Study design is a key issue when initiating a new study for the specific purpose of further investigating a drug safety issue (i.e. hypothesistesting). Important specific issues which should be addressed are: (1) clear specification of the hypothesis to be tested; (2) achieving adequate statistical power to detect as significant clinically important risks compared with those attributable to therapeutic alternatives (a power calculation should always be performed); (3) minimisation of bias (primarily through processes for selecting subjects and collecting exposure and outcome data); and (4) control of confounding in the design (by matching) or analysis (by adjustment, normally using multivariate statistical techniques).
In addition to study design there are many logistic issues to be addressed regarding the processes for assuring the quality of the data and the validity of the study. Particular consideration should be given to potential ethical issues and to how the data are going to be monitored if the study is prospective. The guidelines cited above encourage pharmaceutical companies to set up independent advisory boards to oversee such studies. Advisory boards are now becoming the norm for most studies that are relevant to drug safety.
Was this article helpful?