Medication adherence among people living with multimorbidity: prevalence, predictors and intervention options

Background. Multimorbidity refers to the presence of multiple chronic conditions in an individual person. The prevalence of multimorbidity is increasing globally, presenting a significant challenge for healthcare systems and many individuals. Multimorbidity adds complexity to the clinical- and self-management of illness, including the use of medicines. Medication adherence refers to taking medicines as agreed with a healthcare provider. Non adherence has been described extensively in the literature; however, the focus has largely been on single-condition populations. Multimorbidity is the norm in clinical practice, and may influence the beliefs and behaviours relevant to taking medicines. An evidence base outlining medication non-adherence in this population is therefore needed. Aims. This thesis aimed to (1) synthesise the reported prevalence and predictors of medication non-adherence among people living with multimorbidity, (2) qualitatively explore the experiences of people taking medicines for multiple co-occurring conditions, (3) quantitatively examine theoretical correlates of non-adherence in this cohort, and (4) systematically identify options for a behaviour change intervention to support adherence to medications prescribed for multimorbidity. Methods. Four studies were conducted to address the aims of this thesis. Study 1 involved a systematic review and meta-analysis to synthesise the existing quantitative evidence reporting the prevalence and predictors of non-adherence to medications among people living with multimorbidity. Study 2 used thematic analysis to explore the experiences and perceptions of taking medicines for multimorbidity, using data derived from semi-structured interviews with 16 people recruited in general practice. Study 3 involved a quantitative secondary data analysis using polynomial regression to examine the multidimensional relationship between medication beliefs and adherence in a general practice cohort of 812 people living with multimorbidity. Study 4 applied the Behaviour Change Wheel to select options for a behavioural intervention to support adherence to medications, using the evidence produced in the preceding studies and the wider literature. A panel of Public and Patient Involvement contributors were engaged with throughout the programme of research. Findings. The prevalence of non-adherence among people living with multimorbidity was estimated to be approximately 43% in a meta-analysis of eight studies. Treatment- and illness-related beliefs, memory performance, and the quality of relationships with healthcare providers were among the reported correlates of non-adherence. Several reviewed studies suggested that adherence could differ between conditions within individuals living with multimorbidity. Findings from the qualitative study highlighted that multimorbidity can produce an amplified burden for some people, who seek to relieve this burden through resignation to their need for medications, prioritisation among conditions, and use of supports such as memory aids. These strategies may influence adherence to varying extents, for example prioritisation may lead to disparities in adherence across conditions, while resignation may lead to better adherence but enduring concerns about medicines. The polynomial regression analysis supported this assertion, as people who believed in the necessity of medicines while also having concerns were more adherent than those who did not believe medicines were necessary and who had no concerns. These findings together highlight the complex beliefs and behaviours experienced by some people living with multimorbidity. The identification of intervention options using the Behaviour Change Wheel attempts address this complexity by targeting both perceptual and practical barriers to adherence using cognitive- and behaviourally-focussed techniques. Conclusions. This thesis applied multiple methods to understand and describe adherence to prescribed medications in the context of multimorbidity. Through a synthesis of the available evidence on prevalence and predictors, a qualitative exploration of individual experiences and perceptions, a quantitative examination of a multidimensional theory, and application of the resultant evidence to select options for a behaviour change intervention, the research reported here contributes to the literature on a key behaviour in a prevalent population. By focussing on multimorbidity, the findings from this thesis offer important implications for intervention development, health psychology, and healthcare more broadly.2023-06-2