Medication Appropriateness, Polypharmacy, and Drug-Drug Interactions in Ambulatory Elderly Patients with Cardiovascular Diseases at Tikur Anbessa Specialized Hospital, Ethiopia

Introduction

The world’s aging population is speedily increasing from 9% to 16% between 2019 to 2050.¹ Increasing age is associated with higher rates of co-morbidity, disability (cognitive, mental, and functional impairments),² and multiple medication use.^3–5 Aging influences heart activities by decreasing cardiac muscle elasticity and the ability to respond to changes in pressure.⁶ Hence, it is an independent risk factor for cardiovascular disease (CVD) which poses a great burden to elderly patients.⁷ Globally, 32% of the deaths were due to CVD. Of which, three fourth of them were from low and middle-income countries.⁸

Due to multiple chronic diseases, and age-related physiological changes that affect drugs’ pharmacodynamics and pharmacokinetic properties, pharmacotherapy is complex and challenging in the elderly.⁹ Polypharmacy, non-adherence, inappropriate medication prescription,^10,11 drug-drug interaction,^12,13 and medication-related problems¹⁴ are frequent in elderly patients with CVD. Due to this, elderly patients are at a higher risk of adverse drug events (ADEs).¹⁵

Explicit and implicit criteria are used to assess medication safety in the elderly. The explicit criteria (the 2019 American Geriatrics Society (AGS) Beers Criteria)¹⁶ involves six medication appropriateness issues, while the implicit criteria (Medication Appropriate Index (MAI))^17,18 uses ten criteria. The explicit criteria are primarily concerned with medications and assess how well they meet a set of predetermined criteria, criterion-based with rigid standards. On the other hand, the implicit criteria consider healthcare professionals’ judgment to assess medication appropriateness.

In Ethiopia, the pooled prevalence of CVD was 5% (1–20%) in the general population,¹⁹ but its prevalence is expected to be higher in the elderly population. Despite CVD is the leading cause of death and disability among the elderly,²⁰ data are scarce regarding the appropriate use of medication in Tikur Anbessa Specialized Hospital (TASH). Hence, this study aimed to investigate medication appropriateness using both the 2019 AGS and MAI criteria and associated factors in ambulatory elderly cardiac patients at TASH.

Methods

Study Setting

The study was conducted at TASH, As Ababa, Ethiopia. It is the largest referral hospital in Ethiopia, established in 1964. It has more than 800 beds and serves over 250,000 outpatient and 48,000 inpatients annually. Several clinics serve ambulatory patients. The cardiac outpatient clinic serves more than thousands of patients with CVDs.

Study Design and Period

A retrospective cross-sectional study was performed in a cardiac clinic of TASH, Addis Ababa, Ethiopia. The study was conducted from May 01– August 30, 2021.

Study Population and Sample Size

All elderly patients (aged ≥ 60 years) with CVD were the source population. The sample population consisted of elderly CVD patients who are on follow-up during 2018–2021. As per the cardiac clinic officer, there were a total of 1993 elderly CVD patients during the mentioned follow-up period in TASH. The sample size was then calculated by using a simple proportion formula by considering a confidence interval of 95%, marginal error of 5%, and, the prevalence of 50%. The calculated sample size was 384. Systematic random sampling method was used to choose 384 patients’ medical records. Patients with incomplete medical records were excluded.

Study Variables

The dependent variables were medication appropriateness, polypharmacy and drug-drug interactions (DDIs). On the other hand, the independent variables include socio-demographic characteristics, clinical characteristics (comorbidity, Charlson comorbidity index (CCI)), and prescribed number of medications.

Data Collection Tool and Procedure

The data collection form was prepared based on published study findings and standard tools, which consisted of four sections. The first section included questions on the demographic characteristics of patients. The second section contained clinical and medication-related information. The third section was the AGS Beers Criteria and, the fourth section was MAI tool (Supplementary Table 1).

Data were collected retrospectively from patient medical records. The most recent visit was considered for assessment. Medication appropriateness (Supplementary File) was evaluated according to one explicit (AGS Beers’ criteria) and one implicit (MAI) criteria. Both prescription and Over The Counter (OTC) medications were considered. On the contrary, herbal supplements were excluded as they were not well documented. The collected data were thoroughly reviewed and analyzed by experienced clinical pharmacists. DDI was checked using Drugs.com online drug interaction checker. Polypharmacy is considered when the patient is prescribed five or more medications.

Data Quality Control

Before the actual research, a pre-test was done in 5% of patients to check for data collection instrument clarity, consistency, and acceptability. All required changes were made based on the results of the pre-test.

Ethical Consideration

The ethical clearance was obtained from the ethical review committee of the Department of Pharmacology and Clinical Pharmacy (ERB/SOP/315/13/2021), College of Health Sciences, Addis Ababa University. Then, patients, during follow up time, were briefed about the objectives of the study and the confidentiality of their information. In addition, they were informed of the right to abstain from participation in the study or to withdraw consent to participate at any time without reprisal. After ensuring that the patient has understood the information, the data collectors obtained written informed consent.

