Differential Diagnoses and Clinical Implications of Medication Nonadherence in Older Patients with Chronic Kidney Disease: A Review

Abstract

Older adults with chronic kidney disease (CKD) often have many comorbidities, which requires them to take multiple medications. As the number of daily medications prescribed increases, the risk for polypharmacy increases. Understanding and improving medication adherence in this patient population is vital to avoiding the drug-related adverse events of polypharmacy. The primary objective of this review is to summarize the existing literature and to understand the factors leading to medication nonadherence in older patients with CKD. We discuss the prevalence of polypharmacy, the current lack of consensus on the incidence of medication nonadherence, the heterogeneity of assessing medication adherence, and the most common differential diagnoses for medication nonadherence in this population. Specifically, the most common differential diagnoses for medication nonadherence in older adults with CKD are: 1) medication complexity, 2) cognitive impairment, 3) low health literacy, and 4) systems-based barriers. We provide tailored strategies to address these differential diagnoses and subsequently improve medication adherence. The clinical implications include deprescribing to decrease medication complexity and polypharmacy, utilizing a team-based approach to identify and support patients with cognitive impairment, enriching communication between health providers and patients with low health literacy, and improving health care access to address systems-based barriers. Further research is needed to determine the effects of addressing these differential diagnoses and medication adherence in older adults with CKD.

1. Introduction

In the United States (US), there are approximately 30 million adults with chronic kidney disease (CKD). CKD is defined by the Kidney Disease Improving Global Outcomes (KDIGO) guidelines as an estimated glomerular filtration rate (eGFR) less than 60 ml/min/1.73m² for greater than three months [1, 2]. Most patients with CKD are older adults: 72% of patients with CKD are over the age of 60 years [3]. Older adults with CKD are at an increased risk for polypharmacy [4], defined as the use of 5 or more medications [5]. Polypharmacy is associated with adverse outcomes, such as increased risk of drug-drug interactions, adverse drug effects, and medication nonadherence [4].

Individuals with CKD often have many comorbidities for which they take multiple medications [4]. In fact, research shows that older adults with CKD take more prescribed medications than non-CKD older adults. Further, as kidney function worsens, there is an associated increase in prescribed medications in older adults [6]. Also, older adults with CKD are more likely to report the use of ten or more daily medications than non-CKD older adults [6]. Thus, nearly all individuals with moderate to severe CKD have polypharmacy [5, 6].

Polypharmacy contributes to medication nonadherence, but polypharmacy may be one of many factors affecting medication adherence for older adults with CKD [4, 7]. Since most patients with CKD are older, age-related physical and cognitive changes may affect medication adherence in this patient population [4]. Understanding factors that contribute to medication nonadherence in CKD is essential to providing tailored medication management solutions to individual patients.

This review discusses the current polypharmacy problem in those with CKD. We summarize the literature on medication adherence measurements in patients with CKD, highlighting the heterogeneity in measurements. Finally, we present possible differential diagnoses related to medication nonadherence in CKD: medication complexity, cognitive impairment, low health literacy, and systems-based barriers. Each differential diagnosis is unique, and we provide individualized, targeted strategies for addressing each differential diagnosis.

2. Polypharmacy is problematic for older adults with CKD

Polypharmacy can lead to inappropriately high doses of medications, drug-drug interactions, and adverse drug reactions [8, 9]. Polypharmacy also contributes to medication nonadherence and recurrent hospitalizations for older adults with CKD [4, 7].Medication nonadherence itself is associated with an increased risk for hospitalizations, worsening of medical conditions, and death [10–12]. In older adults with CKD, medication adherence is critical to slow kidney disease progression and death [4, 13–15]. Cardiovascular disease is the leading cause of death in older adults with kidney disease [16]. Medication adherence is a core element of mitigating cardiovascular risk and preventing cardiovascular events [16, 17]. Thus, polypharmacy and associated medication nonadherence are a critical problem for those older adults with CKD [4, 13–17].

