Home » Predicting Medication Adherence using Big Data

Predicting Medication Adherence using Big Data

A Short-Term Scientific Mission experience Italy - Netherlands

Sara Mucherino, Postdoc Researcher at CIRFF - Center of Pharmacoeconomics and Outcome Research, Department of Pharmacy, University of Naples Federico II, Italy

In an effort to improve medication adherence, technology takes on a key role, facilitating the precise monitoring and measurement of adherence behaviors. Among these technological advances, health care-related databases emerge as invaluable resources, leveraging the vast power of Big Data.

Health-related databases are a robust repositories of real-world patient data, serving as crucial tools for assessing medication adherence across various populations. These databases provide a deep understanding of adherence behaviors as they are enriched by different data sources. They encompasses various data sources – such as medical prescription records and pharmacy dispensing records – offering a comprehensive insight into patients’ medication usage patterns, treatment trajectories, and medication taking behaviors within real-world settings.  For instance, pharmacy refill data offer a valid and efficient method for retrospectively assessing medication adherence in large population-based research.

Throughout Short-Term Scientific Missions (STSMs) and collaborative initiatives facilitated by the COST ENABLE Action, researchers have promoted cross-country comparisons and longitudinal assessments of medication adherence. Leveraging advanced analytical techniques, including longitudinal trajectory methods, researchers discern adherence trends, identify influencing factors with the ultimate aim to predict adherence outcomes over time.

Image generated with AI

In today’s digitally-driven healthcare landscape, and upcoming adherence collaborations, the application of Artificial Intelligence (AI) and Machine Learning (ML) algorithms has revolutionized medication adherence research. These technologies, coupled with extensive health electronic databases, empower healthcare professionals to understand and enhance medication adherence practices.

Furthermore, AI algorithms applied to health electronic databases may facilitate the development of predictive models, enabling the anticipation of patients’ medication adherence behaviors. Finally, these models help identify patients at risk of low adherence to treatment and also enable ad-hoc targeted interventions for specific patient populations to improve medication adherence and health outcomes.

As research continues to explore the potential of health technology, AI, and ML in the field of medication adherence research, it becomes imperative to foster collaboration among institutions and researchers from various countries. Through the exchange of expertise, methodologies, and data resources, is possible to collectively advance our understanding of medication adherence and develop innovative strategies to mitigate the challenges associated with low adherence.