|Home | About | Journals | Submit | Contact Us | Français|
In Norway, through our work in a general practice research committee under the Norwegian Medical Association, we have been involved in a voluntary arrangement with pharmaceutical companies where we evaluate research projects initiated by the industry and recruit general practitioners to find patients.1 Our committee makes recommendations to Norwegian general practitioners about whether a project is a real research project rather than marketing camouflaged as research.
However, after we criticised some projects in 2001, this arrangement has been largely boycotted by the industry. One objection in one particular project was that the protocol favoured the company's own drug. We also maintained that exaggerated demands for statistical power led to a grossly overestimated number of people participating. Research ethics should safeguard that type 2 error is prevented, but statistical power beyond reasonable limits (95% or greater power), leads to research where almost any difference reaches significance. According to our view, this also represents a violation of ethical demands in research. Apparently, it is not evident that research committees have expertise that takes such considerations into account. In correspondence in 2001 with the national research ethics committee in Norway, we recommended that ethics committees routinely should ask for and require a positive answer to these two questions:
Our committee has a checklist for quality assurance of multicentre drug research in general practice. All questions require a positive answer. The most important demands for approval are:
Competing interests: None declared.