3. Measures of medication adherence in older adults with CKD rely on patient self-report

How can clinicians accurately measure and assess the severity of medication nonadherence in older adults with CKD? There are many methods to measure medication adherence, but there is no gold standard. Most methods, however, rely on some degree of patient self-report. Whether quantitative or qualitative, self-report measures may be limited by the participants’ bias to answer questions in a positive and desirable manner [18–20]. Self-report questionnaires do maintain some benefits, such as practicality, less intrusiveness, and cost effectiveness [18, 21]. While there is no gold standard for measurement of indirect adherence, triangulation of methods remains the best strategy to assess medication adherence [22].

Previous studies have investigated medication nonadherence in the CKD patient population using quantitative methods, most often through self-report questionnaires [18, 19, 23–28]. While over 43 unique self-report medication adherence scales exist [29], variants of the Morisky Medication Adherence Scale (MMAS) are the most commonly used in studies of patients with CKD [19, 23, 26, 28]. The MMAS is a brief self-report questionnaire, often 4 or 8 items, which asks patients about their medication-taking behavior [30, 31]. Two studies utilized open-ended approaches to assess medication adherence [7, 32]. The first study used a CKD self-care scale, which included five questions related to medication adherence [32]. A second study used semi-structured interviews to determine medication adherence and differentiated between medication adherence and nonadherence using thematic analysis but did not offer concrete definitions for either category [7].

4. Prevalence of medication nonadherence amongst older adults with CKD

Between August 2019 and February 2020, we searched the PubMed database using combinations of these keywords: “medication adherence,” “medication nonadherence,” “CKD,” and “medication regimen complexity”. We excluded studies that focused on children and adolescents, did not extensively research medication adherence, or did not exclusively research patients with CKD. The inclusion criteria are outlined in Figure 1. These criteria included studies in which adults with CKD were the primary study population and medication adherence was a primary focus. We included 10 studies that met these inclusion criteria. The literature on medication adherence specifically in older adults with CKD is sparse, and studies include variable methods for adherence assessment [7, 18, 19, 23–28, 32]. This inconsistency makes it challenging to interpret the degree of medication nonadherence in patients with CKD (Table 1) [7, 18, 19, 23–28, 32].

Figure 1.

Selection of articles for inclusion in review

Table 1.

Summary of studies assessing medication adherence in a chronic kidney disease patient population

Study, N	Differential Diagnoses Assessed	Adherence Definition	Average # of Meds	Mean Age (years)	Mean eGFRa	Adherence Assessment	Limitations
Assessment Type: Self-Report Questionnaire
Parker et. al 2019 [26] N = 157	Medication complexity	MMAS-8, scored 0-8.  Low: <6  Medium: 6-7  High: 8	16	76	< 15b	Low: 17% Med: 28%	Self-reported adherence in those aged 65+ years with CKD 5
Ghimire et. al 2016 [28] N = 53	Medication complexity	MMAS-4, scored 0-4  0: Adherent	11	68	ESRD	Nonadherent: 57%	Self-reported adherence in those with ESRD
Tangkiatkumjai et. al 2017 [19] N = 295	Cognitive impairment, Low health literacy	MMAS-8  Low: <6  Moderate-to-high: 6-8	11	68	39	Low: 21% Med-High: 79%	Self-reported adherence in patients living in Thailand
Muntner et. al 2010 [23] N = 3936	Cognitive impairment, Systems-based barriers	MMAS-4  Low: 2-4  Medium: 1  High: 0	N/A	69	65	Low: 31%	Self-reported adherence in a single study visit
Cedillo-Couvery et. al 2018 [18] N = 3305	Low health literacy	Based on 3 questions, missed or added doses in last week.  Low: >1 purposefully  Medium: >1 accidentally	9	59	42	Low: 15% Med: 17%	Self-reported adherence
Hsu et. al 2015 [24] N = 293	Low health literacy, Systems-based barriers	Based on 3 questions, missed or added doses in last month.	12	66	N/A	Low: 53%	Self-reported adherence
Sontakke et. al 2015 [25] N = 150	Systems-based barriers, Low health literacy, Cognitive impairment	8 questions about medication habits, compared to MMAS-8	8	51	N/A	Nonadherent: 34%	Self-reported adherence
Hong et. al 2016 [27] N = 981	None	6 questions about medication adherence plus pill count  Perfect: 100%  Moderately high: 75-99%  Moderately low: 50-74%  Low: <50%	N/A	55	N/A	Low: 24% Med-Low: 22%	Self-reported adherence to antihypertensives only in African-American patients
Assessment Type: Interview
Rifkin et. al 2010 [7] N = 20	Cognitive impairment, Low health literacy	40-minute interview	8	72	N/Ac	N/A	Self-reported adherence in a small sample from a single US institution
Wang et. al 2019 [32] N = 449	None	Utilized the CKD self-care scale, 5 questions address adherence	N/A	64	41	Mean adherence score: 22.92/25	Focused on self-care, not adherence, in Taiwanese patients

