In this work the methods for synthesizing a decision tree using a complex assessment of a clinical process in decision support systems are reviewed. The main data processing stages are also reviewed. It is noted that only the errors of diagnosis stage are minimized, when synthesizing a binary decision tree using hierarchical clustering in symptom’s area. It’s proposed to make hierarchical clustering in pharmacological area instead of the symptoms area. This is necessary for minimizing a complex error of both the diagnosis stage and the medical rehabilitation stage. Also the necessary metric and the clustering criterion are developed.
diagnosis, rehabilitation, the person making decisions, the decision tree, likelihood ratio, decision rule, clustering criterion, the pharmacological area