Prevalence of potentially inappropriate medication use and its association with hospitalization among older adults with chronic disease in the Amhara region,Ethiopia: a multicenter prospective cohort study

Abstract

Background:

Potentially inappropriate medications (PIMs) are common among older adults with chronic diseases and are linked to adverse outcomes, including hospitalization. Evidence on PIM prevalence and its clinical impact in Ethiopia is limited. This study assessed the prevalence of PIM use and its association with hospitalization among older adults in the Amhara Region, Ethiopia.

Objectives:

To determine the prevalence of PIM use and identify factors associated with PIM exposure and hospitalization in older adults with chronic diseases.

Design:

Multicenter prospective cohort study.

Methods:

Between May 1 and November 30, 2024, 1700 adults aged ⩾60 years were enrolled from five comprehensive specialized hospitals in the Amhara region. PIM use was assessed using the 2023 American Geriatrics Society Beers criteria. Sociodemographic, clinical, medication, and hospitalization data were collected via structured interviews and medical chart reviews. Multivariable logistic regression identified factors independently associated with PIM use and hospitalization.

Results:

PIM use was identified in 41.1% of participants. Exposure to PIMs significantly increased the risk of hospitalization (adjusted odds ratio (AOR) = 3.70, 95% confidence interval (CI): 2.25–4.95, p = 0.023). Independent predictors of PIM use included khat chewing (AOR = 1.95), cor pulmonale (AOR = 2.28), degenerative diseases (AOR = 3.20), Charlson Comorbidity Index >4 (AOR = 4.50), prolonged chronic illness (AOR = 3.07), benzodiazepine use (AOR = 1.80), and concurrent benzodiazepine-opioid use (AOR = 4.02). Regular medication reviews were protective, reducing the risk of PIM use (AOR = 0.55).

Conclusion:

PIM use is highly prevalent among older adults with chronic diseases in the Amhara Region and is associated with increased hospitalization risk. Systematic medication reviews and improved prescribing practices are essential to enhance medication safety and reduce preventable hospital admissions.

Plain language summary

How using unsafe medicines affects hospital visits for older adults with long-term illnesses in Amhara region, Ethiopia: a study in five hospitals

This study in Amhara, Ethiopia, found that 41% of older adults with chronic illnesses were given medicines that may be unsafe for them. These medicines increased their risk of hospital visits. People who chew khat, have many health problems, or use certain drugs like benzodiazepines were more affected. Regular medicine reviews helped reduce unsafe drug use and hospital stays. The study highlights the need for better medicine checks to keep older adults safer.

Keywords

Amhara region beer criteria chronic conditions Ethiopia older adult PIMs

Introduction

The global population is aging rapidly, with projections of 2.1 billion individuals aged 60 and older by 2050. This increase is associated with a rise in chronic diseases prevalent in older adults, necessitating complex medication regimens (polypharmacy), which heighten risks of adverse drug events (ADEs) and medication non-adherence. Potentially inappropriate medications (PIMs) pose additional risks, particularly because of physiological changes in older adults that increase drug sensitivity. Common PIMs, such as benzodiazepines and nonsteroidal anti-inflammatory drugs (NSAIDs), can lead to serious health outcomes like falls, gastrointestinal bleeding, and higher healthcare costs. To enhance patient safety and resource utilization, it is essential to minimize PIM prescribing in older adults, particularly in low-resource settings.^1,2

The 2023 Beers Criteria by the American Geriatrics Society (AGS) serves as a key tool for identifying PIMs in older adults. These evidence-based criteria outline medications to avoid, necessary dose adjustments for comorbidities, and highlight drug interactions that could lead to adverse events. They are adaptable for use in various healthcare settings, including low- and middle-income countries (LMICs) like Ethiopia, where older adults face heightened challenges due to limited access to geriatric care and high rates of polypharmacy. The implementation of such standardized tools can enhance medication evaluations and promote safer prescribing practices, thereby improving health outcomes in older populations globally.^3–5

Globally, about 37% of older adults are prescribed at least one PIM, with the highest prevalence in Africa and South America at nearly 47%. This issue is linked to factors such as limited access to geriatric care and inadequate prescriber training. By contrast, North America and Oceania have lower PIM rates of 29% and 24%, respectively, attributed to better prescribing guidelines and healthcare resources. These disparities highlight the need for targeted interventions to improve medication safety for older adults worldwide.^2,6–8 In Ethiopia, studies suggest a significant issue with PIM use among older adults, with a cross-sectional study revealing that 45.7% of those aged 65 and older received at least one PIM. This practice correlates with heightened risks of ADEs, increased healthcare costs, and higher rates of hospitalization. The findings underscore the urgent need for targeted interventions to enhance prescribing practices and medication safety for the country’s expanding elderly demographic.^2,9

In the Amhara region, systemic challenges increase the risk of inappropriate medication prescribing and hospitalizations among older adults. Limited access to healthcare facilities and a shortage of geriatrically trained professionals hinder comprehensive medication management. High rates of polypharmacy, due to multiple chronic conditions, elevate the risk of adverse drug interactions and other negative clinical outcomes. Addressing these issues is essential for enhancing medication safety and healthcare quality for older adults in the area.¹⁰ Despite the recognized issues linked to PIM use among older adults in Ethiopia, there is insufficient multicenter data on its prevalence and impact on hospitalizations. Current studies tend to be single center or regional, limiting generalizability. Multicenter research is necessary to better understand PIM prescribing patterns, identify at-risk groups, and assess PIM-related hospital admissions. This evidence is crucial for policymakers and healthcare providers to develop targeted interventions to improve prescribing practices and clinical outcomes for older adults in Ethiopia.

This study, therefore, aimed to assess the prevalence of PIM use and its association with hospitalization among older adults with chronic diseases across multiple healthcare settings in the Amhara region, Ethiopia. The findings are expected to inform strategies to improve medication safety and optimize clinical outcomes in older adults within the region.

