Four types of factor have generally been found to predict problems with adherence to medication: regimen characteristics, various patient factors, the relationship between provider and patient and the system of care. The following section focuses on the first three factors; a discussion of factors associated with the system of care is beyond the scope of this report.
A. Regimen-related factors
Complexity of regimen. For many chronic diseases, research has shown that adherence decreases as the complexity of the medication regimen increases (i.e. the number of pills per dose and number of doses per day; the necessity to observe strict requirements related to the intake of food, and the existence of special requirements regarding fluid intake). Adherence to HIV medications is an extremely complicated process that includes both the drugs themselves and the adjustments to daily life necessary to provide the prerequisite conditions for effective drug therapy (13). Some regimens require several doses of medication per day together with various requirements or restrictions on food intake and other activities. These complexities, in addition to the problems of toxicity and side-effects, can greatly influence an individual's willingness and ability to adhere to the therapy (28 - 31).
Many health professionals believe that pill burden strongly influences adherence. However, the effect of pill burden on adherence is closely associated with disease stage. Symptomatic individuals perceive a higher risk for complications of nonadherence to medication, than do asymptomatic patients (32). Dosing schedules and food restrictions or requirements appear to have a more pervasive influence on adherence than pill burden. In the treatment of many diseases, once-daily or twice-daily doses are preferred (33,34). For instance, Eldred and colleagues (33) found that patients on twice-daily doses or less reported better adherence (>80%) and were more likely to take their medications when away from home. Paterson and colleagues (6) also found that a twice-daily dose was associated with better adherence than a three-times-daily dose. However, other studies have failed to confirm this association, including the large Health Care Services Utilization Study with more than 1900 participants (35).Wenger and colleagues demonstrated that the "fit" of the regimen to an individual's lifestyle and schedule, and the individual's attitude towards treatment were better predictors of adherence than dosing schedule (35). It is likely, however, that fewer doses do allow for easier "fitting" of medications into an individual's schedule.
Regimens that involve close monitoring and severe lifestyle alterations together with side-effects may lead not only to frustration and treatment fatigue, but also ultimately to noncompliance (36). Regimens requiring fewer alterations in lifestyle patterns (e.g. fewer pills per day and fewer dietary restrictions) are likely to have a positive influence on adherence to medication.
To the extent possible, regimens should be simplified by reducing the number of pills and frequency of therapy, and by minimizing drug interactions and side-effects. This is particularly important for patients with strong biases against many pills and frequent dosing. There is evidence that simplified regimens that require fewer pills and lower dose frequencies improve adherence (37).When choosing appropriate regimens, the patient's eating habits should be reviewed and the specific food requirements of the regimen discussed so that the patient understands what is required before his or her agreement to such restrictions is sought. Regimens requiring an empty stomach several times per day may be difficult for patients suffering from wasting, just as regimens requiring a high fat intake may be difficult for patients with lactose intolerance or fat aversion.
Side-effects. Side-effects have also been consistently associated with decreased adherence and patients who experience more than two aversive reactions are less likely to continue their treatment (38). HAART regimens usually have temporary side-effects including transient reactions (diarrhoea and nausea) as well as longer-lasting effects (i.e. lipodystrophy and neuropathy).The extent to which side-effects alter a patient's motivation to adhere to a treatment regimen depends greatly on the specific contextual issues surrounding the individual. The literature on side-effects clearly shows that optimal adherence occurs with medications that remove symptoms, whereas adherence is reduced by medications that produce side-effects (13,27). Although HAART may greatly increase quality of life in symptomatic individuals, it probably has a negative effect on quality of life in asymptomatic individuals (39).
Patients quickly discontinue therapy or request changes of medication if they experience side-effects (40).Whether real or perceived, side-effects account for more regimen changes than does treatment failure (30,40,41). One large study of more than 860 HIV-positive patients in Italy reported that more than 25% of treatment-naive patients discontinued their treatment within the first year because of toxicity and other side-effects (30). Another study in France found that the patients' subjective experience of side-effects within the first 4 months of treatment predicted nonadherence more than any other predictors, including sociodemographic variables, number of medications or doses per day (42). The symptoms that cause the most distress are fatigue, diarrhoea, nausea and stomach pain, most of which can be successfully treated (30,42).
