Theoretical Models for Medication Compliance

This is a review of the adherence literature in no way reflects any of the proprietary research/data results from work with our clients.

1. Social Cognition Models
1. Historically, began with the Health Belief Model, which was augmented by the Theory of Reasoned Action, which was in turn augmented by the Theory of Planned Behavior Based on the assumption that attitudes and beliefs affect adherence behavior.
1. Health Belief Model: Expectancies - Perceived benefit of adherence, cost of adherence, cost of not adhering, (susceptibility, severity)
2. HBM works best as a predictor of prevention behaviors (R2 = .25)
3. Criticisms: too simplistic-"barriers" and "benefits" too broad, beliefs underlying these constructs not specified; neglects social factors, leaves out stage of intention.
2. Theory of Reasoned Action
1. Intentions proceed and predict behavior. Intentions are determined by attitudes (toward adherence and subjective norms around adherence). Thus intentions are the bridge between attitudes and behavior
2. Considers social factors, especially normative beliefs:
1. Beliefs and attitudes of friends
2. Beliefs and attitudes of family
3. Beliefs and attitudes of doctors
3. Attitudes about adherence are due to beliefs (expectancies and values). Subjective norms are based on the beliefs of valued others.
3. Theory of Planned Behavior
1. Adds perceived behavioral control-self-efficacy-and perceived barriers to the Theory of Reasoned Action.
2. Includes attributions for beliefs about illness and causes of one's adherence-internal vs. external cause, stability, global vs. specific, universal vs. personal cause, controllability
3. This augments the model to include factors not necessarily under the individual's control
4. Anticipated directions in the evolution of Social Cognition Models--the Social Cognition model's evolution will probably move next to incorporate less rational/conscious processes, specifically:
1. Denial, as either a coping strategy or defense mechanism
2. Constructivist theory Personal Construction of meaning-for example, Weinman's work in developing the Illness Belief Scale
2. Stage Models
1. Stage models share a core base of following stages of adherence from pre contemplation stage through a motivation stage to the initiation and maintenance of adherence behavior. The patient's desire to comply is not assumed.
2. They all assume that different cognitions are important at different stages.
3. Major criticism of all stage models include:
1. They pay scant attention to the maintenance stage. For example, how is the motivation to continue maintenance maintained?
2. The main barrier to change recognized by stage models is low self-efficacy. This is an important variable, but not likely a comprehensive explanation for capturing all barriers.
3. The models are simplistic and linear, no reciprocal causality or feedback processes are conceptualized.
4. Apparently irrational responses, such as denial, are not accounted for.
5. Individual differences in personal meaning of illnesses are not addressed.
6. The Transtheoretical Model is one of several early stage models. Its limitations (that it is linear, simplistic, and narrow) are much more comprehensively addressed by the self-regulation model.
4. Major early stage models include:
1. Transtheoretical Model (stages of change)--evolves from clinical work and research on treatment of addictive/compulsive behaviors. Adds decision theory (pro's vs. con's--Janis) to stages.
2. Health Action Process Approach
3. Precaution Adoption Process
4. Goal Setting Theory
5. The Self-Regulation Model built on earlier models to address deficits.
   Key differences include emphasis on the coping appraisal processes, and the resultant feedback effects on cognition, emotion and behavior.
   The interaction between cognition and behavior is a dynamic process, not the result of a single stage decision.
   The selection of a coping response (behavioral--take medication, coping--denial or distraction) is caused by beliefs about the illness threat, an appraisal of the attempted solution,
1. Fear, an action plan, ideas about one's illness and personal experiences of symptoms are key dimensions.
2. Personal representations of illness drives coping and appraisal. People form common sense models about their illness along five dimensions:
1. Identity - Physical signs and symptoms and an abstract label for them
2. Cause - What makes you sick, how you got the illness
3. Consequences - Expected physical, psychological, and social outcomes
4. Time-line - Expected course and duration of the illness
5. Cure - Beliefs about potential for cure or control. Beliefs that an illness is acute predict treatment drop-out (Myer's hypertension study), beliefs that an illness chronic predicts adherence
3. The patient is an active problem solver. Adherence behavior is intended to close the gap between current health and future goals for one's health. Threats to health are the problem, patient's behavior is an attempt to solve the problem.
4. The self regulation model defines responses to illness as involving the following three steps, which operate in a continuous feedback loop, not in a linear progression:
1. Cognitive representation of, and emotional reaction to, the health threat which is identified by internal cues-symptoms, and external cues-information.
2. The development and implementation of an action plan or coping process to deal with the threat
3. Appraisal of, and feelings about, the outcome of the plan, or coping response. Appraisals could lead to a new plan (increased adherence), a new definition of the illness (its more serious than I thought), or a different coping strategy (denial).
5. General and Specific beliefs are distinguished. General beliefs are about how the illness affects the average person, specific beliefs are exclusively about the individual. One must assess both. General beliefs are likely to be objective, rational and scientific. Specific beliefs are more likely to be subjective and idiosyncratic.
1. Support for this distinction seen in Myer's hypertension study, subjects believed at the general level that hypertension is a symptomatic. At the specific level, many believed (erroneously) that they could tell when their pressure was up, and took their medication accordingly-less when they thought it was down, more when they thought it was up.
