Identification of cardiovascular patient groups at risk for poor medication adherence: a cluster analysis

Background Poor medication adherence limits the secondary prevention of cardiovascular diseases (CVDs) and leads to increased morbidity, mortality, and costs. Identifying groups of patients at risk of poor adherence behavior could enable an intervention to be developed and target patients appropriately. Objective The first aim of this study was to identify homogeneous subgroups of cardiovascular outpatients based on their cardiovascular risk factors. Subsequently, differences in medication adherence between these groups were examined. Methods In this retrospective, observational study, patients with an established CVD were included. Well-known cardiovascular risk factors such as smoking, diet, exercise, blood lipid levels, blood pressure, and body mass index were collected. To identify patient subgroups, a 2-step cluster analytic procedure was performed. Differences between the groups on medication adherence were determined on the outcome of the Modified Morisky Scale. Data collection took place between October 2011 and January 2013. Results Cardiovascular risk factors of 530 patients were included in the cluster analysis. Three groups were identified. Compared with other clusters (clusters 1 and 2), cluster 3 contained significantly fewer patients who could be classified as highly adherent and more patients classified as medium adherent (23% and 57%, respectively; P = .024). This group was characterized by a younger age (53% were <55 years old) and using a relatively low number of different medications (41% used