[Clinical trials--methodological gold standard or naive reductionism?].

The methodological concept of the randomized controlled clinical trial is a relatively recent invention. It is widely accepted to be the only means of conclusively establishing a cause effect relationship between a given therapeutic intervention and an observed outcome. There are, however, numerous pitfalls. Ethical concerns and/or practical problems can seriously affect both proper randomisation as well as the use of a placebo treatment as the control intervention. Sometimes blinding of the patients or the therapist may be difficult or even impossible, depending on the availability of a similar treatment without the specific therapeutic effect, whereas a satisfactory investigator blinding can be achieved in most instances. Practical problems can be minimised by adapting the trial design to the given clinical situation. The fact that a number of drawbacks exist, must, however, not be used as an argument to abstain from performing such studies--simply because there is no better alternative.