6. The limited empirical testing of self-regulation theory has supported the model (hypertension, diabetes, and cardiac rehabilitation studies). Because of its complexity it is more difficult to operationalize. Weinman's Personal Representations of Illness scale may provide researchers with a good model for operationalizing the self-regulation model.
6. Future Directions for the Self-Regulation Model
1. Future Directions: The Self-Regulation Model should be extended to include representations and beliefs about medication, as well as illness beliefs, as called for by Horne in his review of qualitative studies about beliefs about medication. Horne asserts that medication beliefs may have a more direct effect on adherence than illness beliefs. Factors emerging from this research include those at the general and specific levels:
1. Specific/necessary - the specific value of medication for health
2. Specific fears - the extent to which specific proscribed medication is believed to be harmful (long-term side effects) or addictive, or disruptive to lifestyle. Specific fears may be rational, but not true, and thus would be a marker for poor doctor-patient communication.
3. General fears - the extent to which medication in general is feared for side-effects and addiction,
4. General overuse - the extent to which doctors are believed to over-rely on medication
2. Refinements in the concept of nonadherence -Horne distinguishes between passive nonadherence and active nonadherence.
1. Passive nonadherence includes forgetfulness-forgetfulness in not a memory problem (except in extreme cases) beliefs about illness and medication lead to low saliency of medication, which leads to forgetfulness
2. Active non-adherence is due to Specific Concerns about medication. Taking less medication is an active coping strategy to deal with fears of long term dependence, and long term side-effects--medication is a necessary poison to be titrated-titration balances "objective" need for medication, fears of consequences of illness, and fears of consequences of using medication. The process cycles through appraisal processes that seek the optimal balance between internal emotion fear states. Ultimately, the patient is directly guided by achieving the optimal internal emotional state, analogous to adjusting weights on a multi-dimensional balance scale.
3. Prospective longitudinal studies are crucial for testing the model.
1. Multiple measures of adherence are needed
2. Initial response and long term response to medication must be differentiated. Hypothesis tested should included that general and specific medication beliefs will affect treatment preferences and initial orientation to medication. Long term compliance will rest on the dynamic, cyclical reciprocal processes through which specific experiences of symptoms affect initial beliefs.
3. Intervention studies are needed to test whether beliefs can be changed, and to then quantify the degree to which adherence behavior changes in response to changes in beliefs. Doctor-patient relationship variables and other relationships that influence the patient's medication and illness beliefs are the avenues by which beliefs will change. Costs of these interventions must be tested against the value of improved adherence.
7. Physician-Patient Relationship as a determinant of adherence
1. The traditional model of the doctor patient relationship was too simplistic.
1. Traditional model assumed an expert physician and a compliant, passive patient who did as told, or refused to cooperate. Noncompliance is a patient-locus problem
2. A more current model is based on the patient who comes with a complex agenda, and who are active in his or her own health care. They modify or distort suggested treatment regimens, rather than completely accept or reject them. Low adherence is a relationship problem. Underreporting of adherence is often a patient strategy to preserve the relationship with the doctor, and avoid confrontation (Fletcher, 1989).
2. Communication implications of major models
1. Health Belief Model - Identifying patients' beliefs improves adherence, Training doctors on why patients might not adhere increases time spent on patient education, improves knowledge and beliefs, increases adherence (Inui, 1976, hypertension study)
2. Locus of Control - Optimal strategy for internal/external is one that matches the situation, not an absolute. Assessment by doctor should lead to tailoring strategies, e.g., if internal LOC, focus on what patient can/should do.
3. Stages of Change - Motivational assessment is crucial to identify stage patient is in, tailor approach accordingly. Support for this intervention comes from smoking cessation, weight loss and diet modification studies.
4. Self-Regulation Model - Miscommunication in one of three crucial areas leads to decreased adherence: 1) patients' beliefs about illness and medication, 2) patients' action plan for dealing with illness, 3) patients' appraisals of effectiveness of the plan. Doctors typically address nature and consequences of illness, not how to implement a plan, cope with problems, how best to appraise results/effects of patients' coping efforts (e.g., what kind of improvement is realistic to expect over what period of time).
3. Typical communication problems
1. Doctors over control interviews-interrupt patients (on average, after 18 seconds) even though patient problem descriptions typically last 60 seconds. Don't ask open-ended questions. Thus doctors obtain, on average, only 50% of information available.
2. Doctors avoid psychosocial issues, restricted to biomedical issues. Studies find this lowers adherence (DiMatteo & Hayes, Starfield).
3. Doctors tend not to inquire about adherence, adherence improves when they do (Hall).
4. Doctors feel uncomfortable and poorly trained at patient education, and typically underestimate the amount of information patients want. Doctors rarely check to see if information has been understood as intended. This is problematic-adherence, satisfaction, recall and understanding are all related to amount and type of information communicated (Hall, Burgoon). Svarstad found that 70% of patients in a general practice clinic who understood recommendations were adherent, only 15% of those who made one or more mistake in recalling recommendations adhered. 50% don't recall accurately.

