Antiretroviral Therapy: Factors Associated With Adherence Ann Williams, RN, EdD Clinical effectiveness of the new highly active antiretroviral therapies depends in large part on patients ' ability to adhere to demanding medication regimens because the suboptimal drug levels associated with nonadherence are, in turn, associated with the development of antiretroviral resistaiwe. However, definitions of adherence are inconsistent, and the concept is difficult to measure. Adherence to medical and health regimens is the outcome of a dynamic process of human behavior and interaction. Factors influencing this process include characteristics of the regimen, the provider, the patient, and society.
Adherence, compliance, nonadherence, antiretroviral therapy T h e past 2 years have witnessed tremendous progress in both the basic science and clinical management of HIV. It is now possible--with the availability of protease inhibitors in addition to the older reverse transcriptase inhibitors--to reduce the number of HIV particles in the blood to levels undetectable by current laboratory methods. In addition, because the new methods measure the virus directly, rather than waiting for the results of viral replication to be reflected in falling CD4+ cell numbers, it is also now possible to closely monitor the effects of antiretroviral therapy and rapidly modify treatment as necessary. These impressive advances have created, predictably, complicated new challenges. Moving from the relatively controlled world of the laboratory and clinical trials environment into the real world of people who live with a chronic, long-term, life-threatening Ann l~lliams, RN, EdD, is an associate professor at Yale School of Nursing and a family nurse practitioner in the Yale AIDS Care Program, New Haven, CT.
infection is not simple. However, it has been part of the legacy of the HIV epidemic that, as we meet each new challenge, we learn something about ourselves and our society that casts light on larger issues. In the first decade of the epidemic, many of those issues were related to the social and political context in which HIV transmission occurred, as a spotlight was thrown on the dynamics of intimate human relationships against a background of poverty, exclusion, and vulnerability. In the second decade, it appears that we will be challenged to examine the dynamics of relationships between health care providers and patients as we try to understand and then influence, once again, human behavior and social conditions. The new circumstances challenge us to rethink relationships between clinicians and patients, the health care system and patients, and prescribers and patients. While considering these individual behaviors, we must also examine relationships between patients and society and between medical institutions and the communities where they are located because ultimately, societal factors are the most powerful determinants of health.
Patient Adherence and HIV Drug Resistance It is difficult, in the midst of all the "hype" and hope, to face the fact that not all patients are benefiting from the new antiretroviral regimens. There is, in fact, a rising tide of frustration among experienced patients and among clinicians as we are faced with many patients who experience antiretroviral "failure." The complete picture of this phenomenon is not yet drawn; we have neither a consistent definition nor a clear explanation of the reasons for the failure to respond to pharmacologic interventions.
JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. 8, Supplement, 1997, 18-23 Copyright 9 1997 Association of Nurses in AIDS Care
Williams / Antiretroviral Therapy
In most instances, a high viral burden and declining CD4+ cell count in someone takinga powerful combination of agents (highly active antiretroviral therapy or HAART) indicates the presence of a virus that is resistant to the medication (Feinberg, 1997). Pharmacologic factors potentially associated with antiretroviral failure and with the development of resistance include drug-drug interactions, changes in metabolism, and inadequate absorptionmall variables that might alter the effective serum drug levels. Drug-drug interactions are complex, and although patients may be taking their antiretroviral agents as prescribed, they may also be taking an interacting drug that precludes the antiretroviral effects from being exerted. Inadequate absorption of the medications is another possibility. In addition, antiretroviral therapy failure may reflect preexisting resistant strains of the virus (Condra & Emini, 1997). Resistance is the result of the incomplete suppression of viral replication (Vella, 1997). Current concern with patient adherence arises from the assumption that persons following a strict medication regimen will exhibit good virus suppression and that their virus will not have the chance to develop mutations and rapidly reproduce. Resistance will be unlikely if viral suppression is sufficient (Vella, 1997). Suboptimal adherence and suboptimal drug levels are associated with the development of antiretroviral resistance. Figure 1 shows a schema that describes the relationship between patient adherence and viral drug resistance. The area of concern in this arc is the section across the top, where there is only partial suppression of the virus, allowing the development of mutations in a replicating population of the virus. This model emphasizes that the development of drug resistance is primarily a problem for individuals who take some of their drugs some of the time. Those who do not fill their prescriptions or take their pills at all (0% compliance) are not contributing to the pool of resistant virus any more than are those who take all of their pills all of the time (100% compliance) (B laschke, 1997).