eGFR = estimated glomerular filtration rate, MMAS-8 = 8-item Morisky Medication Adherence Scale, CKD = chronic kidney disease, ESRD = end-stage renal disease, MMAS-4 = 4-item Morisky Medication Adherence Scale, N/A = not applicable, US = United States

^a)eGFR measured in mL/min/1.73m²

^b)Included those with eGFR < 15 mL/min/1.73m² only

^c)Twelve participants had CKD stage 3-5; eight participants had ESRD on dialysis

The degree of medication nonadherence reported by older adults with CKD was variable [18, 24]. One medication adherence study found that 15% of patients showed low adherence to their medication regimen [18], while another. using a different adherence assessment, found that up to 53% of patients showed low adherence [24]. One study found that patients with lower self-reported medication adherence were prescribed more pills that those with higher medication adherence: 14 pills per day compared to 9 pills per day [24]. A second study found that individuals with lower adherence were more likely to be of minority race, have more comorbid diseases, and have more depressive symptoms [18].

In several studies, self-reported medication nonadherence was associated with an increased risk for CKD progression [18, 19]. Medication nonadherence in older adults with CKD was also associated with a higher prevalence of adverse medication events [24]. For example, medication nonadherence was associated with higher systolic blood pressure variability between CKD visits, a factor that contributes to the risk of cardiovascular events and death [23, 27].

5. “Differential diagnoses” for medication nonadherence in patients with CKD

Alongside identifying medication nonadherence, it is critical to investigate contributing factors to nonadherence. In the ten studies included in Table 1, eight specifically assessed differential diagnoses to medication nonadherence in older adults with CKD. The most commonly cited differential diagnoses were: 1) medication complexity, 2) cognitive impairment, 3) low health literacy, and 4) systems-based barriers.

5.1. Medication complexity may lead to medication nonadherence

Clinical practice guidelines often advocate for the prescription of medications to treat chronic conditions, but these guidelines rarely contain information for specifically treating older patients with life-limiting illnesses and multimorbidity. Additionally, these guidelines often lack information about when it is appropriate to reduce and remove medications [33]. This contributes to increased medication complexity.

Medication complexity derives from multiple characteristics of a medication regimen, such as a medication’s dosage form, administration route and frequency, as well as administration instructions and timing [26, 34]. Medication complexity may be quantified by the Medication Regimen Complexity Index (MRCI), a tool that quantifies drug regimen complexity with weighted scores for dosage forms, dosing frequency, and additional administration instructions [26, 35]. Quantitatively, higher MRCI scores were associated with a shorter time to hospital readmission amongst older adults with CKD [36].

Qualitatively, increased medication complexity is associated with negative consequences amongst older adults with CKD, including polypharmacy [26]. Increased medication regimen complexity is also associated with both physical and mental health-related quality of life [37]. In older adults with pre-dialysis CKD, high medication complexity and increased dosing frequency were associated with higher perceived medication burden. In the same study, increased perceived medication burden was associated with increased medication nonadherence, suggesting that perceived medication burden, defined as patients’ self-reported beliefs regarding the burdensomeness of their medication regimen, could be a modulator between actual medication complexity and nonadherence [37].

