Pharmacovigilance is an applied science which aims to ensure the safe use of medicines. Our aim is to develop methods that facilitate the early detection of unknown adverse events, characterize better the safety profile of the medicines, and improve the communication of safety information to the patients and healthcare professionals.
For statistical signal detection disproportionality-based statistics are used on spontaneous reporting datasets (including tens of millions of adverse event reports), and we utilize network theoretical and machine learning models to improve the current standards. We also investigate the impact of risk minimization measures, and survey the knowledge of patients and healthcare professionals that helps to increase the knowledge about safe use of medicines.
- Pharmacovigilance (for healthcare students or professionals):
- Adverse event characterization
- Signal detection
- Risk-benefit assessment
- Risk minimization
- Patient and healthcare professional engagement
- Method development (for healthcare or IT students or professionals)
- Pharmaceutical data management
- Big data in pharmaceutical industry
- Application of network theory
- Application of machine learning
Project leader: Mátyás Pétervári PharmD
Adverse drug reactions yield a heavy burden on the public health sector. These adverse reactions are implying severe and prolonged potential risks, therefore the earliest detection is important in order to prevent wide public health effects .
Our aim is to develop methods that facilitate the early detection of unknown adverse events, characterize better the safety profile of the medicines, and improve the communication of safety information to the patients and healthcare professionals.
For statistical signal detection well-established industrial standard methods are applied on spontaneous reporting datasets including tens of millions of individual case safety reports (e.g. FAERS, EudraVigilance). Our internally developed software is able to calculate different correlations between drug and adverse events that helps to characterize the adverse event profile of a drug or drug class, or detect previously unknown possible adverse reactions (signals).
We have also developed network theoretical solutions to describe the correlations within adverse event reports. In our accepted manuscript  in Drug Safety we presented a new signal detection method by using network analytics, and we also made available Vigilace website  that allow access for researchers to run network analysis on spontaneous reporting datasets. In another of our projects machine learning approach was applied to generate machine-readable features of drugs and adverse events based on reporting patterns.
Patient and healthcare professional engagement plays essential role in pharmacovigilance, as they need to apply the novel information originating from pharmacovigilance procedures. Therefore we also research the effectivenes of risk minimization measures, and the pharmacovigilance and drug safety knowledge and attitude of patient and healthcare professionals
- Pétervári M., Benczik B., Balogh O.M., Petrovich B., Ágg B., Ferdinandy P., Network analysis for signal detection in spontaneous adverse event report database: Application of network weighting normalization to characterize cardiovascular drug safety. Drug Safety [accepted, unpublished]