Utility of Big Data to Explore Medication Adherence in Māori and Non-Māori Community-Dwelling Older Adults with Heart Failure in Aotearoa New Zealand: A Cross-sectional Study

BACKGROUND: Medication adherence improves morbidity and mortality-related outcomes in heart failure, and knowledge of patterns of medication adherence supports patient and clinician decision-making. Routinely collected national data facilitate the exploration of medication adherence and associated factors in older adults with heart failure, including the association between ethnicity and adherence. There are known inequities in access to medicines between Māori (Indigenous People of Aotearoa New Zealand) and non-Māori, yet ethnic variation in medicines adherence in community-dwelling older adults with heart failure has not been explored. OBJECTIVE: Here we identify medication adherence rates for community-dwelling older adults diagnosed with heart failure and differences in adherence rates between Māori and non-Māori. METHODS: Cross-sectional analysis of interRAI (comprehensive standardised assessment) data in a continuously recruited national cohort from 2012 to 2019. RESULTS: Overall, 13,743 assessments (Māori N = 1526) for older community-dwelling adults with heart failure diagnoses were included. The mean age of participants was 74.5 years [standard deviation (SD) 9.1 years] for Māori and 82.3 years (SD 7.8 years) non-Māori. In the Māori cohort, 21.8% did not adhere fully to their medication regimen, whereas in the non-Māori cohort, this figure was 12.8%. After adjusting for confounders, the Māori cohort were more likely to be medication non-adherent than non-Māori [prevalence ratio 1.53, 95% confidence interval (CI) 1.36-1.73]. CONCLUSIONS: There was a significant disparity between Māori and non-Māori concerning medication adherence. Given the international use of the interRAI-HC assessment tool, these results have significant transferability to other countries and allow the identification of underserved ethnic groups for which culturally appropriate interventions can be targeted.

https://link.springer.com/article/10.1007/s40266-023-01044-2