Recognizing Nonadherence Nonadherence, or noncompliance, is not a new concept, and yet it is still not well defined (Wright, 1993). To understand why an HIV-infected person might not "comply" with his or her antiretroviral regimen--in
Emerging Resistance PROBABLITY OF RESISTANT
Figure 1. Relationship Between Drug Resistance and Drug Compliance. NOTE: Resistance to antiretroviral agents can be prevented or at least significantly delayed if the HIV viral load is sufficiently suppressed. However, resistance readily develops with suboptimal doses of antiretroviral agents, such as is seen with nonadherence.
order to develop and test interventions to increase adherence--a definition is needed that has sufficient nuance to distinguish major categories of behavior, but is not so detailed that it is cumbersome in the clinical environment. At least two maj or categories of behavior are related to medication adherence. The first is active nonadherence, in which an individual decides, based on his or her evaluation of the risks and benefits of the recommended regimen, not to follow the recommendation. Patients who appear faithfully for all visits and submit generously to all the examinations may not take all of their medications. They may know from the very beginning that they will not take medication, but they choose not to inform their provider. For example, a patient with a high viral load and low CD4+ count may choose not to take prescribed antiretroviral medications because he is asymptomatic and believes the drags are too risky. In this situation, the discrepancy between what the clinician intends and what the patient does is the result of a conscious difference of opinion or judgment. To alter the situation, one or more of the parties must make a different decision. The second category comprises those who intend to comply, who may in fact believe that they do comply but in truth do not adhere to the regimen. There are
20 JANAC Vol. 8, Supplement, 1997
numerous factors potentially contributing to the difficulty that the patient has in managing the r e g i m e n n some, but not all, of which might be altered. It should be noted that a broad view of adherence and compliance includes behaviors associated with keeping professional appointments and making lifestyle changes such as diet, exercise, and drug use as well as taking medications correctly. As experience with the protease inhibitors has reminded us so dramatically, these other aspects of compliance are also crucial to antiretroviral effectiveness. Adherence to medical and health regimens is the outcome of a dynamic process of human behavior. In the simplest model, it is shaped by the dynamics of the relationship between the patient and the provider. Contributing factors include characteristics of the regimen, the provider, and the patient, all of which affect the outcome: adherence or nonadherence (Wright, 1993).
Adherence: Medication Factors The ease or difficulty of adhering to a regimen is affected by many characteristics of the medications, such as how often pills must be taken, how many pills must be taken (each time and each day), what the medication tastes like, how big the tablets are, and how the medications must be stored. Other factors include requirements such as whether the pills must be taken with or without food, what the side effects are, and how effective the drug is at accomplishing its primary purpose. The frequency of the daily dose of a prescribed medication makes a difference in the ease with which an individual can reliably take the medication. More frequent doses are more difficult to adhere to than are less frequent doses (Cramer, Mattson, Prevey, Scheyer, & Quellette, 1989; Eisen, Miller, Woodward, Spitznagel, & Przybeck, 1990), although this may not extend to a daily regimen versus a three-times-a-week regimen. Patients are more likely to miss afternoon or evening doses than morning doses and to skip doses on weekend days compared to weekdays. In addition, the total number of pills people have to take clearly has an impact. Some patients find that the size of the pills or capsules influences their willingness to take specific medications. Side effects are a particular problem when the medication is prescribed for a condition that is asymptomatic.
Table 1. Adherence to Medication: Clinician Factors Clinical style Interpersonal style Stability Availability Assessment skills Communication skills Clinical management skills
The new highly aggressive approach to antiretroviral therapy and the addition of protease inhibitors have significantly complicated antiretroviral regimens along all of the above lines. For example, a substantial number of pills has been added to many patients' daily quota. In addition, the efficacy of the protease inhibitors appears to be very sensitive to the timing of doses and to the presence or absence of food. The effects of deviations from recommended schedules and directions on absorption, and thus on drug serum levels and presumably the potential for viral resistance, are profound. Almost all of the current antiretroviral regimens are associated with a significant incidence of side effects, some quite severe, altering the quality of daily life (Decks, Smith, Holodniy, & Kahn, 1997).
Adherence: Clinician Factors Although the clinician is an important part of the therapeutic relationship, the influence of clinician characteristics on compliance has not been as well examined as patient characteristics. For purposes of this discussion, the term clinician refers not only to the people who write the prescriptions but also to the people who support that activity with the patients. Numerous characteristics of the clinician may influence a patient's success in adhering to a medication regimen (see Table 1). Clinicians, like patients, may be more or less optimistic about the effectiveness of new drugs. The roller coaster of psychological responses to new developments in HIV science has been partially the result of rapid discoveries, unrealistic expectations, and sincere hope. It has led to a skepticism on the part of some providers that may restrict the enthusiasm with which new regimens are introduced to patients. In any case, it is helpful when the clinician's style is consistent with that of the patient.