Methods

Study period, setting, and design

This multicenter prospective cohort study was conducted in five similar comprehensive specialized hospitals in the Amhara region from May 1 to November 30, 2024. A total of 1700 older adults were enrolled, with sample sizes proportionally allocated across study sites based on patient load (see Table S3 for hospital characteristics and allocation details).

Operational definition

PIM: Medications that are considered potentially inappropriate for use in older adults according to the 2023 AGS Beers Criteria.

Chronic disease: Long-term health conditions diagnosed by a physician, including hypertension, diabetes mellitus, chronic obstructive pulmonary disease, and heart failure.

Polypharmacy: The concurrent use of five or more medications by a patient.

Multimorbidity: The presence of two or more chronic diseases in the same individual.

Hospitalization: Admission to a hospital for at least 24 h during the study follow-up period.

Readmission: Hospitalization occurring within 30 days after discharge from a previous hospital stay.

Length of hospital stay: The number of days from hospital admission to discharge.

Lost to follow-up: Participants who could not be contacted or did not complete the study follow-up period.⁶

Follow-up period: The duration over which participants were observed for hospitalization outcomes, from enrollment until the end of the study or censoring.

The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement,¹¹ and the corresponding STROBE checklist is provided as Supplemental Material.

Enrollment and follow-up timing

Participants were enrolled from May 1 to November 30, 2024, with follow-up until either hospitalization or a maximum of 6 months. This prospective follow-up facilitated accurate documentation of incident hospitalizations. The study primarily measured the prevalence of PIMs and secondarily assessed hospitalization outcomes.

Outcome measurement

PIM exposure was assessed using the 2023 AGS Beers Criteria by trained pharmacists through structured interviews and medical chart reviews. Medications were classified as PIMs based on their risks of adverse effects or interactions, particularly for older adults. The primary outcome was all-cause hospitalization within 6 months, defined as any hospital admission lasting over 24 h, which included various medical complications. Data were collected from hospital records and patient interviews to establish a link between PIM exposure and hospitalization. Hospitalization was analyzed as a binary outcome, utilizing logistic regression to assess the relationship between PIM use and hospitalization while controlling for factors such as age, sex, comorbidities, polypharmacy, and behavioral influences.

Study population and eligibility criteria

The study focused on patients aged 60 and above, who had been on medication for a minimum of 3 months for any condition while attending outpatient clinics or being admitted to certain hospitals. Exclusions applied to those with severe cognitive or psychiatric issues, those in palliative care, patients with incomplete medical records, transfers from outside the Amhara region, or individuals who refused participation.

Sample size and sampling technique

The sample size was determined using the single population proportion formula, based on a previously reported prevalence of PIM use of 37% among older adults in Ethiopia,¹² which employed a similar assessment method.⁹ With a 95% confidence level (Z = 1.96) and a 5% margin of error, the initial sample size was calculated as

n_{0} = \frac{{(Z_{α / 2})}^{2} \times p \times (1 - p)}{d_{2}} = \frac{{(1.96)}^{2} \times (0.37 \times 0.63)}{{(0.05)}^{2}} = 359

To account for clustering due to the multicenter design, a design effect of two was applied, yielding:

n_{1} = 359 \times 2 = 718

After adjusting for a 10% non-response rate:

n = \frac{718}{0.90} = 798

To improve the accuracy of prevalence estimates and provide sufficient statistical power for subgroup analyses and multivariable modeling, the sample size was increased to 1700 participants, which reduces the margin of error and enhances the validity and generalizability of findings among diverse patient populations. Proportional patient allocation from each hospital is detailed in Supplemental Table 7.

Data collection instruments and procedures

Data were collected by trained pharmacists using standardized, validated instruments, including a pretested questionnaire and a chart review checklist.^1,9,13,14 The study involved pretesting a questionnaire for clarity and cultural appropriateness and training pharmacists on study protocols. Eligible participants, older adults (⩾60 years) in the Amhara region, provided informed consent before structured interviews, and medical chart reviews were conducted to gather sociodemographic data and clinical histories. PIMs were assessed based on the 2023 AGS Beers Criteria, with high inter-rater reliability among pharmacists. The primary outcome measured was all-cause hospitalization over 6 months, utilizing hospital records and follow-up interviews to link PIM exposure to hospitalization events. Quality assurance measures included addressing missing data and implementing bias mitigation strategies. Covariates included sociodemographics, clinical characteristics, and medication-related factors to adjust for confounding effects in the analysis.

Data quality assurance, reliability, and bias mitigation

The study’s data collection tool was validated and refined through a pretest. Supervisors and data collectors underwent extensive training on study goals and ethical standards. Daily reviews by the principal investigator and corrections by supervisors ensured data integrity, supported by continuous supervision and refresher courses. To enhance reliability, validated instruments and the 2023 AGS Beers Criteria were employed, with strong inter-rater reliability (Cohen’s kappa = 0.82). Missing data were minimized through verification and follow-up, with fewer than 5% cases handled via multiple imputations. Bias mitigation was achieved through proportional participant allocation, standardized procedures, double data entry, and statistical adjustments, reinforcing the robustness of findings linking inappropriate medication use to hospitalization.

Data processing and analysis

Descriptive statistics were calculated for categorical variables, summarizing frequencies and percentages. Bivariate analysis identified factors linked to PIMs and hospitalization, with variables having p < 0.25 included in multivariable logistic regression models. The dependent variable was PIM use, and adjusted odds ratios (AORs) with 95% confidence intervals (CIs) were reported. Covariates included sociodemographic factors, comorbidity burden, medication-related factors, and behavioral factors, evaluated for model fit using the Hosmer–Lemeshow test and multicollinearity via variance inflation factors. Subgroup analyses explored the link between PIM use and hospitalization. All covariates in the models are detailed in Tables S1 and S2. Analyses were conducted using Stata version 17.

