Computation of adherence to medicines to assist clinicians in the outpatient setting NHS Greater Glasgow and Clyde

Adherence to medicines is an important indicator of the quality of medication management, and impacts health care outcomes and cost effectiveness. Patients with long-term conditions, such as epilepsy, are often poorly adherent, which results in increased use of health care and higher risk of death.

Currently, specialist clinicians within the outpatient setting rely on self-reported adherence, which is unreliable and prone to bias. This project from NHS Greater Glasgow and Clyde aimed to provide access to an estimate of adherence at the point of care using information on medicines dispensed to patients in the community.

The programming language R was used, and natural language processing to identify regular expressions in prescription dose instructions, and an R-package (AdhereR) was used to calculate adherence and to visualise the medication history.  

These advanced analytics were incorporated into the Medicines Adherence Visualisation Information System (MAVIS). This innovative app allows clinicians to flexibly visualise individual medication histories, measure adherence patterns and access verified health psychology educational resources to effectively guide adherence support. Evaluation of clinicians’ experience of using MAVIS to address patient adherence issues in outpatient clinics is currently ongoing. 

The project team gained valuable experience in using R and successfully adapted to the challenges of COVID-19 to deliver sustainable changes in analytical capabilities for health care delivery. They plan to continue to build on their new-found skills by implementing R in future quality improvement initiatives and making it the ‘business as usual’ software for data analysts.

Contact information

For more information about this project, please contact Seán MacBride-Stewart, Lead for Medicines Management Resources, NHS Greater Glasgow and Clyde.