Dimensions - Important

8. Doctor variables - physician job satisfaction,
9. Doctor patient relationship-
1. partnership (empowerment, role (patient participation, choice, and responsibility) vs. traditional doctor patient role,
2. quality of communication,
3. affective quality of the relationship,
4. quality of alliance.
10. Patient beliefs and cognitions
1. faith and optimism - adherence to placebo leads to improved outcomes-post M.I. mortality
2. appraisals of disease
3. appraisals of effectiveness of coping/adherence behaviors
11. Characteristics of treatment - complexity (more than number of doses), number of drugs, fitting dosage schedules into daily routines; duration, number of drugs goes up with age, thus complexity is increasing as memory may be decreasing. Type of medication
1. cardiac and diabetic drugs-80-90%
2. antibiotics-75%
3. anti-hypertensives & diuretics-60-70%
4. tranquilizers, analgesics-40-50%
5. musculoskelatal drugs-40%
12. Beliefs-beliefs are not always "good" or "maladaptive. Adaptive value of a belief is situational.
1. Efficacy-Outcome efficacy and self-efficacy (confidence in being able to perform necessary adherence behaviors. Efficacy becomes more important to the extent an adherence behavior is difficult
2. Internal vs. Health Locus of Control- Health locus of control much more relevant to adherence behavior and general locus of control
3. Dispositional optimism
13. Coping-situational also, not always "good" or "maladaptive.
14. Stress-Negative events
15. Memory-Ability to accurately recall proscribed regimen.
16. Depression-This is an important moderator variable-it undermines other strengths and exacerbates other risks, such as low self-efficacy
17. Quality of Life/ satisfaction with various dimensions of life-family, friends, meaningfulness of life, quality of daily activities, finances, health, mood, etc. set the context in which more proximal variables operate.
18. Importance of distinguishing between short-term experienced side effects and feared long-term side effects
19. Information-adequacy, relevance, ease of comprehension
20. Social norms-beliefs and attitudes about illness and treatment by significant others, including family and close friends.
21. Cost
22. Packaging, especially for elderly-up to one quarter could not open their childproof containers.

Dimensions - No effects

23. Demographics
24. Stable personality traits

Measurement Issues

Patients can be very accurate in reporting the likelihood that they will adhere to treatment if asked simply and directly.

Still, pill counts yields higher estimates of non-adherence than self-report

Prescription checks more valid if the check is on dispensing (pt picks it up), not just called in.

MEMS-electronic monitoring yields higher estimates of non-adherence than self report or pill checks.

80 percent is a reasonable standard of adherence

Interactions of predictors of adherence need to be assessed, not just main effects-how do factors combine, and effect each other differentially, to affect compliance. Are risk factors additive, or multiplicative. For example, do depression, poor

Back