One serious side-effect that may affect adherence to HIV medications is lipodystrophy. Kasper and colleagues (43) found that 37% of their respondents either stopped or changed their medications because they developed lipodystrophy. Of those who were adherent, 57% stated that they had seriously considered discontinuation of therapy, while 46% stated that they would change medications if symptoms worsened.
Lipodystrophy affects between 30% and 60% of persons on HAART (44,45). Physical manifestations vary greatly but can include fat accumulation on the upper back and neck (buffalo hump), under the muscles of the abdomen (crix belly or protease paunch), lipomas and breast enlargement; it may also cause peripheral wasting of fat in the face, legs, arms and buttocks (46 - 48). Physiologically, these physical deformities are usually preceded by hyperglycaemia, insulin resistance, hypercholesterolaemia and hypertriglyceridaemia. The exact relationship of these physiological changes to lipodystrophy is unclear. Nonetheless, lipid abnormalities must be treated and this can increase the complexity and side-effects of already complex regimens. Selecting regimens that do not contribute to dyslipidaemia or lipodystrophy may allay fears of disfigurement and support adherence.
In the light of these findings, simplified regimens with fewer pills and fewer doses, and that minimize side-effects, are desirable for achieving maximum adherence (38).
B. Patient-related factors
A patient's behaviour is the critical link between a prescribed regimen and treatment outcome. The most effective regimen will fail if the patient does not take the medication as prescribed or refuses to take it. Consequently, all things being equal, the most important factors influencing adherence are patient-related (27).
Psychosocial issues. Perhaps more than anything else, life stress can interfere with proper dosing of protease medication regimens (49,50), and such stress is experienced more often and to a greater degree by individuals of low socioeconomic status. Although studies of most demographic characteristics of patients have generally failed to establish consistent links with adherence to medication, some recent studies have described several variables that have a possible association. Adherence is apparently most difficult for patients with lower levels of education and literacy, and a few studies have reported lower adherence among blacks and women, although this finding has not been consistent (38).Women have cited the stress of childcare as being related to missed doses (36). The abuse of alcohol and intravenous drugs and the presence of depressive symptoms have also been linked with poor adherence to medication.
Although some studies have demonstrated that a history of substance abuse is unrelated to adherence (51,52), active substance abuse is one of the stronger predictors of nonadherence (53,54). Nevertheless, even active substance abusers can achieve good adherence if the provider takes the time to address the patient's concerns about the medications, including anticipation of, and management of, side-effects. Mocroft and colleagues (52) demonstrated that intravenous drug abusers were significantly less likely to begin antiretroviral therapy, but among those who did, the response to therapy was similar to that of other exposed groups.
Psychological distress has also been shown to affect adherence. Depression, stress, and the manner in which individuals manage stress, are among the most significant predictors of adherence, but correlations with other psychiatric comorbidities are weaker (6,53 - 57). Hopelessness and negative feelings can reduce motivation to care for oneself and may also influence a patient's ability to follow complex instructions. Adolescents with HIV who reported high levels of depression demonstrated lower adherence than did their peers who were not depressed (56). These findings are similar to those of studies on other chronic conditions that have demonstrated a relationship between adherence and depression (58).
Just as social support acts as a buffer for many psychosocial problems, it also affects adherence behaviour. Patients with supportive friends and families tend to adhere to HAART better than those without these supports (6,59,60). In addition to the support that can be provided by clinic staff in the form of a good relationship between providers and patients, recommendations for improving adherence often include providing a telephone-counselling line where messages can be left for nurses, and enlisting the support of pharmacists (61). It is important to encourage patients to involve family and friends in their care, and to follow up on referrals to support groups, peer-counselling and community-based organizations.
Several psychosocial predictors of acceptable levels of adherence to HIV medications have been identified in a large-scale, multisite investigation of HAART (62). These include:
- availability of emotional and practical life support;
- the ability of patients to fit the medications into their daily routines;
- the understanding that poor adherence leads to resistance;
- the recognition that taking every dose of the medications is important; and
- feeling comfortable taking medications in front of other people.
Such psychosocial aspects of treatment may be easily overlooked yet have been documented as being crucial to consistent adherence to HIV medication regimens.
