[Adequate and inadequate trials in "clinical" pharmacology].

Our definition, based on 14 years of practical experience, is as follows: In clinical pharmacology, a procedure is adequate if it can be used in normal subjects in a non-invasive way and if it generates relevant information concerning pharmacotherapy in patients. Six examples are discussed in detail; the variable body weight may cause wide variations, for example, in absorption characteristics; unhomogeneous groups of subjects are likely to lead to poorly reproducible results. Body position and food intake (individual eating habits) may disguise drug effects by creating additional "noise". By titrating the heart rate during work load on the ergometer in order to achieve and maintain a target rate, individual differences in physical fitness and skill can be eliminated. Orthostatic regulation of cardiovascular variables may be impaired without concomitant psychic lability. The latter indicates predisposition to placebo responses, e.g. in studies using the standardized tourniquet pain model and mild analgesics. In normal subjects computer analysis of cerebral biosignals in combination with psychometric and behavioural tests usually gives more reliable information as compared to the patient pretreated with different drugs.