5.2. Cognitive impairment may lead to medication nonadherence

Importantly, forgetfulness may be an early sign of cognitive impairment [38]. Forgetfulness accounts for about 30% of medication nonadherence [39] and is common amongst older adults with CKD [7, 19, 23–25]. In one study, 44% of older adults with CKD cited forgetfulness as the reason for medication nonadherence [19]. In another study, 46% answered affirmatively to forgetting to take a pill at least once in the preceding thirty days [24].

Cognitive impairment is prevalent in older adults with CKD: 20-50% individuals with moderate CKD have coexisting cognitive impairment [40]. In advanced CKD, up to 70% of patients have cognitive impairment [41]. This may be due to CKD’s association with other vascular comorbidities and the high rate of cardiac events in older adults with CKD [42]. Individuals with CKD are more likely to have cerebrovascular and vascular disease, major contributors to the development of cognitive impairment [43]. Cognitive impairment should be considered as a possible differential diagnosis for medication nonadherence, especially when patients report forgetfulness. Other CKD-related factors such as uremic metabolites, anemia, and drug-drug interactions can negatively affect cognitive function, and should be considered as well [43–45].

5.3. Low health literacy may lead to nonadherence

Another differential diagnosis to medication nonadherence is low health literacy [46]. Regardless of their age, patients with low health literacy may not understand the need to take chronic medications every day [46]. Fear may also contribute to medication nonadherence in patients with low health literacy; patients may miss doses or intentionally skip doses due to either fear of becoming “dependent” on a medication [46] or fear of medication side effects [25]. In an observational study, 47% of individuals with CKD identified a fear of adverse events and 27.5% reported being unaware of the seriousness of their condition as contributors to their nonadherence [25].

In a systematic review on health literacy in older adults with CKD, the prevalence of limited health literacy was 25%, with between-study heterogeneity. Also in this review, nonwhite ethnicity and lower socioeconomic status were independently associated with lower health literacy, while age was not [47]. A separate review on health literacy in patients with CKD also reported a prevalence of low health literacy of approximately 25% with disproportionately lower levels of health literacy in those with lower socioeconomic status and nonwhite ethnicity. This review also suggested that lower health literacy increases the risk of inferior health outcomes and encouraged future research studies to investigate this possible relationship further [48].

For asymptomatic conditions, missing a dose of a medication may not result in immediate adverse health effects. A lack of understanding of each medication’s importance can lead to patients with complex medication regimens prioritizing medications, adhering to some medications but not others [7, 24, 25]. For example, older adults with CKD have reported prioritizing adherence to antihypertensives over medications for asymptomatic comorbidities of their CKD (i.e. lipid-lowering agents) [7]. Many of these patients were unable to explain how their prescribed medications were beneficial for their kidney disease, suggesting low health literacy as a contributor to nonadherence [7]. Assessing a patient’s knowledge and beliefs about their medications is critical to understanding how health literacy affects medication adherence [18].

5.4. Systems-based barriers may be underlying factors in medication nonadherence

Systems-based barriers, such as medication cost, limited pharmacy access, and insurance status, may be on the differential diagnoses for medication nonadherence [23–25, 49]. Cost can be a barrier to medications as we age. Older adults in the US may be retired and on a fixed income, relying on Medicare for prescription drug coverage [23, 24, 50]. Patients with chronic illnesses require many medications for extended periods of time, leading to sustained medication costs [51]. Medications to treat CKD and its sequelae cost the individual and Medicare significantly, as older adults with CKD are living well into their 80s [52]. The cost of CKD and the medications associated with chronic disease state management can compound, leading to decreases in medication adherence. Amongst those with CKD, 63% cited high cost as a reason for medication nonadherence [25]. Indeed, individuals with CKD report skipping doses to make their medication supplies last longer to ultimately decrease costs [50].

Changes in physical function with age or CKD progression may reduce mobility. This can lead to issues with access to care, including obtaining medication refills [53]. Also, older adults living in a less densely populated area may have limited access to health professionals and live further away from a pharmacy to fill prescriptions [54]. Lack of insurance coverage is also associated with medication nonadherence [24, 55]. Outside of Medicare coverage, one study found that the majority of study participants with CKD did not have health insurance, and 54% believed that the government should subsidize their medications [25]. Older adults with CKD with no insurance were more likely to delay visits to the doctor and delay filling their medications [24].

