Abstract
The limitations of conventional group comparative therapeutic trials are discussed. They include: the heterogeneity problem (due to biological variations within the sample) and the extrapolation problem (i.e. the problem of external validity). These problems may to some extent be overcome by multiple cross-over studies in a single patient. Such studies may be useful in selected cases in daily clinical work, but could also be used for detecting treatment effects in rare diseases. Usually, it is necessary to measure the clinical effect by means of clinical scores, for which reason ranking methods must be used. The statistical significance may be judged by means of a permutation test. The risk of committing a type II error in single patient studies is usually high.

This publication has 1 reference indexed in Scilit: