Faster knowledge of side effects via artificial intelligence

The research project PHAIR – Pharmacovigilance by Al Real-time Analyzes – will make it possible to pick up unwanted effects of medicines faster and thereby increase the quality of treatment with medicines. This must be done by combining national health data and using artificial intelligence in the form of algorithms and pattern recognition. Data in the project include, among other things, national health registers, patient records, and the patients' own experiences of side effects.

Greater knowledge of side effects

The project's first case is the Covid-19 vaccines. The project will thus deepen specific knowledge about possible side effects of vaccines and contribute to new complex methods that can also be used to learn about the side effects of other medicines.

The project's innovative approach, combined with the participation of many parties and professionals within the health sector, gives the research project a broad and solid foundation that will ultimately increase patient safety for the individual citizen.

- At KU we are really happy about the perspectives in the project. Last year we established a pioneer center for artificial intelligence at the University of Copenhagen because artificial intelligence will change our world. Artificial intelligence lives on large amounts of data, and in PHAIR we will be able to develop models that can be beneficial in many areas. It will give us the opportunity to push the international use of artificial intelligence in relation to health data, says Mads Nielsen, professor at the University of Copenhagen, Department of Computer Science and co-lead in the Pioneer Centre for AI.

Bispebjerg and Frederiksberg Hospital are also participating in the project, and here chief physician Espen Jimenez Solem believes that the project has great potential for healthcare, patients, and the research environment.

- Our phase 4 research – ie the research that takes place once the medicines have received marketing authorization and are on the market – gets a huge boost. There are no other countries that have access to data like that for an entire population. It will attract both companies and research from the rest of the world. We are going to give Denmark as a research nation a proper kick in the rear. For the patients, we get a much better understanding of the effect, price, and side effects when we can follow the treatment in real-time, he says.

For the benefit of patients

Jesper Kjær, head of the unit for the Danish Medicines Agency's data analysis center (DAC), sees opportunities to use the fine-grained Danish data to develop new solutions for the benefit of future patients:

- We are happy to join the project because it points to the future. Denmark already has a unique position in relation to health data. The connection of patient-reported data, national register data, and artificial intelligence give us tools that increase patient safety in the entire value chain, from drug approval to side-effect monitoring.

Danish patients also see great potential for patient safety in using data to pick up patterns and potential side effects faster than today.

- It is important that the patient perspective is considered from the start so that the project meets the patient's needs, wishes, and concerns. We know that they are happy to make their data available when it can create improvements for the benefit of others, and we must protect that trust by involving patients and communicating the project at eye level with the patients. We are happy to contribute to that, says Morten Freil, director of Danske Patients.

At Trifork, Karen Skjerbæk Jørgensen, VP & CCO Trifork Digital Health, has great potential in entering into a public-private partnership that combines artificial intelligence and health data.

- PHAIR has the potential to create a common information model so that inappropriateness in relation to side effects can be found both nationally and internationally, she says.