6. Tailor strategies to address the differential diagnoses for medication nonadherence

Uncovering the underlying factors that contribute to medication nonadherence is essential [46]. Then, clinicians can implement tailored strategies addressing these factors to improve medication adherence for older adults with CKD. The recommended individualized and targeted strategies for these differential diagnoses are outlined in Table 2.

Table 2.

Individualized and targeted strategies to address differential diagnoses to medication adherence

Differential Diagnosis	Individualized, Targeted Strategy
Medication Complexity	• Deprescribe inappropriate or unnecessary medications [4, 54, 55]   • Optimize medication regimens by reducing administration frequency and converting to long-acting medications when possible [4, 56, 57]   • Utilize combination therapy when appropriate to minimize the daily number of medications taken by a patient [60–62]   • Consider a clinical pharmacy consult for medication review and assistance [55, 58, 59]
Cognitive Impairment	• Screen for cognitive impairment early [38]   • Work with an interprofessional team to care for patients with cognitive impairment [46]   • Use technology, such as automatic and electronic reminders, to enforce medication adherence [63, 64]   • Encourage caregiver support when necessary [7, 50, 65]
Low Health Literacy	• Effectively communicate with patients by eliminating medical jargon and speaking in plain language [69, 71]   • Use teach-back method to confirm patient’s understanding [71]   • Provide a combination of verbal instructions and printed materials to patients [72–75]   • Complete a “brown-bag review” with patients to address all of their questions and concerns about their prescribed medications [8, 76]   • Build a lasting and trusting relationship with patients [46, 50]
Systems-Based Barriers	• Implement policy changes to improve healthcare access [73, 77]   • Choose an individualized drug regimen that will be affordable to patients who have concerns with drug cost [50, 78, 79]   • Work with health systems to advocate for fair drug pricing for patients [46, 73, 77]

6.1. Deprescribe to address polypharmacy and medication complexity

Deprescribing involves the systematic elimination of unnecessary and/or inappropriate medications. Deprescribing helps reduce polypharmacy and associated medication costs. Evidence-based deprescribing can improve medication adherence [4, 54, 55] and is associated with increased patient satisfaction and a lower risk of adverse events [4, 56, 57]. Collaboration with a pharmacist to deprescribe medications can ensure that CKD medication regimens are simplified, safe, and effective [58]. Pharmacist-performed deprescribing may lead to lower medication costs for patients and may improve the efficiency of prescription refills through refill consolidation and synchronization, both of which may improve medication adherence [55, 59]. Pharmacist-led medication reviews can optimize medication appropriateness in older patients with CKD [59].

The utilization of combination therapy can help to reduce pill burden and thus improve medication adherence [60]. Fixed-dose combination therapy is defined as a single pill containing multiple medications. By using combination therapy, the total number of daily medications taken by a patient is minimized [60]. Patients with CKD and other comorbid conditions, such as hypertension or hyperlipidemia, can benefit from the use of combination therapy when appropriate [61, 62].

6.2. Consider a team-based approach to addressing cognitive impairment

When older adults report forgetting to take their medications, consider evidence-based methods for assessing cognition [38]. It is crucial to reach out to an interprofessional team with experience supporting patients with cognitive impairment if and when it is identified. Technology may improve medication adherence in those with mild cognitive impairment. For example, clinicians, family members, or caregivers may set up automated reminders to improve medication adherence [46]. Telephone and tele-video reminders can also reinforce medication adherence in older adults with CKD experiencing cognitive impairment [63, 64].

Memory aids and pill boxes can positively improve medication adherence for those with mild cognitive impairment; however, these strategies will fail as cognition worsens [7, 50]. As cognition worsens, involving family members or caregivers can improve medication adherence [7, 50, 65, 66]. Older adults with CKD have reported that assistance from family members and/or caregivers improves medication adherence [7]. Consider consulting a geriatrics and/or home health team if patients are no longer able to manage their medications due to cognitive impairment, regardless of whether or not a family member or caregiver can do so [63, 67, 68].

