be prevented by having groups of a size that provides sufficient statistical power and thus rules out sample variation. Differences in groups can be prevented by randomization, and double blinding the subjects prevents investigator bias.