Results

Study enrollment and overview

In a study conducted between May and November 2024, 1700 older adults aged 65 and above were enrolled across five specialized hospitals in the Amhara region, Ethiopia. Participants were tracked from admission to discharge or readmission, achieving over 98% data completeness with no exclusions due to missing key variables.

Sociodemographic and behavioral characteristics

The study population had a mean age of 69.3 years, with 59% males. Most participants lived in rural areas (76%), and 33% had no formal education. Marital status showed that 63% were married and 20% divorced. Health financing was mainly through community-based insurance (66%), with 29% relying on out-of-pocket payments. Behavioral habits included khat chewing (58%), alcohol consumption (34%), cigarette smoking (20%), and traditional medicine use (22%; Table 1).

Table 1.

Baseline sociodemographic and behavioral characteristics of older adults in comprehensive specialized hospitals in the Amhara region from May to November 2024 (N = 1700).

Characteristics	Categories/units	n (%)/mean ± SD
Age	Years	69.33 ± 8.39
Sex	Male	1002 (58.94)
Residence	Rural	1292 (76)
Education	No formal	561 (33)
	Primary	663 (39)
	Secondary	119 (7)
	Tertiary	357 (21)
Marital status	Married	1071 (63)
	Single	204 (12)
	Divorce	340 (20)
	Widowed	85 (5)
Expected primary payers	Insurance	1122 (66)
	Out of pocket	493 (29)
	Others	85 (5)
Behavioral factors
Smoking	Yes	340 (20)
	No	1360 (80)
Alcoholic drink	Yes	578 (34)
	No	1122 (66)
Khat chewing	Yes	986 (58)
	No	714 (42)
Traditional medicine intake	Yes	368 (21.65)
	No	1332 (78.35)

Clinical features and comorbidities

A total of 170 participants (10%) reported a family history of chronic illness, with a mean duration of chronic disease of 28.4 ± 33.5 months. Nearly half (49%) had three concurrent chronic conditions, and 30% had four or more, indicating significant multimorbidity. The mean Charlson Comorbidity Index (CCI) score was 3.06 ± 1.77, with 36% of participants having a CCI ⩾4. Predominant cardiovascular disorders included coronary heart disease (43%), heart failure (34.5%), and hypertension (20%). Other common comorbidities were chronic kidney disease (18%), anemia (21%), and diabetes mellitus (13%; Table 2).

Table 2.

Baseline clinical characteristics and comorbidity profile characteristics admitted older adults among comprehensive specialized hospitals in the Amhara region from May to November 2024 (N = 1700).

Features	Categories/units	n (%)/Mean ± SD
Family history of chronic conditions	Yes	170 (10)
Duration of chronic conditions	Months	28.37 ± 33.46
Number of comorbidities	1	17 (1)
	2	340 (20)
	3	833 (49)
	4 and more	511 (30)
CCI	<4	1088 (64)
	⩾4	612 (36)
CCI	Score	3.06 ± 1.77
Key comorbidities
Heart failure	Yes	586 (34.45)
Hypertensive heart disease	Yes	153 (9)
Coronary heart diseases	Yes	731 (43)
Arrhythmia	Yes	238 (14.43)
Kidney disease	Yes	306 (18)
Hypertension	Yes	342 (20.12)
Respiratory tract infection	Yes	306 (19.35)
Diabetes mellitus	Yes	220 (12.94)
Liver diseases	Yes	256 (15.06)
Anemia	Yes	357 (21)
COPD/asthma	Yes	187 (11)
Cardiomyopathy/LVH	Yes	170 (10)
CRVDH	Yes	272 (16)
Pulmonary hypertension	Yes	85 (5)
Degenerative disease	Yes	78 (4.59)
Others		30 (1.76)

CCI, Charlson Comorbidity Index; COPD, chronic obstructive pulmonary disease, CRVDH, Chronic Rheumatic Valvular Heart Disease; LVH, Left Ventricular Hypertrophy.

Prevalence of PIMs

Among all participants, 704 (41.4%) received at least one PIM requiring intervention. The most frequently prescribed PIM classes were benzodiazepines (22.9%), proton pump inhibitors (PPIs; 20.2%), NSAIDs (18.2%), and anticholinergic agents (17.3%).

Regarding renal safety, 198 (28.1%) had kidney function assessments documented, and 162 (23.0%) received appropriate renal-dose adjustments. Medications requiring cautious use included sodium–glucose cotransporter 2 (SGLT2) inhibitors (9.7%) and non-benzodiazepine sedatives (7.4%). Clinically significant drug–drug interactions were observed, including benzodiazepine plus opioid (6.4%), warfarin plus antibiotics (3.7%), and NSAIDs plus corticosteroids (3.0%).

Only 150 (21.3%) patients with PIM exposure received regular structured medication reviews or deprescribing assessments, indicating limited implementation of medication optimization strategies (Table 3). These results demonstrate a substantial burden of PIM use and highlight the need for improved medication safety interventions among older adults in Ethiopia (Table 3).

Table 3.

Prevalence of subtype potentially inappropriate medication use and possible reasons for each subtype on admitted older adults among Comprehensive specialized hospitals in the Amhara region from May to November 2024 (N = 1700).

