Kahasse’s background is in statistics and health data science. Kahasse’s research interests include, but are not limited to, the impacts of inequalities on health outcomes and the application of machine learning, considering the algorithm bias in the healthcare

Project Information

Research Driver Programme: Medicines in Acute and Chronic Care

Title: Impact of inequalities on adverse drug reactions in patients with polypharmacy: developing approaches to reduce adverse outcomes and taking into account fairness of data algorithms.


Use of multiple medicines (also known as polypharmacy) can increase the risk of side-effects but there are not many guidelines for doctors on how to minimise these risks. In addition, individuals living in deprived areas are more likely to be prescribed multiple medications due to their higher levels of comorbidities. Also, they are more likely to suffer from side-effects with medicines.

The aim of this project is to use large datasets with patient medical records (that have patient’s name and address removed) and to identify important information that indicates higher risks of side-effects in patients with polypharmacy. We will also explore possible problems with the data (also known as biases) which can happen if patients who have difficulty accessing their doctors (for example, patients living in very deprived areas).

The final part of this project will work on how to give feedback to general practitioners on how they can improve treatments in patients with polypharmacy (for example, what medicines should be reviewed by them with the patients to consider stopping it).

What is your motivation for undertaking this project and how will this funding impact your research?

Overprescribing comprises 10% of prescriptions in UK primary care, according to a recent Government report. One of its main consequences is problematic polypharmacy that puts patients at higher risk of complications. This has huge impact both in health outcomes and financial costs. Hence, the findings of this research is believed to be valuable input for policy and guidelines for minimising ADR and Polypharmacy. It is very true that I couldn’t start this research without the funding from HDR UK.