This multicenter study tested the actuation of a new model of pharmacovigilance, focused on three pharmacological wide-used categories (non-steroidal anti-inflammatory drugs, NSAID, oral anticoagulants, and antihypertensive drugs). Besides the traditional way of pharmacovigilance, an active investigation was performed, using a phone-structured interview. Patients discharged from the participating hospitals were included into the study, if their prescribed therapy included some of the above drugs and after informed consent. Three hundred subjects were interviewed, 100 for each pharmacological category. For a period of six months after patient's discharge from the hospital, a traditional pharmacovigilance survey was carried out. About 30 days after discharge from the hospital, patients were interviewed by the medical staff and data recorded. NSAID group stratification evidenced a significant percentage of severe haemorrhage among the patients who were using acetylsalicylic acid (ASA) as antiaggregant (6.8%) compared to the patients who were using non-ASA NSAID, at therapeutic dosage (1.8%). From this data, it seems that the active pharmacovigilance model was able to better highlight a real problem for the NSAID category, in particular it evidenced a pharmacological subclass (ASA) more prone to cause ADR than expected from literature data related to whole pharmacological class. Given the required economical effort, this pharmacovigilance method could take place as a selected tool when pharmacovigilance signals from the international databases become consistent or for new wide-used drugs, to screen potentially dangerous pharmacological subclasses, normally “hidden” because of a “camouflage” among ADRs of the entire pharmacological class.