The COM-B Framework as a Diagnostic Schema for Medication Nonadherence

Abstract

Medication nonadherence is a common problem encountered by hospitalists. A systematic approach is needed to identify contributors to medication nonadherence in clinical settings. The COM-B model is a commonly used framework to understand drivers of behavior. This perspective illustrates how clinicians can use the COM-B framework as a diagnostic schema for medication nonadherence.

Background

Medication nonadherence is a pressing issue for patients and physicians. Patients who are less adherent to their chronic medications have increased risk for emergency room visits and hospitalization.(1) The international prevalence of hospital admissions related to medication nonadherence is estimated as 4.3%, with rates ranging from 0.7% to 10.8%.(2) Estimates across the US indicate that 31% of patients with hypertension, one of the most common conditions encountered by internists, are nonadherent to their medications.(3)

Patient nonadherence is a frustrating clinical issue because the solution is apparently so simple: “If this patient would just take their medication, this episode could have been avoided.” Physicians report that addressing medication nonadherence is important,(4) and they have a responsibility to assess reasons for nonadherence to medication regimes.(5) Addressing medication nonadherence is a critical step to improve patients’ health.

Assessing for Nonadherence

Nonadherence can be clinically assessed by the everyday hospitalist with several different methods, including by unstructured patient or proxy self-report, a structured questionnaire (such as the Adherence to Refills and Medications Scale or ARMS(6)), prescription fill data, or biochemical detection of drug levels.(7) Other methods, such as direct observation, pill counting, electronic drug monitoring, or digital sensors, are often impractical in the hospital setting. Experts encourage using methods for assessing for nonadherence relevant to the specific phases of nonadherence, including: (1) noninitiation, (2) incorrect implementation, and (3) nonpersistence.(8) For example, if a hospitalist wants to assess if a readmitted patient initiated a medication after a recent hospital discharge (e.g., assessing for noninitiation), then prescription fill data may be the most appropriate measure. Alternatively, if a hospitalist wants to assess if a patient is missing doses, taking them at the wrong time, or improperly administering the medication (e.g., assessing for incorrect implementation), then self-report or a structured questionnaire may be more appropriate. There is no gold standard for measuring adherence, and providers may have to use multiple methods for a proper assessment. However, once it has been identified, a systematic evaluation is required so that clinicians can identify an accurate diagnosis and implement patient-specific interventions.

Diagnostic Schemas

Medical educators often use diagnostic schemas to guide clinical thinking. A diagnostic schema is a framework that clinicians use to shape their clinical reasoning for a particular issue.(9) Schemas allow clinicians to activate a mental framework about a problem and unwind it methodically, rather than generating exhaustive differential diagnoses for every presentation. Physicians use diagnostic schemas every day, such as hyponatremia or fever in a hospitalized patient. Medical students learn how to assess acute kidney injury based on pre-renal, post-renal, and intrinsic causes. These diagnostic schemas are passed down to us from our attendings, fellows, senior residents, and colleagues. Our individualized schemas become more complex and nuanced over time and with accumulating experience. We shape these mental frameworks to develop quicker and more accurate clinical decisions. Medication nonadherence deserves a diagnostic framework due to its prevalence and clinical importance.

The COM-B Framework

Just as there is no gold standard for measuring adherence, the Capability, Opportunity, Motivation, Behavioral (COM-B) framework is one of many social-behavioral models available to assess underlying barriers and facilitators to adherence.(10) Each model brings a unique perspective, but there are several common constructs across models, which include information, knowledge or awareness; beliefs; motivation; abilities or skills; and environment at the micro, meso, and macro levels. Although no single behavioral model has been shown to be superior to another, I suggest the COM-B framework is well-suited as a diagnostic schema for medication nonadherence because it is both practical and comprehensive. Its brief acronym makes it easy to remember, it is widely used in implementation science to identify determinants and guide intervention design, and it encompasses individual, interpersonal, and health-system influences on medication adherence. In addition, its use as a bedside clinical diagnostic schema has not, to my knowledge, been previously described.