Characteristics	Patients’ n (%)	N (%)	Reason(s) or notes
Medications to avoid in most older adults (n = 704)
PIMs that require deprescribing or substitution? (N = 1700)	Patients	704 (41.41)	Patients identified with PIMs need action
Does the patient receive regular medication reviews for polypharmacy and PIM reduction? (n = 704)	patients	150 (21.3)	Percentage of patients receiving regular medication reviews
Benzodiazepine	Yes	161 (22.9)	Commonly prescribed but high risk for cognitive impairment, falls, and fractures
Proton pump inhibitors	Patients	142 (20.2)	Often overused beyond 8 weeks, risk of fractures and infections
NSAIDs	Patients	128 (18.2)	Risk of GI bleeding, kidney injury, and heart failure, especially in the elderly
Anticholinergic drugs	Patients	122 (17.3)	Risk of confusion, dry mouth, constipation, urinary retention
Antipsychotics	Patients	72 (10.2)	Increased risk of stroke and mortality, especially in dementia patients
Medications to avoid based on diseases/syndromes (n = 704)
Falls or fractures	Patients	41 (5.8)	Common consequences of sedatives, benzodiazepines, and NSAIDs
Cognitive impairment or delirium	Patients	29 (4.1)	Often linked to benzodiazepines, anticholinergics
Severe constipation or urinary retention	Patients	16 (2.3)	Related to anticholinergic drugs
Kidney function and dosing (n = 704)
Gastrointestinal bleeding	Patients	14 (2.0)	Associated with NSAIDs and corticosteroids
Has the patient’s kidney function	Patients	198 (28.1)	Important for detecting renal impairment
Are any medications that require kidney dose adjustment being prescribed accordingly	Patients	162 (23.0)	Reflects appropriate prescribing based on kidney function
Medications to use with caution (n = 704)
SGLT2 inhibitors	Patients	68 (9.7)	Use caution if history of recurrent urinary/genital infections
Non-benzodiazepine	Patients	52 (7.4)	Similar risk profile as benzodiazepines, often prescribed for insomnia
Drug–drug interactions (n = 704)
Benzodiazepine + opioid	Patients	45(6.4)	High risk of respiratory depression, sedation, overdose
Warfarin + antibiotics	Patients	26 (3.7)	Increased bleeding risk due to an interaction affecting warfarin metabolism
NSAIDs + corticosteroids	Patients	21 (3.0)	Increased risk of GI bleeding and ulcers

NSAID, nonsteroidal anti-inflammatory drug; PIM, potentially inappropriate medication; SGLT2, sodium–glucose cotransporter 2.

Denominator (n = 704) represents all patients identified with at least one PIM requiring action out of the total cohort (N = 1700). Subcategory counts indicate the number of patients within this PIM group exposed to each specific medication or risk category. Percentages reflect the proportion of PIM users affected by each subtype.

Hospitalization characteristics

A total of 1258 participants (75.5%) had prior hospital admissions, averaging 2.38 admissions. The median time to readmission was 78 days, with a 10-day median length of stay. Most admissions (73%) occurred in internal medicine wards, while 25% were in cardiac units. Time since the last admission was ⩽12 months for 29.8%, 13–24 months for 46.4%, and >24 months for 23.8% of participants (Table 4).

Table 4.

Previous hospital admissions and current hospitalization characteristics of admitted older adults among comprehensive specialized hospitals in Amhara region from May to November 2024 (N = 1700).

Characteristics	Categories/units	n (%)/mean ± SD
Previous hospital admission
Any previous admission	Yes	1258 (75.51)
Number of previous admissions	month	2.38 ± 0.612
Time since last admission (months)	0–12	425 (29.76)
	13–24	663 (46.43)
	>24	340 (23.81)
Current hospitalization details
Readmission date (median) (IQR) 78 (38–180)
LOS (days) median 10 (6.5–18) with minimum and maximum (2–35)
Admitting unit (N = 1700)	Cardiac unit	425 (25)
	Internal medicine	1241 (73)
	Others	34 (2)

LOS, length of hospital stay.

Factors associated with PIM use

In multivariable analysis, several factors were significantly associated with PIM use. Older age (⩾80 years) increased the odds of PIM exposure (AOR = 1.58, 95% CI: 1.22–2.04), as did multimorbidity with a CCI ⩾4 (AOR = 2.10, 95% CI: 1.55–2.85). Lifestyle factors such as khat chewing (AOR = 1.41, 95% CI: 1.17–4.24) were also linked with higher PIM use. By contrast, patients receiving regular medication reviews were less likely to be prescribed PIMs (AOR = 0.65, 95% CI: 0.48–0.88). Full regression models with all predictors and interaction terms are presented in Tables S2.

Impact of PIM use on hospitalization

PIM use was strongly associated with hospitalization (AOR = 1.75, 95% CI: 1.30–2.35). Additional predictors of hospitalization included advanced age (⩾80 years; AOR = 1.95, 95% CI: 1.32–2.88), CCI ⩾4 (AOR = 2.10, 95% CI: 1.55–2.85), and benzodiazepine use, particularly when combined with opioids (AOR = 2.35, 95% CI: 1.60–3.40). Importantly, regular medication review remained protective, reducing the risk of hospitalization by approximately 35%. Full regression models with all predictors and interaction terms are presented in Table S2.

Discussion

Overview of findings

This study aimed to determine the prevalence of PIM use and examine its associated factors with PIM use among this vulnerable population with chronic conditions in the selected region, northwest Ethiopia. The findings revealed a substantial burden of PIMs, with a significant association between PIMs and an increased risk of hospitalization. Furthermore, the study identified several other predictors, including behavioral factors, comorbidities, and certain medication classes, while the protective effect of regular medication reviews.

