Applying COM-B to medication adherence

On average only fifty percent of people with long term conditions are adherent to their treatment across diverse disease and patient groups (Holloway & van Dijk, 2011; Sabaté, 2003). Medication non-adherence leads to reduced clinical benefit, avoidable morbidity and mortality and medication wastage (DiMatteo, Giordane, Lepper, & Croghan, 2002). With increases in life expectancies as well as the number of patients managing chronic illnesses, this problem may well become worse in the next few years. Consequently, policy makers have called for successful interventions to address the causes of non-adherence and improve the population’s use of medicines (Holloway & van Dijk, 2011; Horne, Weinman, Barber, Elliott, & Morgan, 2006; Nunes et al. , 2009; Sabaté, 2003). Indeed, it has been estimated that $269 billion worldwide could be saved by improving patient medication adherence (IMS Institute for Healthcare Informatics, 2012). Unfortunately, many adherence interventions to date have not been effective (Haynes, Ackloo, Sahota, McDonald, & Yao, 2008). Medical Research Council guidelines recommend that appropriate theory and evidence should be identified to inform the development of an intervention (Craig et al. , 2008). However, most adherence interventions are developed without a sound theoretical base, which may be one of the reasons they have not been effective (Horne et al. , 2006). Successful interventions have often involved a level of complexity that would be too difficult and expensive to implement in practice (Haynes et al. , 2008). Explanations and models of medication adherence/non-adherence have changed over the years. Early work tended to focus on the role of doctor-patient communication and its effects on patient satisfaction, understanding and forgetting as key determinants of subsequent treatment adherence (Ley, 1988). However, health behaviour research has consistently demonstrated that the provision of information alone is not an effective way to change behaviour, and so research has now moved onto approaches and models which focus on patients’ beliefs, motivation and planning abilities as the core explanatory variables. Many of these are social cognition or self-regulatory models which emphasize the importance of the beliefs which individuals have about their illness and treatment as well as their own ability to follow the treatment and advice which they are given (see Conner & Norman, 2005). Existing models and frameworks are not comprehensive since they neglect automatic processes such as habit (for example, Ajzen, 1985; Bandura, 1977, 1986; Horne, 1997, 2003; Leventhal, Nerenz, & Steele, 1984; Pound et al. , 2005; Rosenstock, 1974), do not describe dynamic behaviours whereby the experience of adherence/non-adherence can alter predisposing factors such as beliefs about medication (for example, Ajzen, 1985; Bandura, 1977, 1986; Horne, 2003; Pound et al. , 2005; Rosenstock, 1974) and neglect factors at a systems level (for example, Horne, 2000, 2003; Leventhal et al. , 1984; Pound et al. , 2005; Rosenstock, 1974). In addition, the often used Christina Jackson