The COM-B framework was developed to combine a large number of pre-existing frameworks used to describe behavior and behavioral change.(11) It has been applied in research settings to identify factors associated with medication nonadherence.(12) COM-B is founded on three conditions of behavior: capability (physical and psychological capabilities), opportunity (physical and social opportunity), and motivation (automatic and reflective motivation). A provider can consider three questions to assess medication adherence: (1) “Is the patient capable of adhering to the recommend treatment plan?” (2) “Does the patient have the opportunity to adhere to the recommend treatment plan?”, and (3) “What is the patient’s motivation to adhere to the recommend treatment plan?” The simplicity of this framework allows COM-B to be easier to implement in a busy clinical setting.

There are many contributors to medication nonadherence, and management plans should address the underlying causes. Examples illustrating how hospitalists can use the COM-B framework to identify potential causes of medication nonadherence are presented in Table 1. Even if we understand the many nuances of adherence, failing to dissect the label of “noncompliant”, “nonadherent”, or “poorly controlled” puts the patient at risk of premature closure and other implicit biases.(13) Thus, clinicians should be encouraged to view nonadherence as an invitation for further investigation, just as we would do for chest pain, anemia, or altered mental status. By using the COM-B framework, clinicians can untangle the “differential diagnosis” of nonadherence.

Table 1.

The COM-B framework and its clinical application to medication nonadherence.

Behavioral Domain		Examples of Medication Nonadherence	Potential COM-B–Derived Interventions
Capability	Physical capability	A patient with proliferative diabetic retinopathy who does not have the ability to self-administer the proper dose of insulin.	Consult with an endocrinologist, diabetic nurse educator, pharmacist, or ophthalmologist to develop a safe and effective insulin regimen.
		A patient with hand osteoarthritis admitted for a COPD exacerbation. She has difficulty using her home inhalers.	Consult with a pharmacist, respiratory therapist, pulmonologist or rheumatologist to determine which medications and formulations best accommodate the patient’s physical limitations.
	Psychologic capability	A patient with heart failure and repeated hospitalizations due to volume overload who has limited health literacy.	Have a detailed discussion with the patient to assess their understanding of their illness and prognosis. Use communication strategies like “ask, tell, ask.” Provide clear, written instructions in the patient’s language and at their educational level.
		A patient with untreated substance use disorder and substance-induced mood disorder which prevent them from taking their heart failure medications.	Offer evidence-based treatment for their substance use disorder and mood disorder. Consult with a psychiatrist to identify any other mental health conditions that also may need to be addressed.
Opportunity	Physical opportunity	A patient with heart failure who runs out of their medications. The pharmacy is a significant distance away from their home.	Connect the patient with a mail-order pharmacy, consider prescribing 90-day duration of medications, and/or assist the patient with requesting automatic refills.
	Social opportunity	A patient with unstable housing who does not have a safe place to store their medications.	Consult with a social worker or with community organizations from the patient’s cultural background. Consider patient challenges with tele-communication, transportation, paying for medications (insufficient income, no insurance), appointments, and housing.
Motivation	Automatic motivation	A patient from a community that does not trust the healthcare system and declines MMR vaccination for their children or themselves.	Build trust in healthcare providers by establishing a consistent, reliable, and welcoming relationship with the patient. Consider how provider interactions (verbal and non-verbal) may be interpreted by the patient in the context of their world view.
	Reflective motivation	A patient with diabetes is prescribed metformin but declines to take it due to gastrointestinal side effects.	Practice shared decision-making with the patient to discuss benefits and risks of medications. Provide alternative medications or treatments with well-established effectiveness and side effects.

Abbreviations: COPD, chronic obstructive lung disease; MMR, measles, mumps, and rubella.