Williams / AntiretroviralTherapy 21 Table 2. Adherenceto Medication:PatientFactors Physical health Mental health, includingaddictivedisease Material resources,includingliving situation Cultural beliefs Self-efficacy Social support Personal skills HIV knowledge
Patients who have been able to establish an ongoing relationship with a single primary care provider may find it easier to adhere to difficult regimens. When the clinician is consistently available to answer questions, reinforce recommendations, and provide positive feedback, especially in the early weeks of a difficult regimen, the patient's chances of success are enhanced (Davis, Canniff, Andradas, Cohen, & Hellinger, 1997). Other issues to consider include the clinician's assessment skills and ability to hear what the patient has to say. Clinicians who listen to patients and recognize their anxiety and concerns are better able to help them determine when they are ready to initiate therapy and develop an adherence plan. Communicating clearly with the patient about the purpose of the therapy as well as what to expect while taking the medications is a key management skill, as is the ability to manage side effects as they appear. Many side effects are limited in duration, and astute symptom management combined with anticipatory guidance may salvage a regimen that the patient is otherwise prepared to abandon.
Adherence: Patient Factors Numerous patient characteristics have been examined in relation to compliance with acute and chronic medication and health regimens (Besch, 1995; Haynes, Sackett, & Taylor, 1980), some of which are listed in Table 2. The table represents the general characteristics that have been associated with adherence to medication. Each of these factors comprises many facets and/or levels. HIV knowledge, for example, includes knowing not only the purpose of antiretroviral therapy but also such highly practical information as the effect of drug holidays on antiretroviral efficacy. The key for clinicians is to address these factors
when collaborating with a patient to develop an individualized medication adherence plan. Medication adherence will be affected by whether an individual has the physical ability to do what needs to be done, the mental capacity to comprehend directions and organize the activity, and the material resources to obtain and maintain supplies. Clearly, a person's physical and mental capacities to handle a regimen are important. It is likely that when health care providers make it a priority, it is possible to help people with physical and mental disabilities adhere to fairly complicated regimens. We have a successful history with this process, such as when we asked patients to manage complicated IV medications for cytomegalovirus, fungal infections, and various other difficult regimens at home. The same is true for material resources. Patients with limited resources, including homeless patients, have managed successfully with the help of other professionals, such as home care nurses and case managers who play a large role in the community and home setting. Other patient factors influencing medication adherence include individual and cultural beliefs about illness and health, the availability of supportive significant others who also understand and agree with the regimen, whether the individuals believe they are capable of following the regimen, and whether they believe it will be effective (Besch, 1995). Individual understanding and beliefs about illness and health are embedded in human culture and strongly influence the decisions that patients make related to engagement in care. In particular, religion, spirituality, folk medicine, and alternative medicine play a role in medication adherence behavior. In addition, the majority of Americans living with HIV represent one or more subcultures that are distinct and distanced from the more mainstream culture of most health care providers. There may be a great deal of information sharing within these fairly closed, discrete patient communities, along with strong beliefs and attitudes about effective AIDS treatment. In addition to community, the social support available from family and significant others may have a positive or negative effect on a patient's ability to maintain a medication regimen. Concerned and caring partners are essential elements in the daily adherence plans of many people living with HIV. In contrast, in
22 JANAC Vol. 8, Supplement, 1997 Table 3. AssessingAdherence Counting Self-report Chart review Pharmacy review Pill counts Drug assays Biologic markers Electronic monitoring
some situations, the colleagues and family of the infected person do not know they have AIDS. The effort to maintain secrecy can distract from careful medication adherence, especially when the medications must be taken around mealtimes or at the workplace. The concept of personal skills includes the desire and ability to organize many activities of dally life. Some people are very compulsive and will be able to organize the complicated drug regimens that HIV therapy may demand. Others may find the requirements of these regimens unacceptable. Personal skills and the ability to adhere are not associated with demographic factors such as age, race, gender, education, or socioeconomic class. Finally, the more patients understand about the purpose of their therapeutic regimen and the reasons for the restrictions required by the medications, the better able they are to continue to take them. In a crosssectional survey of 202 patients receiving antiretroviral therapy for at least 6 months, adherence was greater among those who understood the purpose of antiretroviral therapy (Eldred, Wu, Chaisson, & Moore, 1997).
