This document discusses how medical data mining of Danish healthcare registries can be used to link diseases, drugs, and adverse reactions. It describes how the Danish registries contain structured and unstructured data from hospitals and central registries dating back to 1968. Text mining of unstructured clinical notes allows identification of diseases, drugs, and possible adverse drug reactions. Analysis of this data through methods like comorbidity, diagnosis trajectories, and temporal correlations can provide new insights into disease relationships and help discover new adverse drug reactions and their frequencies.