Biological variability and predictability: an exercise in plausibility

The inherent variability in the responses of the human body to stimuli means that it is often necessary to utilize statistical methods to evaluate causal relationships in clinical research. It is necessary, however, to be cautious with statistical methods and it is imperative that scientific plausibility is included in the analysis. This article addresses some of these issues in relation to the assessment of so-called adverse effects associated with medical devices.