Prevalence of PIMs use and its significance

The prevalence of PIMs among older adults with chronic conditions in the Amhara region is 41.4%, indicating a significant issue with inappropriate prescribing. This figure is consistent with global data, where PIM rates among the elderly range from 20% to 50%. Africa records the highest PIM prevalence at 47.0%, suggesting that the challenges around PIM use are widespread, not limited to developed nations. The risks associated with PIM use are considerable, with over 40% of older adults in this study exposed to at least one PIM, leading to increased ADEs, hospitalizations, and mortality. Common PIM classes, such as benzodiazepines and NSAIDs, contribute to serious complications, highlighting the need for effective medication safety strategies. To combat this, strategies like structured medication reviews and adherence to evidence-based guidelines are essential to reduce PIM exposure and improve patient safety in low-resource settings.^1,8,13

This study thus provides a critical benchmark, emphasizing that PIM use may be under-recognized in many LMICs due to limited local data.¹⁵ The results are particularly relevant for Ethiopia, which is undergoing a demographic transition toward an aging population and an increasing burden of chronic disease. This shift coincides with healthcare system constraints, including limited access to specialized geriatric services, inadequate medication review processes, and widespread use of traditional medicines, all of which exacerbate the risk and impact of PIM exposure.

Given the broader challenges faced by Sub-Saharan Africa in delivering effective health services and medication management for older adults, the documented 41.4% prevalence underscores the urgent need for targeted interventions. These data can inform the development of clinical guidelines, healthcare provider training, and resource allocation strategies aimed at optimizing medication safety and improving health outcomes among older adults, thereby transforming PIM use from a largely unrecognized problem into a priority for evidence-based policy and practice.

PIM use is a major driver of hospitalizations

A secondary finding of this study was the strong association between PIM use and hospitalizations. Patients exposed to PIMs faced an approximately fourfold increased risk of being hospitalized (AOR = 3.7, 95% CI: 2.254–4.95). This implies powerful predictors of hospital admission in this vulnerable population. PIMs contribute to hospitalizations primarily through ADEs and medication-related complications. This is consistent with findings from other studies globally, which highlight ADEs as significant causes of hospitalization.^16,17 The study highlights the negative effects of PPIs, including cognitive impairment, falls, fractures, infections, kidney injury, heart failure, and stroke risk. Factors contributing to hospitalization include chewing khat, degenerative diseases, chronic illness, medication use, and clinical symptoms.

The mediating role of inappropriate medication use

The study highlights the mediating role of inappropriate medication use in hospitalization risk among older adults with chronic illnesses. Khat chewing was associated with nearly double the risk of using PIMs (AOR = 1.95), potentially due to its stimulant effects causing cardiovascular and neuropsychiatric complications that lead to drug interactions, exacerbation of underlying conditions, and altered medication adherence. Previous studies have shown that khat use increases blood pressure and heart rate, which are linked to myocardial infarction and psychosis, important contributors to PIMs.^18,19 This local cultural context necessitates tailored public health interventions addressing khat use to effectively reduce PIM exposure.

Cor pulmonale, indicative of advanced cardiopulmonary disease, was also a significant predictor of PIM use among older adults with chronic illnesses (AOR = 2.28). This association may be explained by the complex pharmacological management required for cor pulmonale, which contributes to polypharmacy and increases the risk of ADEs.^20,21 In addition, the high hospitalization risk among these patients aligns with epidemiologic data showing that older adults with chronic respiratory and cardiac conditions have substantially higher rates of potentially preventable hospitalizations compared to younger populations.^22,23

Degenerative diseases were identified as another significant clinical factor associated with increased PIM risk (AOR = 3.20). These chronic and progressive conditions commonly coexist with multimorbidity and functional decline in older adults. Patients with degenerative disorders often require complex pharmaceutical regimens, raising the risk of polypharmacy and exposure to PIMs.²⁴ Furthermore, PIMs were extensively used to treat chronic symptoms commonly linked with degenerative illnesses. This finding is consistent with prior research indicating that PIM use partially mediates the connection between degenerative diseases and hospitalization. Persons with neurodegenerative disorders are particularly vulnerable to PIM-related effects due to reduced cognition, frailty, and poor medication adherence, especially in settings lacking organized deprescribing methods.^25–27

The CCI score greater than 4 was identified as the strongest clinical predictor of PIM use (AOR = 4.50), indicating that patients with a high comorbidity burden were 4.5 times more likely to be hospitalized than those with lower scores. This significant association is consistent with extensive literature highlighting multimorbidity as a major predictor of hospital admissions and mortality.²⁸ The clinical complexity of individuals with numerous chronic illnesses often necessitates complicated pharmaceutical treatments, increasing the likelihood of polypharmacy and PIM exposure.^28,29 The inclusion of PIMs in multimorbid patients is especially concerning because these medications may worsen existing diseases, cause ADEs, and contribute to avoidable hospitalizations. Evidence from low-resource settings confirms that the combination of high comorbidity burden and inappropriate prescribing dramatically increases healthcare use and worsens clinical outcomes.^2,10

Prolonged duration of chronic illness, defined as a median duration exceeding 2 years, was associated with a threefold increase in hospitalization risk (AOR = 3.07). This aligns with findings that chronic disease progression leads to increased frailty and a higher incidence of acute health crises. Chronic illnesses elevate hospitalization risk due to organ damage and declining physiological reserve.^30,31 Moreover, longer illness duration often requires complex and sustained pharmacological regimens, increasing the likelihood of polypharmacy and PIM exposure. Prior research supports the strong correlation between disease duration, PIM use, ADEs, and subsequent hospitalization.^32,33

Medication-related factors also played a significant role. Benzodiazepine use increased the odds of hospitalization by 80% (AOR = 1.80), consistent with studies linking benzodiazepines to falls, cognitive impairment, and delirium in older adults.^34,35 These adverse effects contribute substantially to morbidity and healthcare utilization in this population. In addition, concurrent use of benzodiazepines and opioids was associated with a fourfold increase in hospitalization risk (AOR = 4.02), likely due to synergistic central nervous system depressant effects that impair respiratory drive and increase sedation.³⁶ This interaction is recognized globally as a major contributor to medication-related hospitalizations and mortality, prompting public health warnings and regulatory guidelines.³⁷