Patients are less likely to adhere to their recommended medication treatment plan as the number of contributors to nonadherence increases. For example, a patient with stage 3 chronic kidney disease may have low health literacy (psychological capacity), lack health insurance (social opportunity), and have a general preference for more “natural” treatment options than medications (reflective motivation). Clinicians may have to use multiple strategies to effectively address medication adherence. Interventions that have empirical evidence to improve adherence include patient education, behavioral counseling, simplifying medication regimens, synchronizing refills, clinical pharmacist consultation, reminders, and financial incentives.(7, 14)

Hypothetical Case Study #1

A 68-year-old man with obesity, diabetes mellitus, hypertension, and heart failure with preserved ejection fraction is hospitalized with acute hypoxemic respiratory failure. This is his third hospitalization in six months. The hospitalist reviews the patient’s record and observes that “medication nonadherence” is noted in several recent discharge summaries. In the emergency room, the hospitalist discovers that the patient had not been taking his furosemide, spironolactone, and dapagliflozin as prescribed.

The hospitalist uses the COM-B framework and discovers that the patient had multiple contributing factors to his medication nonadherence, including mild cognitive impairment (psychological capability), difficulty with transportation to his local pharmacy (physical opportunity), and a desire to take fewer medications (reflective motivation). To resolve these issues, the hospitalist works with a pharmacist to (1) review and consolidate the patient’s medications, (2) identify a mail-order pharmacy covered by his insurance to directly deliver his medications to his home, and (3) establish shared decision-making with the patient to identify the most essential medications to maintain his health and maximize his quality of life.

Hypothetical Case Study #2

A 24-year-old woman with moderate asthma and cannabis use is admitted with increased shortness of breath. This is her fourth hospitalization and fifth emergency room visit in the past twelve months. The patient’s one-liner reads “24 yo woman with poorly controlled asthma who p/w SOB. Frequent admissions in setting of missed clinic visits and noncompliance.” The hospitalist uses the COM-B framework and learns that the patient has been prescribed several different inhalers (including a short-acting beta agonist, inhaled corticosteroid (ICS) only, ICS-formoterol, and ICS-salmeterol) and nebulization treatments by multiple prescribers, leading to confusion (psychologic capability). Her mother is showing signs of early-stage Alzheimer’s dementia which has increased her caregiving responsibilities thereby limiting her ability to attend medical appointments outside of work (social opportunity). She and her husband own a car but he uses it to get to work (physical opportunity). She has increased her cannabis use from a few times per week to daily to help with stress and anxiety (psychologic capability). She admits that her cannabis use has likely impacted her health and motivation to schedule follow-up appointments (automatic motivation). She is motivated to reduce her cannabis use (reflective motivation) but has had several unsuccessful attempts to do so.

The hospitalist organizes a multidisciplinary conversation with the pulmonology consultant, respiratory therapist, pharmacist, and social work to come up with a treatment and discharge plan. The pharmacist identifies an inhaler that matches the patient’s formulary and completes a full medication reconciliation with the patient. The respiratory therapist educates the patient on proper inhaler use. The pulmonologist schedules a telemedicine appointment one week after hospital discharge. The social worker helps connect the patient with daytime caregivers and transportation assistance. The hospitalist evaluates the patient for cannabis use disorder, screens the patient for anxiety and depression, and provides evidence-based treatment.

Conclusions

Nonadherence to medication is a common yet complex scenario encountered by hospitalists. It is influenced by a myriad of external and internal factors and three distinct stages of time (i.e. noninitiation, incorrect implementation, and nonpersistence).(8) Framed this way, medication adherence can be considered a quintessential implementation problem. Implementation science seeks to bridge the “know-do gap” between what we know works (i.e. evidence-based interventions) and what we do in clinical practice. Through the issue of nonadherence, we see how medicines with even the strongest evidence-base fail to be implemented in everyday practice. Using the COM-B framework as a diagnostic schema for nonadherence can help bring implementation science to the bedside.(15)

These hypothetical case studies were not intended to be comprehensive or discuss every factor affecting medication adherence. But a diagnostic schema is not meant to be exhaustive. Its purpose is to guide a clinician’s thinking to the most common and most pressing causes when confronted with a common condition. The diagnosis of medication nonadherence is broad, but the COM-B framework of behavior modification is a cognitive tool that can help hospitalists consider additional contributors of nonadherence, and avoid overlooking capability and opportunity factors that can be recognized and directly mitigated.