Patient-belief system. A patient's knowledge and beliefs about disease and medicine can influence adherence. Understanding the relationship between adherence and viral load and between viral load and disease progression is integral to good adherence behaviour (53).Wenger and colleagues (35) found better adherence in patients who believed antiretroviral medication to be effective. Negative beliefs regarding the efficacy of HAART may also affect adherence behaviour. For example, many African Americans were found to be reluctant to take zidovudine because they believed that it was toxic. Siegel and colleagues (63) showed that African American men were more likely than Caucasian men to report scepticism about medications and their ability to adhere to those medications. Other beliefs such as those regarding interference with the actions of HAART by alcohol and drugs can also affect adherence (64).
The list below, adapted from the NIH Antiretroviral Guidelines (62), lists additional patient- and medication-related strategies to improve adherence.
• Inform patient, anticipate, and treat side-effects.
• Simplify food requirements.
• Avoid adverse drug interactions.
• If possible, reduce dose frequency and number of pills.
• Negotiate a treatment plan, which the patient understands and to which he or she is committed.
• Take time, and use several encounters, to educate the patient and explain the goals of therapy and the need for adherence.
• Establish the patient's readiness to take medication before the first prescription is written.
• Recruit family and friends to support the treatment plan.
• Develop a concrete plan for a specific regimen including dealing with side-effects and relate it to meals and the patient's daily schedule.
• Provide a written schedule with pictures of medications, daily or weekly pill boxes, alarm clocks, pagers or other mechanical aids to adherence.
• Set up adherence support groups, or add adherence issues to the regular agenda of support groups.
• Develop links with local community-based organizations to help explain the need for adherence using educational sessions and practical strategies.
• Consider "pill trials" with jelly beans.
Confusion and forgetfulness are major obstacles in achieving adherence to HIV medication regimens. Difficulty in understanding instructions has also been reported to affect adherence. Requirements and/or restrictions on the intake of food and water, or the temporal sequences of dosing can be confusing. Misunderstandings may arise as a result of a complex regimen, and/or from poor instructions from the health care provider. In the AIDS Clinical Trial Group, 25% of the participants failed to understand how their medications were to be taken (53). In another study, less adherent individuals reported significantly greater confusion than did adherent individuals over how many pills to take and how to take them (41).
The most commonly cited reason for nonadherence is forgetfulness (51,53,65); for example, Chesney and colleagues (53) reported that 66% of their respondents gave this as the main reason for nonadherence. Ostrop and colleagues (51) demonstrated that not only is forgetfulness the most common reason for nonadherence, but also that the middle dose in a three-times-a-day regimen is the most commonly forgotten. Although other studies have not confirmed this finding, doses are more commonly missed in three-times-daily regimens than in once-daily or twice-daily regimens.
Patient - provider relationship. A meaningful and supportive relationship between the patient and health care provider can help to overcome significant barriers to adherence (37,59,66), but few providers routinely ask about adherence or offer counselling (67). Factors that strengthen the relationship between patient and provider include perceptions of provider competence, quality and clarity of communication, compassion, involving the patient as an active participant in treatment decisions and convenience of the regimen (27). Conversely, patients become frustrated with health care providers when misunderstandings occur, treatment becomes complex, the patient is blamed for being a "bad patient" or side-effects go unmanaged. These frustrations may lead to poor adherence. Specific strategies for clinicians and health teams, as suggested in the NIH Antiretroviral Guidelines (62) are listed below:
• Establish trust.
• Serve as educator, source of information, continuous support and monitoring.
• Provide access between visits for questions or problems by giving the patient a pager number, and arranging for coverage during vacation periods and conferences.
• Monitor adherence; intensify management during periods of low adherence (e.g. by means of more frequent visits, recruitment of family and friends, deployment of other team members, referral for mental health or chemical-dependency services).
• Utilize health team for all patients, for difficult patients and for those with special needs (e.g. peer educators for adolescents or for intravenous drug users).
• Consider the impact of new diagnoses (e.g. depression, liver disease, wasting, recurrent chemical dependency), on adherence and include adherence intervention in their management.
• Enlist nurses, pharmacists, peer educators, volunteers, case managers, drug counsellors, physician's assistants, nurse practitioners and research nurses to reinforce the message of adherence.
• Provide training on antiretroviral therapy and adherence to the support team.
• Add adherence interventions to the job descriptions of HIV support-team members; add continuity-of-care role to improve patient access.