This study also reported the importance and the protective effect of regular medication reviews. Patients who underwent consistent medication evaluation had significantly lower risks of hospitalization, underscoring the critical role of systematic medication assessment in improving outcomes among older adults with chronic illnesses. This highlights a feasible and impactful intervention point for healthcare providers and policymakers in Ethiopia and similar low-resource settings. Implementing structured medication review protocols could help identify and deprescribe PIMs, reduce ADEs, and optimize pharmacotherapy management, ultimately lowering preventable hospital admissions. Given the high prevalence of PIM use observed, integrating regular medication reviews into routine clinical practice should be prioritized as a key strategy to enhance patient safety and reduce healthcare burdens.^38–40

Cognitive impairment and delirium were also predictors of hospitalization, increasing the likelihood by approximately 78% (AOR = 1.78). These conditions are common in older adults and are often caused or exacerbated by PIMs such as benzodiazepines, anticholinergic drugs, and sedative-hypnotics.^41–43 PIMs contribute to neurocognitive decline by intensifying central nervous system depression, increasing anticholinergic burden, and disrupting neurotransmitter function, thereby increasing susceptibility to delirium and cognitive impairment. This pharmaceutical impact complicates clinical progression, raises the risk of falls, infections, and other acute events, and ultimately increases hospitalization likelihood.^44,45 Longitudinal studies have shown that deprescribing PIMs reduces delirium and cognitive impairment severity, leading to shorter hospital stays and improved outcomes.^46,47

Finally, gastrointestinal bleeding was associated with a 58% increased risk of hospitalization. This risk is exacerbated by PIMs, particularly NSAIDs and corticosteroids, which compromise the gastrointestinal mucosa and increase bleeding risk.^48,49 PIMs may mediate the relationship between chronic conditions and hospitalization by exacerbating gastrointestinal vulnerability and precipitating ADEs.

PIM use mediates the comorbidities-related hospitalization risk

The current study identified that PIM use mediates the relationship between hospitalization risk and comorbidity burden, as indicated by a CCI score greater than 4. While both PIM use (AOR = 1.75) and elevated comorbidity burden (AOR = 2.10) independently associated with hospitalization risk, a significant interaction effect (AOR = 2.80) suggests a synergistic relationship. This indicates that PIM use amplifies the impact of multimorbidity on the likelihood of hospitalization. The finding reflects the clinical complexity associated with multimorbidity, where patients with multiple chronic conditions often require polypharmacy, inherently increasing the risk of inappropriate prescribing.⁵⁰ These results emphasize that PIM exposure is a critical factor that translates a high comorbidity burden into increased hospitalization rates.⁵¹ Although previous research has independently linked both multimorbidity and PIM use to hospital admissions, fewer studies have explicitly explored PIMs as a mediating factor in this relationship.

The study also found that the use of PIMs in combination with khat chewing increased the risk of hospitalization (AOR = 1.41), with PIM use further elevating this risk (AOR = 1.65). Since khat is a commonly used stimulant in the study area and is associated with gastrointestinal and cardiovascular complications, this interaction likely exacerbates hospitalization risk. Substance use behaviors such as khat chewing may worsen side effects related to inappropriate medications, underscoring the importance of healthcare providers considering lifestyle factors when reviewing prescriptions and counseling patients.⁵²

Furthermore, the study showed that PIM use alone increased the risk of hospitalization (AOR = 2.30), and advanced age (⩾80 years) was independently associated with a higher hospitalization risk (AOR = 1.95). This aligns with extensive literature demonstrating that older adults are more vulnerable to adverse medication effects, polypharmacy, and hospital admissions.^53,54 The findings highlight the critical need to avoid PIMs in the elderly, given age-related physiological changes that affect drug metabolism and increase susceptibility to medication-related harm.

Lastly, use of benzodiazepines alone and in combination with opioids significantly increases hospitalization risks (AOR = 1.88 and AOR = 2.35, respectively). Risks are heightened with PIMs, leading to complications like overdose, falls, and cognitive decline. The findings highlight the need for cautious prescribing, close monitoring, and adherence to guidelines to reduce medication-related hospitalizations and recommend careful dose management and patient selection.^55–58

Policy relevance and future research

The study highlights critical implications for clinical practice and health policy in Ethiopia and similar low-resource settings due to a high prevalence of PIM use among older adults. It emphasizes the need for systematic interventions, including structured medication reviews, adherence to national or institutional guidelines like the AGS Beers Criteria, and targeted training for healthcare providers. Future research should assess the effectiveness of these interventions in reducing PIM exposure and adverse outcomes while exploring barriers to guideline adherence and focusing on patient-centered outcomes such as quality of life.

Limitation

This study has notable strengths, including its large multicenter prospective cohort design that enhances generalizability across various healthcare settings in the Amhara region. It utilized the 2023 AGS Beers Criteria to identify PIMs, with data collected by trained pharmacists under rigorous quality control to ensure high integrity. However, limitations include the observational design, which restricts causal inferences, and the potential for residual confounding. The study’s focus on specialized hospitals may not adequately represent primary care or rural populations, and the 6-month follow-up may not capture long-term effects from PIM exposure. In addition, incomplete renal function data could impact dose adjustment accuracy, and traditional medicine interactions were not comprehensively evaluated. Despite these limitations, the study highlights the prevalence of PIMs and their association with hospitalization, emphasizing the need for systematic medication reviews and enhanced education to mitigate PIM-related harm among older adults.

Conclusion

This multicenter prospective cohort study in the Amhara region of Ethiopia found a 41.4% prevalence of PIM use among older adults with chronic diseases. PIM exposure correlated with increased hospitalization risk, notably in individuals with multimorbidity, advanced age, and those using benzodiazepines in combination with opioids. Behavioral factors, including khat chewing, also raised PIM use and hospitalization risk. Regular medication reviews were identified as key in reducing both PIM exposure and hospitalization risk, highlighting their importance in this demographic. Clinical implications suggest healthcare providers prioritize routine medication evaluations and address local practices to enhance medication safety. Future research should focus on sustainable medication review integration in low-resource settings and examine the long-term outcomes of PIM reduction and the role of traditional medicine in medication safety.

