A considerable amount of empirical, descriptive, research has identified correlates and predictors of adherence and nonadherence. These include aspects of the complexity and duration of treatment, characteristics of the illness, iatrogenic effects of treatment, costs of treatment, characteristics of health service provision, interaction between practitioner and patient, and sociodemographic variables. Many of these variables are static, and may not be amenable to intervention. They have been well described in the main text of this report and will not be discussed further here. While such findings help to identify risk factors, they tend to be discrete and atheoretical, and not very helpful in guiding a clinical approach to this problem.
This section describes several important variables that are behavioural in nature and are also dynamic, and therefore amenable to intervention. First we identify key behaviours of health care providers, health system factors and attributes of patients. Then we discuss promising behavioural science theories and models that help to explain behavioural change. These serve as helpful heuristics both for understanding nonadherence and for addressing it.
A. Provider behaviours
Variables related to how health care providers interact and communicate with their patients are key determinants of adherence and patient health outcomes (4,6,17,29,30). The health care providers prescribe the medical regimen, interpret it, monitor clinical outcomes and provide feedback to patients (31).
Correlational studies have revealed positive relationships between adherence of patients to their treatment and provider communication styles characterized by, providing information, "positive talk" and asking patients specific questions about adherence (32). The clarity of diagnostic and treatment advice has been correlated with adherence to short-term but not to long-term regimens and chronic illnesses. Continuity of care (follow-up) is a positive correlate of adherence. Patients who view themselves as partners in the treatment process and who are actively engaged in the care process have better adherence behaviour and health outcomes (33).Warmth and empathy of the clinician emerge time and again as being central factors (34). Their patients of providers who share information, build partnerships, and provide emotional support have better outcomes than the patients of providers who do not interact in this manner (35). Patients who are satisfied with their provider and medical regimen adhere more diligently to treatment recommendations (36). Findings such as these can guide providers to create a treatment relationship that reflects a partnership with their patients and supports the discussion of therapeutic options, the negotiation of the regimen and clear discussion of adherence.
Health care providers often try to supply information to patients and to motivate them, and recognize the importance of behavioural skills in improving health. However, there is evidence that, in practice, they give limited information (37), lack skills in motivational enhancement (38), and lack knowledge and experience frustration in teaching patients behavioural skills (39).More structured, thoughtful and sophisticated interactions between provider and patient are essential if improvements in adherence are to be realized.
B. Health system factors
The health care delivery system has great potential to influence the adherence behaviour of patients. The policies and procedures of the health system itself control access to, and quality of, care. System variables include the availability and accessibility of services, support for education of patients, data collection and information management, provision of feedback to patients and health care providers, community supports available to patients, and the training provided to health service providers. Systems direct providers' schedules, dictate appointment lengths, allocate resources, set fee structures and establish organizational priorities. The functioning of the health system influences patients' behaviour in many ways.
- Systems direct appointment length and duration of treatment, and providers often report that their schedules allow insufficient time to address adherence behaviour adequately (40).
- Health systems determine reimbursements and/or fee structures, and many health systems lack financial coverage for patient counselling and education: this threatens or precludes many adherence-focused interventions.
- Systems allocate resources in ways that may result in heightened stress for, and increased demands upon, providers and that have, in turn, been associated with decreased patient adherence (41).
- Systems determine continuity of care and patients demonstrate better adherence when they receive care from the same provider over time (42).
- Systems direct information sharing - the ability of clinics and pharmacies to share information regarding patients' behaviour towards prescription refills has the potential to improve adherence.
- Systems determine the level of communication with patients - ongoing communication efforts (e.g. telephone contacts) that keep the patient engaged in health care may be the simplest and most cost-effective strategy for improving adherence (43).
C. Patient attributes
Patient characteristics have been the focus of numerous investigations of adherence. However, age, sex, education, occupation, income, marital status, race, religion, ethnic background, and urban versus rural living have not been definitely associated with adherence (26,44). Similarly, the search for the stable personality traits of a typical nonadherent patient has been futile - there is no one pattern of patient characteristics predictive of nonadherence (34,42).With the exception of extreme disturbances of functioning and motivation, personality variables have not emerged as significant predictors. Recent studies of patients with mental health problems have provided evidence that depression and anxiety are predictive of adherence to medical recommendations (45 - 48). Almost everyone has difficulty adhering to medical recommendations, especially when the advice entails self-administered care.
Illness-relevant cognitions, perceptions of disease factors, and beliefs about treatment have stronger relationships to adherence. In particular, factors such as perceived susceptibility to illness, perceived severity of illness, self-efficacy and perceived control over health behaviours appear to be correlates (26,49). For adherence to occur, symptoms must be sufficiently severe to arouse the need for adherence, be perceived as being resolvable and acute, and remedial action must effect a rapid and noticeable reduction in symptoms (50).
Knowledge about an illness is not a correlate of nonadherence, but specific knowledge about elements of a medication regimen is, although apparently only for short-term, acute illnesses (51). Some of the above variables, and several others, form the basis of various theories and models of behaviour change and we now turn our attention to these.