Monitoring human metabolism is of crucial importance for personalized medicine. Many metabolic diseases can be treated by controlling various endogenous metabolites (e.g., glucose, lactate, cholesterol, etc.). Similarly, measuring the metabolism of exogenous compounds (e.g., etoposide, ftorafur, cyclophospamide, etc.) can enhance the effectiveness of a therapy as applied to the individual patient, since the response rate of different patients to the same pharmacological treatment and dose typically varies widely. The objective of this work is the systematic study of the use of electrochemical readout for advanced diagnosis and drug monitoring. Whereas to date various electrochemical principles have been studied and successfully tested, they typically operate on a single target molecule and are not integrated in a full data measurement and analysis chain. Our work addresses simultaneous multi-target detection as well as full integration of biosensors and readout electronics in silicon-based realizations that can be implanted in animals and humans. Our sensors exploit as probes both the protein class of oxidases - for sensing endogenous metabolites - and cytochromes P450 - for sensing drug compounds. We fabricated nanostructured electrodes that use carbon nanotubes (CNT) as intermediary between the probes and the electrodes to achieve higher sensitivity and lower detection limit and we developed circuits for in-situ data analysis. We developed an integrated platform to support multiple applications within the same architecture, thus extending the scope of utilization of the biosensing analysis chain.