Supplemental Material

sj-docx-1-taw-10.1177_20420986251410989 – Supplemental material for Prevalence of potentially inappropriate medication use and its association with hospitalization among older adults with chronic disease in the Amhara region, Ethiopia: a multicenter prospective cohort study

Supplemental material, sj-docx-1-taw-10.1177_20420986251410989 for Prevalence of potentially inappropriate medication use and its association with hospitalization among older adults with chronic disease in the Amhara region, Ethiopia: a multicenter prospective cohort study by Getachew Yitayew Tarekegn, Fisseha Nigussie Dagnew, Samuel Agegnew Wondm, Tilaye Arega Moges, Zufan Alamrie Asmare, Teklie Mengie Ayele, Sisay Sitotaw Anberbr, Dawit Haile Zeben, Tigabu Eskeziya Zerihun, Abel Temeche Kassaw, Desalegn Addis Mussie, Teferi Bihonegn Melese and Samuel Berihun Dagnew in Therapeutic Advances in Drug Safety

Supplemental Material

sj-docx-2-taw-10.1177_20420986251410989 – Supplemental material for Prevalence of potentially inappropriate medication use and its association with hospitalization among older adults with chronic disease in the Amhara region, Ethiopia: a multicenter prospective cohort study

Supplemental material, sj-docx-2-taw-10.1177_20420986251410989 for Prevalence of potentially inappropriate medication use and its association with hospitalization among older adults with chronic disease in the Amhara region, Ethiopia: a multicenter prospective cohort study by Getachew Yitayew Tarekegn, Fisseha Nigussie Dagnew, Samuel Agegnew Wondm, Tilaye Arega Moges, Zufan Alamrie Asmare, Teklie Mengie Ayele, Sisay Sitotaw Anberbr, Dawit Haile Zeben, Tigabu Eskeziya Zerihun, Abel Temeche Kassaw, Desalegn Addis Mussie, Teferi Bihonegn Melese and Samuel Berihun Dagnew in Therapeutic Advances in Drug Safety

Footnotes

Acknowledgements

The authors thank the staff of Debre Tabor University and all study participants for their support and cooperation.

Declarations

ORCID iDs

Getachew Yitayew Tarekegn

Samuel Agegnew Wondm

Tilaye Arega Moges

Samuel Berihun Dagnew

Supplemental material

Supplemental material for this article is available online.

References

By the 2023 American Geriatrics Society Beers Criteria^® Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria^® for potentially inappropriate medication use in older adults. J Am Geriatr Soc 2023; 71: 2052–2081.

Tian

Chen

Zeng

, et al. Prevalence of use of potentially inappropriate medications among older adults worldwide: a systematic review and meta-analysis. JAMA Netw Open 2023; 6: e2326910.

Borodo

Jatau

Mohammed

, et al. The burden of polypharmacy and potentially inappropriate medication in Nigeria: a clarion call for deprescribing practice. Bull Natl Res Cent 2022; 46: 178.

Tesfaye

Tessema

Yizengaw

, et al. Potentially inappropriate medication use among older adult patients on follow-up at the chronic care clinic of a specialized teaching hospital in Ethiopia: a cross-sectional study. BMC Geriatr 2021; 21: 530.

Nigussie

Demeke

Potentially inappropriate medications use and associated factors among older patients on follow-up at the chronic care clinic of Hiwot Fana Comprehensive Specialized Hospital in Eastern Ethiopia. Curr Ther Res Clin Exp 2024; 100: 100730.

Paulamäki

Jyrkkä

Hyttinen

, et al. Regional variation of potentially inappropriate medication use and associated factors among older adults: a nationwide register study. Res Social Adm Pharm 2023; 19: 1372–1379.

Randles

O’Mahony

Gallagher

PF.

Frailty and potentially inappropriate prescribing in older people with polypharmacy: a bi-directional relationship?

Drugs Aging 2022; 39: 597–606.

Motter

Fritzen

Hilmer

, et al. Potentially inappropriate medication in the elderly: a systematic review of validated explicit criteria. Eur J Clin Pharmacol 2018; 74: 679–700.

Bhagavathula

Gebreyohannes

Fialova

Prevalence of polypharmacy and risks of potentially inappropriate medication use in the older population in a developing country : a systematic review and meta-analysis. Gerontology 2022; 68: 136–145.

10.

kitaw

Haile

RN.

Prevalence of polypharmacy among older adults in Ethiopia: a systematic review and meta-analysis. Sci Rep 2023; 13: 17641.

11.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61: 344–349.

12.

Bhagavathula

Gebreyohannes

Fialova

Prevalence of polypharmacy and risks of potentially inappropriate medication use in the older population in a developing country: a systematic review and meta-analysis. Gerontology 2022; 68: 136–145.

13.

O’Mahony

Cherubini

Guiteras

, et al. STOPP/START criteria for potentially inappropriate prescribing in older people: version 3. Eur Geriatr Med 2023; 14: 625–632.

14.

Tesfaye

Bosho

Dissassa

, et al. Potentially inappropriate medicine use and predicting risk factors in hospitalized older adult patients: findings of a prospective observational study from Ethiopia. J Pharm Policy Pract 2023; 16: 164.

15.

Tian

Chen

Prevalence of polypharmacy and potentially inappropriate medications use in elderly chinese patients: a systematic review and meta-analysis. Front Pharmacol 2022; 13: 1–8.

16.

Sahilu

Getachew

Melaku

, et al. Adverse drug events and contributing factors among hospitalized adult patients at Jimma Medical Center, Southwest Ethiopia: a prospective observational study. Curr Ther Res Clin Exp 2020; 93: 100611.