6.3. Better communication can improve medication adherence for patients with low health literacy

Medical encounters can be thorough, yet because of the complexity of the visit, patients, regardless of their age, may remember as little as 50% of the content discussed during the visit [46, 69]. According to the Program for the International Assessment of Adult Competencies, 19% of adults in the United States operate at low levels of literacy and 29% operate at low levels of numeracy [70]. Clinicians can effectively communicate with their patients using the following strategies. Clinicians can eliminate medical jargon in favor of simplified, plain language when communicating with all of their patients. The teach-back strategy, asking patients to explain a concept back to the healthcare professional, may also deepen their understanding, and has been shown to improve medication adherence [71]. Complementing verbal instructions with printed materials is also effective [72, 73]. Finally, the use of pictorial medication lists that include medication names and dosage information have been shown to improve medication adherence for those with low health literacy [63, 74, 75].

A “brown bag review” is another way to improve medication-related communication [8, 76]. During the brown bag review, the patient brings all of their medications to a visit with a clinician. This is an opportunity for patients to ask questions about their medications. The clinician may assess medication-taking behavior and discuss strategies for improvement, all of which may improve medication adherence. Simply knowing the indication for each medication has led to improved medication adherence [50]. Similarly, acknowledging patient experiences with medication side effects and working toward managing them increases medication adherence [7]. Finally, the brown bag review may allow the clinician to clarify discrepant information that the patient received which may have negatively affected their adherence.

The final communication tool is building trust [46, 50]. When clinicians leverage patient-centered communication tools, over time, a trusting relationship may be built [24, 46]. Continuity of care and of relationships between patients and clinicians are both associated with increased medication adherence, reduced likelihood of hospitalization, and lower healthcare expenses [46]. Additionally, involving the patient in the decision-making process is a critical component of building trust and improving medication adherence [24, 50]. Older adults with CKD have reported being more likely to adhere to medications when they felt involved and knowledgeable about their healthcare plan [50].

6.4. Addressing the factors contributing to systems-based barriers may improve access to care and medication adherence

Improved access to healthcare insurance improved medication adherence [46]. Macro-level policy changes, such as the Affordable Care Act, may improve health literacy and enable individuals to enroll in insurance plans and gain access to healthcare treatment [73, 77]. For those with or without insurance, medication cost may be an issue. Older adults with CKD have noted that the high cost of buying prescriptions was a barrier to medication adherence [50]. Extensive research has been conducted in the diabetes patient population, and there are significant correlations between decreased medication and treatment costs and improved medication adherence [78–81]. When possible, clinicians should work with their health systems to advocate for fair drug pricing for their patients.

7. Conclusions

Understanding the factors influencing medication adherence can improve future health outcomes in older adults with CKD. Older patients with CKD are prescribed a greater number of medications than older adults without CKD, leaving CKD patients at an increased risk for polypharmacy. While there is currently no gold standard for measuring medication adherence, clinicians can use validated measures, like the MMAS, to assess the presence and degree of medication nonadherence. When medication nonadherence is identified, clinicians should consider potential differential diagnoses leading to this nonadherence. The most common differential diagnoses are: medication complexity, cognitive impairment, low health literacy, and systems-based barriers. As recommended in this review, deprescribing to decrease medication complexity, utilizing a team-based approach to address cognitive impairment, enhancing physician-patient communication to increase health literacy, and improving access to healthcare are strategies to address these differential diagnoses. Future research is needed to establish a new standard of care for identifying and addressing medication nonadherence in older adults with CKD.

Key Points.

• Older patients with CKD are prescribed greater numbers of medications, leaving them at increased risk for polypharmacy.

• The differential diagnoses for medication nonadherence include medication complexity, cognitive impairment, low health literacy, and systems-based barriers.

• Strategies to address these barriers to medication nonadherence are deprescribing, utilizing a team-based approach, enhancing the physician-patient relationship, and improving access to healthcare.

Availability of Data and Material

Not applicable