Assessing Adherence It is often of critical importance to know exactly which medications patients are taking and on what schedule, but obtaining that information can be a challenge. Some approaches to measuring medication adherence are listed in Table 3. Other than asking the patient (self-report), the majority of these are not practical for daily clinical use. Serum drug assays similar to those used to measure theophylline levels are potentially of great benefit but are not presently available for most antiretroviral agents. Similarly, a biologic marker such as quantitative HIV RNA may someday be useful to determine by
extrapolation some degree of compliance. Electronic monitoring is a newer approach. One device, the MEMs T M cap, incorporates a small computer chip that records every time the bottle is opened. It, of course, will not reveal where the pill goes, but it does report that the bottle was opened (Cramer et al., 1989). These are, however, very expensive. For most patients and clinicians, medication adherence will continue to be assessed by asking the patient what pills he or she has been taking. By taking a careful medication history at each visit, it is likely that in most instances, this information will be as accurate as the patient's memory. It is often extremely helpful to ask patients to bring all bottles of pills they are taking to the clinic at each visit. Another approach is a home visit to review medication schedules. This practice brings the clinician into the patient's world, which is a very different one than what is seen in the clinic, and allows a more accurate assessment.
Supporting Adherence Strategies to improve HIV medication adherence and behaviors often use the tuberculosis model as an analogy. Diabetes, however, may provide a better analogy. In that illness, providers and patients talk about tight versus loose control, whereas with HIV, the choice is between complete viral suppression versus viral load reduction. And effective diabetes management requires a set of recommended behaviors--medical behaviors---that are quite intrusive and force a change in one's daily life--similar to HIV regimen situations. When health care professionals discuss adherence, there is a tendency to focus on issues related to characteristics of the medications and on the behavior of the patients. However, experience with HIV prevention over the past decade has demonstrated that it is very difficult to change human behavior. In addition to patient behaviors, there are professional behaviors that might be modified to take advantage of what is known about promoting adherence. The noncompliance of providers is reflected in our failure to implement adherence interventions that have been demonstrated to be effective. A rich body of research on adherence in the nursing literature can be drawn on in the same way that literature in oncology, infectious disease, and addiction was drawn on in the beginning of this epidemic to design
Williams / Antiretroviral Therapy 23
programs of prevention and care. In particular, psychiatry, chronic illness, and pediatrics may be useful sources of models that can be examined and adapted. The relevance of current nursing knowledge related to adherence for patients with HIV should be evaluated, and interventions should be designed and tested for effectiveness.
References Besch, C. L. (1995). Compliance in clinical trials. AIDS, 9, 1-10. Blaschke, T. E (1997, January). Noncompliance and resistance to protease inhibitors. Paper presented at the Fourth Conference on Retroviruses and Opportunistic Infections, Washington, DC. Condra, J. H., & Emini, E. A. (1997, January/February). Preventing HIV-1 drug resistance. Science & Medicine, pp. 2-11. Cramer, J., Mattson, R. H., Prevey, M. L., Scheyer, R. D., & Quellette, V. L. (1989). How often is medication taken as prescribed? Journal of the American Medical Association, 261, 3273-3277. Davis, S. M., Canniff, J. M., Andradas, V., Cohen, C. J., & Hellinger, J. A. (1997, January). Successful ritonavir induction:
Intensive patient management. Paper presented at the Fourth Conference on Retroviruses and Opportunistic Infections, Washington, DC. Deeks, S. G., Smith, M., Holodniy, M., & Kahn, J. O. (1997). HIV-1 protease inhibitors: A review for clinicians. Journal of the American Medical Association, 277, 145-153. Eisen, S. A., Miller, D. K., Woodward, R. S., Spitznagel, E., & Przybeck, T. R. (1990). The effect of prescribed daily dose frequency on patient medication compliance. Archives of Internal Medicine, 150, 1881-1883. Eldred, L., Wu, A., Chaisson, R. E., & Moore, R. D. 0997, January). Adherence to antiretroviral therapy in HIV disease. Paper presented at the Fourth Conference on Retroviruses and Opportunistic Infections, Washington, DC. Feinberg, M. (1997). Hidden dangers of incompletely suppressive antiretroviral therapy. Lancet, 349, 1408-1409. Haynes, R. B., Sackett, D. L., & Taylor, D. W. (1980). How to detect and manage low patient compliance in chronic illness. Geriatrics, 35, 91-93, 96-97. Vella, S. (1997). Clinical implications of resistance to antiretroviral drugs. AIDS Clinical Care, 9, 45-47, 49, 52. Wright, E. C. (1993). Non-compliance--Or how many aunts has Matilda? Lancet, 342,909-913.