Statistical Analysis and Interpretation

Data were entered, cleaned analyzed using the Statistical Package for Social Sciences (IBM SPSS Statistics for Windows, version 23). Descriptive statistics were performed to summarize the sociodemographic and clinical characteristics. The results were presented in frequencies, percentages, tables, and figures. Logistic regression was used to determine factors associated with inappropriate medication use which is assessed by the AGS Beers Criteria. The results of the multivariable logistic analysis were reported using adjusted odds ratio (AOR) and 95% CI. Statistical significance was declared at p < 0.05.

Results

Demographic and Clinical Characteristics of the Study Participants

A total of 384 ambulatory elderly cardiac patients were included in the study. The mean age of patients was 68.1 ± 6.5 years, ranging from 60–92 years. Majority (82.6%) of patients were young-old (60 to 74 year age). In addition, 52.6% of them were females, and the majority of (93.7%) the elderly patients had at least one comorbidity (mean 2.68 ±1.2). The average duration of the disease was 8.14 ±7.4 years. Further, 57.6% of patients had ≥ 5 CCI (Table 1).

Table 1 Demographic and Clinical Characteristics of the Study Participants (N=384)

Disease and Medication Use Pattern

The most frequent CVD encountered were hypertension (78.4%), ischemic heart disease (31.3%), hypertensive heart disease (26.3%), degenerative valvular disease (16.7%), and atrial fibrillation (11.2%).

A total of 1698 medications were prescribed, averagely 4.4 ± 1.9 per patient (range; 1–12 medications). The most frequently prescribed medications were atorvastatin (65%), enalapril (58.6%), aspirin (46.9%), furosemide (44.3%), and amlodipine (37.2%) (Table 2).

Table 2 The Most Frequent CVD Medications Prescribed Among Study Participants

Medication Appropriateness Among Ambulatory Elderly Patients

Explicit Beer’s Criteria (AGS Beers Criteria)

Over one-third of (28.1%) of elderly CVD patients were prescribed at least one potentially inappropriate medication (PIM), averagely 1.51 ±1.3 per patient. Of these patients, 19.8%, 26.6%, 17.4%, and 6% were prescribed one, two, three, and four or more PIM, respectively. A total of 617 medication-related issues were identified. 54.9% (277), 28.1% (108), 13.4% (38), 4.6% (18) and 1% (4) of patients prescribed medications that should be used with caution, PIM as per the Beers criteria, medications that should be avoided or reduce dose in renal failure, potentially inappropriate medications to avoid in older adults with certain conditions, and medications with clinically important DDIs to be avoided, respectively.

The frequently prescribed inappropriate medications were diuretics (46.29%) and antiplatelet agents (31.5%). (Figure 1) In addition, the frequently prescribed medications that should be avoided in the elderly were amitriptyline (29.6%), proton pump inhibitors (21.3%), digoxin (20.3%), glibenclamide (16.3%), and non-steroidal anti-inflammatory medications (NSAIDs) (12.3%).

Figure 1 Inappropriate drug classes based on the AGS Beers criteria.

Implicit Criteria (MAI Criteria)

The majority (365, 95.1%) of elderly patients were prescribed one or more PIM. The average MAI score per patient was 18.2 ±12.3 (median 17; range 0–90). The criteria with the highest MAI score were effectiveness, correct dosage, unnecessary duplication, and indication-related issues. On the other hand, the lowest score was the duration of therapy. Considering the cumulative MAI, 320 medications (84.4%) met inappropriate ratings of 0–30 weighted average MAI, 57 (14.8%) met inappropriate ratings of 31–60 weighted average MAI and 3 (0.8%) met inappropriate rating of ≥ 61 weighted average MAI (Table 3).

Table 3 Distribution of Inappropriate Prescriptions Using MAI Criteria

Polypharmacy and Drug-Drug Interactions in the Elderly

More than half the study participants had polypharmacy (53.1%, 212) and, the majority had (90.1%, 346) potential DDIs which can range from mild to major interactions. Of all DDIs, 15.1%, 75%, and 83.3% were major, moderate, and minor interactions, respectively. (Table 4) The most frequently prescribed major DDIs were enalapril and spironolactone (73%, 46) (causes hyperkalemia), amlodipine and carbamazepine (11.1%, 7) (reduces the blood levels of amlodipine), and enalapril and allopurinol (3.2%, 2) (increases hypersensitivity to allopurinol) (Table 5).

Table 4 Polypharmacy and Drug-Drug Interaction in the Study Participants

Table 5 List of Severe DDI, Their Prevalence, and Expected Negative Effects

Factors Associated with Inappropriate Medication Prescription

Patients with polypharmacy were 5.1 times riskier to have inappropriate prescriptions (AOR=5.1, 95% CI [0.110,0.386] p<0.001) than patients having less than five medications (Table 6).

