This article describes findings about effective utilization of electronic medical databases from a viewpoint of a pharmaceutical company, with a progress of MIHARI Project in mind. Purposes of this article are (1) to understand merits to perform a database study and a drug use result survey as an additional pharmacovigilance activity for each risk in a risk management plan, and (2) to find points to be considered when their activities are to be selected. A database study has advantages as follows: a comparison group can be set easily; a power is high as a database has many cases; and no need to consider some points such as reporting bias and recall bias. On the other hand, some issues of a technical point (e.g. understanding characteristics of each database) and a regulatory system (e.g. clarifying association to the drug reexamination system) should be considered. A drug use result survey has advantages as follows: data can be collected, even the data that databases do not have; and a drug use result survey in a specific condition can be useful. As a usual design of drug use result surveys however is hard to answer a research question on each risk, more various designs should be considered. Each pharmacovigilance activity has different points to be considered. Therefore, we have to clearly understand the purpose to perform an additional activity and select the most appropriate activity to maximize the benefit of medication users.

Good Post-Marketing Study Practice (GPSP) changed in April 2018, allows pharmaceutical companies to use a real-world data for pharmacovigilance activity. On the other hand, it is known that there are 3 major dimensions of pharmacovigilance： “Monitoring, vigilance, and science：building the best evidence” ,“Regulation, industry, and legal system：ensuring public health” and “Medicine, medicines, and uncertainty : doing good to patients” . Therefore, in this article, we consider how the change of GPSP reflects on the 3 features of pharmacovigilance.In general, it is thought that the change of GPSP contributes pharmacovigilance, considering the 3 features of pharmacovigilance. On the other hand, there are some points to improve pharmacovigilance system：1) how a safety question for pharmacovigilance should be addressed, 2) how information of routine pharmacovigilance should contribute to a safety question to be addressed, 3) how a feasibility assessment (assessment of data source before conducting a formal comparative activity) should be conducted, and 4) a necessity of a variety of methodology and data sources such as descriptive studies and disease registry. These improvements will contribute to global standardization and give us global competence.Overall, it is very difficult to consider the best safety question, data source and methodology from many options. However, it is thought that keep considering them in order to accumulate experiences is important for our ultimate goal, which is to help our patients. We expect more discussions among all the stakeholders together.