News

Predicting treatment adherence: a review of computational approaches

Treatment non-adherence affects up to 50% of patients with chronic conditions, leading to poorer health outcomes, higher hospitalisation rates, and increased mortality. This study titled “Methods and computational techniques for predicting adherence to treatment: A scoping review” and involving BEAMER partners provides a structured overview of computational methods used to build predictive models for patient adherence, including the BEAMER Model B-COMPASS.

A scoping review of 29 studies revealed that supervised learning techniques—such as generalized linear models (21.7%), logistic regressions (20%), and random forests (18.3%)—are the most commonly applied. Over half of the studies focused on chronic metabolic conditions like diabetes and hypertension, with predictors often including treatment-related, socio-demographic, and economic factors.

While these models show promise, few account for the complex interplay of factors influencing adherence. The findings highlight the need for advanced analytical techniques to better capture these dynamics, paving the way for personalized care, improved outcomes, and reduced healthcare costs.

Access the full publication here.

We Would Be Happy to Hear Your Comments

Feel free to share your thoughts by contacting us

Healthcare Professional

The primary aim of the end user personas is to support the creation of materials to support the implementation of the BEAMER model framework and to help define requirements for the elements of the BEAMER model framework. Hence, healthcare professionals (HCPs) represent the primary envisaged end user group of the BEAMER model framework and the associated Adherence Intelligence Visualisation Platform (AIVP)

It is one learning from the joint design process that the job titles of healthcare professional team members do not necessarily predict the roles they would play within the change management process for implementing BEAMER and installing it as a standard model within healthcare. Additionally, the role and responsibilities of certain job titles, for example nurse, varies across different healthcare systems and would affect how they interact with the BEAMER model outputs and the access they would be permitted and so it would not be helpful to include these job titles: The four personas represent role-independent archetypes within the group of HCPs. They encompass a Managerial HCP Persona, an Implementer HCP Persona, a Support HCP Persona, and a Established HCP Persona.

These healthcare professional personas may be further tailored to specific healthcare settings depending on the needs of the individual pilot sites. Thus, adapted or spin-off versions of these original personas may be considered. The persona displays include a summarising statement, goals, challenges, experience, and needs to enhance the accessibility and usability of the model while minimising user burden.

Patient Organisation

Patient organisations are considered potential users of the model outputs. Consequently, personas were designed for these groups to assure that the implementation materials may also support their needs in the longer term, thus fostering sustainability of the project outputs.

The identified focus areas within this persona are goals, needs, skills and tools, along with potential challenges anticipated during the implementation process. The persona emphasises awareness-raising, capacity building, education, peer support provision, and the promotion of research and development in therapeutic care.

The patient organisation persona serves as a theoretical framework representing how patient organisations could benefit from and include the BEAMER model framework in their therapy and care related as well as their organisational work. This persona comprises the needs, goals, challenges and necessary tools, facilitating preparation and implementation of the model and optimising the user experience of patient organisations as end users of the BEAMER model framework. It can be used as a guide to identify potential obstacles and understand the prerequisites for a patient organisation to successfully adopt and integrate the BEAMER model framework.

“In implementing the BEAMER model, we want to be able to respond to the different needs of our patients to ensure their adherence, build a supportive community and improve outcomes.”