[Research based on clinical observations].

The author tries to define the field of "Clinical observations research". According to this definition there are three main causes for several false developments in medicine: the valuation of observations and statements on the level of hypothesis, the identification of statistical correlations with causal connections, and the stopping of investigations on the level of twodimensional analysis. To get a high level of probability of results in "Clinical observations research" it is necessary to use methods of multivariante analysis, as it is demonstrated in three examples from the field of general surgery and traumatology.