The method and the device are used for evaluating recorded measurement data. In particular, the use of the method and of the device makes it possible to detect the individual risk of a living being in respect of certain disease states and, if appropriate, to automatically evaluate correspondingly recognized risks in a device control. Measurement data in respect of heart-specific and/or circulation-specific parameters or respiration-specific parameters of a patient are preferably recorded and evaluated. The measurement data are determined over a defined period of time and the measurement data are evaluated taking into account the time information. A particularly effective implementation of the method can be based on pattern recognition in which individual patterns and/or sequence patterns are evaluated.