Table 6 Factors Associated with Inappropriate Medication Use

Discussion

This study analyzed medication appropriateness, polypharmacy and DDI in the ambulatory elderly (aged 60 years and above) patients with CVD at TASH, Addis Ababa with a focus on prevalence and associated factors of PIM.

Our findings showed that 28.1% of the patients received at least one PIM according to the AGS Beers Criteria. Ethiopian studies conducted in geriatric patients at Tirunesh Beijing Hospital, University of Gondar Teaching Hospital chronic care clinic, and Jimma Medical Center reported that 46.9%,²¹ 47.2%,²² and 83.1%²³ of older adults were prescribed at least one PIM, respectively. In these studies, medication appropriateness was assessed using Beers criteria in two studies,^22,23 while the remaining study²¹ used both WHO core prescribing indicators and Beers criteria. In addition, our finding was lower than a study done in the United States, with 87.4% of the patients receiving at least one PIM.²⁴ As it is seen in different studies, PIM is common in the elderly. However, the different prevalence may be due to variation in case-mix, availability of medications, type and number of comorbidities.

Our study also revealed that more PIM was detected using MAI than Beers criteria (MAI:95.1%, Beers criteria: 28.1%), which is comparable to a study conducted in Spain (94.1% PIM by MAI, 68.8% by Beers criteria).²⁵ In addition, Awad et al²⁶ detected more cases of PIMs using MAI (74%) than Beers criteria (53.1%). The reason for this could be the Beers criteria have rigid criteria that focus on the drug/drug class, and it does not address patient preference, while MAI focuses on the patients who address the entire medication regimen with the clinical conditions.

Diuretics (46.3%), antiplatelets (31.5%), and tricyclic antidepressants (5.7%) were the frequently prescribed PIMs in our study. However, angiotensin-converting enzyme inhibitors (ACEIs) (15.6%), a combination of aspirin and Vitamin K Antagonist (14%), and statins (3.16%) were the common PIMs reported by Abegaz et al.²⁷ In addition, Tesfaye et al²³ found diuretics (46.3%), antiplatelets (31.5%) and tricyclic antidepressants were the common PIMs. Further, a study conducted in Serbia showed that proton pump inhibitors (15.7%) were the frequent PIM according to Beers criteria.²⁸ Such variation may be due to the different clinical conditions, comorbidities, and disease severity of the studied population.

More than half the patients (53.1%) were prescribed polypharmacy, which is higher than the studies conducted in various parts of Ethiopia (10.8–42.5%).^21–23 As aging is associated with multiple comorbidities,^29,30 polypharmacy is common in the elderly. Similarly, DDI is common in elderly patients.³⁰ Our study revealed that 90.1% of patients had potential mild to major DDIs, which lies in the recent systematic review finding (DDIs: 80.5–90.5%).³¹ However, it is higher than Bhagavathula et al (2.8%)²² and Assefa et al (84.3%) studies. Of all the DDIs, enalapril and spironolactone interactions (73%) were the most frequent DDI like de Oliveira et al study result.³¹ On the other hand, van Heerden et al found that central nervous system medications (30.6%) were the frequent causes of DDIs.³² The variation could be due to differences in geriatric circumstances as well as methodological discrepancies, particularly in the methods/software utilized to detect DDIs. In addition, our study revealed that 15.1% of patients had major DDIs, which is relatively lower than the studies conducted in studies conducted at Ethiopian Yekatit 12 medical college hospital (17.3%),¹² and in the South African elderly ambulatory patients (25.8%).³² The difference could be due to the variation in the case-mix of targeted patients, and professional expertise.

Regarding risk factors, polypharmacy is significantly associated with PIM (AOR=5.1 95 CI [0.110,0.386], p=0.000). Polypharmacy, which was more common with older people, is recognized as the primary contributor of PIM in the Serbian study.³³ In addition, a study done in France reported that the presence of comorbidities, taking multiple medications, and the higher CCI were associated with the likelihood of PIM.³⁴ As the number of medications is increased, the possibility of potential DDIs and drug-disease interactions are increased. This predisposes the patient to have PIM. The variation of PIM risk factors could be due to the type and number of comorbidities, and the case mix of the study population.

The strengths of our study include; firstly, patient-specific factors were considered to evaluate PIM. Secondly, two methods of medication appropriateness were used. On the other hand, there were certain limitations which include (i) doctors were not interviewed to identify their reasons for prescribing the identified PIMs; (ii) the exclusion of herbal supplements may have resulted in an underestimation of PIMs; (iii) the study was done in a single study, and (iv) some medication appropriateness lists were difficult to analyze due to the retrospective nature of the study.

Conclusion

The study found a considerable number of patients prescribed PIM. MAI identified more PIM than the Beers criteria. More than half of the patients prescribed polypharmacy. In addition, the majority of the patients experienced potential mild to major DDIs. Of which, around 15% had major potential DDIs. Moreover, PIM is significantly associated with polypharmacy. These findings highlight the need for interventions that improve appropriate prescription in elderly patients, reduce the number of medications and reduce DDIs.