Clinical Decision Support Systems (CDSSs) have been one of the challenging real-world applications of artifi- cial intelligence techniques for decades. CDSSs built so far mainly deal with explicit medical knowledge, typically encoded into a knowledge base. The rationale is that medical personal knowledge may be updated or time constraints limit physicians’ ability to properly manage that knowledge. However, inter-user variability due to subjectivity in feature evaluation has been typically ignored when building clinical decision support systems. In some cases implicit medical knowledge can be important to share, especially for novices’ education. In this work we present an approach for building CDSSs facing this problem through different artificial intelligence techniques: a machine learning approach for a critiquing module and a case based reasoning approach for a consulting module. The purpose is to build a system accounting for the specific skill an expertise of the user, providing new or enough information to take proper decisions.