The Association Between Medication Adherence Across Multiple Medications and Health Outcomes in Ageing Populations

Aim: The overarching aim of this thesis is to explore the association between medication (non-) adherence and a range of health outcomes, particularly healthcare utilisation and health-related quality of life, in older adults. A specific focus of the thesis was the objective measurement of adherence across multiple medications in older adults with multimorbidity.

Methods and results: Four inter-related studies were conducted to address the thesis aims. The first study involved a systematic review of the literature of the association between medication adherence and health outcomes in ageing populations. Sixty-six studies were identified but due to significant heterogeneity, results were only pooled across 3 outcomes; hospitalisations, emergency department visits and mortality. Overall, there was a significant inverse association between adherence and hospitalisation and mortality. The second study explored the association between antihypertensive medication (AHTM) adherence and subsequent healthcare utilisation in a nationally-representative sample of Irish adults aged ≥50 years (TILDA). Good AHTM adherence (PDC≥80%) was associated with a significant decrease in subsequent GP visit rate, although no significant association was observed with secondary healthcare utilisation rates (hospitalisations, ED visits and outpatient department visits). The third study applied the RxRisk-V morbidity tool in the objective measurement of adherence across multiple medications and associations with patient-centred outcomes including EQ-5D utility and vulnerability were estimated. Data from an existing prospective cohort of community-dwelling Irish adults aged ≥70 years (CPCR) were used. Summary adherence estimates over a 24-month period between wave 1 and wave 2, were estimated using the CMA7 function within the AdhereR programme. In addition, adherence trajectories were identified using group-based trajectory modelling (GBTM); a useful statistical approach for grouping participants based on longitudinal adherence patterns. The association between different methods of measuring adherence among multimorbid participants and patient-centred outcomes, captured at wave 2, was examined. Higher levels of adherence had a small, significant positive association with EQ-5D utility. Five adherence groups were identified using GBTM with no significant differences in utility scores found between the adherence groups. Adherence had a significant inverse association with vulnerability risk. In the fourth study, the same method of adherence measurement was applied over a 12-month period to a nationally-representative sample of community-dwelling (TILDA), multimorbid adults aged ≥70 years and the association with subsequent self-reported healthcare-utilisation was examined. There was a significant inverse relationship between adherence and both the GP visit and outpatient visit rate. No significant association was found between adherence and self-reported hospitalisation rate.

Conclusion: Findings from this thesis suggest that the association between medication adherence and health outcomes in ageing populations is inconsistent, with variability associated with adherence measurement, outcome typology and population under study. Overall, suboptimal medication adherence is associated with increased healthcare utilisation in older people, but the effect size may be small, depending on the service use measured. Suboptimal medication adherence may be associated with decreased health-related quality of life in older multimorbid people, but further research is needed to confirm the nature and significance of this relationship. Suboptimal medication adherence should be monitored and addressed in older, multimorbid populations, potentially as part of a comprehensive medication management intervention.