Introduction: The Caring Does Matter (CDM) programme aimed to improve cardiovascular disease risk (CVR) management in Pacific people, targeting medication adherence problems. This paper presents cluster analysis of CDM data to model medication adherence and cardiovascular risk factors in high-CVR patients.

Methods: Changes in cardiovascular medication adherence status and in physiological measures of high-CVR Pacific patients, as well as their baseline physiological measures, demographics and other risk factors are included in the analysis. Differences in resulting clusters are described to provide insight into the population.

Results: 1786 Pacific patients were identified with high CVR (≥ 10%, 5 year event risk) at baseline and were still enrolled with the thirteen participating general practices at CDM end. Two of three models attempted produced significant clusters: a two-cluster model indicating patients who failed to improve adherence during the programme had higher prevalence of diabetes; and a three-cluster model where one cluster was characterised by higher but improving blood pressure, and another characterised by higher but improving HbA1c.

Discussion and Conclusion: Cluster analysis reveals statistically distinct patient groups, including differences in characteristics of patients less responsive to the programme intervention. This provides a basis for further efforts to understand the population and better tailor interventions.