17.

Budnitz

Lovegrove

Shehab

, et al. Emergency hospitalizations for adverse drug events in older Americans. N Engl J Med 2025; 365: 2002–2012.

18.

Naeem

Situt

Shetty

, et al. Understanding the intricacies of khat-associated cardiovascular disease. J Clin Med 2025; 14: 1–11.

19.

Mega

Dabe

NE.

Khat (Catha edulis) as a risk factor for cardiovascular disorders: systematic review and meta-analysis. Open Cardiovasc Med J 2017; 11: 146–155.

20.

Sheikh-Taha

Asmar

. Polypharmacy and severe potential drug-drug interactions among older adults with cardiovascular disease in the United States. BMC Geriatr 2021; 21: 233.

21.

Gale

Hurst

Hawkins

, et al. Identification and management of cardiopulmonary risk in patients with chronic obstructive pulmonary disease: a multidisciplinary consensus and modified Delphi study. Eur J Prev Cardiol 2025; 32: zwaf119.

22.

Costanzo

Di Castelnuovo

Panzera

, et al. Polypharmacy in older adults: the hazard of hospitalization and mortality is mediated by potentially inappropriate prescriptions, findings from the Moli-sani Study. Int J Public Health 2024; 69: 1–11.

23.

Boateng

Pascual

Grassby

The impact of polypharmacy on health outcomes in the elderly. Authorea, 2022, pp. 1–10.

24.

Aubert

Kabeto

Kumar

, et al. Multimorbidity and long-term disability and physical functioning decline in middle-aged and older Americans: an observational study. BMC Geriatr 2022; 22: 910.

25.

Sheikh Rezaei

Šinkovec

Schöberl

, et al. Utilization of potentially inappropriate medication and risk of adverse drug events among older adults with chronic renal insufficiency: a population-wide cohort study. BMC Geriatr 2021; 21: 117.

26.

Shaw

Briscoe

Nunns

, et al. What is the quantity, quality and type of systematic review evidence available to inform the optimal prescribing of statins and antihypertensives? A systematic umbrella review and evidence and gap map. BMJ Open 2024; 14: e072502.

27.

Kim

Lee

Jang

, et al. Polypharmacy and elevated risk of severe adverse events in older adults based on the Korea Institute of Drug Safety and Risk Management-Korea adverse event reporting system database. J Korean Med Sci 2024; 39: e205.

28.

Hong

Lai

FTT

Wang

, et al. Age-specific multimorbidity patterns and burden on all-cause mortality and public direct medical expenditure: a retrospective cohort study. J Epidemiol Glob Health 2024; 14: 1077–1088.

29.

Tian

Chen

Prevalence of polypharmacy and potentially inappropriate medications use in elderly Chinese patients: a systematic review and meta-analysis. Front Pharmacol 2022; 13: 862561.

30.

Salari

Najafi

Rasoulpoor

, et al. The global prevalence and associated factors of loneliness in older adults: a systematic review and meta-analysis. Humanit Soc Sci Commun 2025; 12: 985.

31.

Pengpid

Peltzer

Anantanasuwong

Bidirectional association between functional disability and multimorbidity among middle-aged and older adults in Thailand. Front public Heal 2022; 10: 1055699.

32.

Doumat

Daher

Itani

, et al. The effect of polypharmacy on healthcare services utilization in older adults with comorbidities: a retrospective cohort study. BMC Prim Care 2023; 24: 120.

33.

Hung

Kim

Pavon

JM.

Deprescribing in older adults with polypharmacy. BMJ 2024; 385: e074892.

34.

Hwang

Son

Kim

, et al. Association of zolpidem with increased mortality in patients with brain cancer: a retrospective cohort study based on the National Health Insurance Service database. J Clin Neurol 2022; 18: 65–70.

35.

Hofe

vom Stricker

Vernooij

, et al. Benzodiazepine use in relation to long-term dementia risk and imaging markers of neurodegeneration: a population-based study. BMC Med 2024; 22: 266.

36.

Sun

Dixit

Humphreys

, et al. Association between concurrent use of prescription opioids and benzodiazepines and overdose: retrospective analysis. BMJ 2017; 356: j760.

37.

Asfaw

Boden

LI.

Impact of workplace injury on opioid dependence, abuse, illicit use and overdose: a 36-month retrospective study of insurance claims. Occup Environ Med 2020; 77: 648–653.

38.

Degen

Chen

L-J

Schöttker

Medication reviews in hospitalised patients for reduced hospital readmission and mortality. Systematic review, meta-analysis and meta-regression of RCTs. Ageing Res Rev 2025; 104: 102661.

39.

Grischott

Rachamin

Senn

, et al. Medication review and enhanced information transfer at discharge of older patients with polypharmacy: a cluster-randomized controlled trial in Swiss hospitals. J Gen Intern Med 2023; 38: 610–618.

40.

Bülow

Clausen

Lundh

, et al. Medication review in hospitalised patients to reduce morbidity and mortality. Cochrane Database Syst Rev 2023; 1: CD008986.

41.

Joshi

Thomas

Braff

, et al. Anticholinergic medication burden-associated cognitive impairment in schizophrenia. Am J Psychiatry 2021; 178: 838–847.

42.

Marcantonio

ER.

Delirium in hospitalized older adults. N Engl J Med 2025; 377: 1456–1466.

43.

Reallon

Gervais

Moutet

, et al. Impact of cumulative exposure to anticholinergic and sedative drugs on cognition in older adults: a memory clinic cohort study. Alzheimers Res Ther 2024; 16: 163.

44.

Evelo

Licher

Stricker

, et al. Temporal trends of anticholinergic drug exposure among older adults: a 25-year population-based study. Drug Saf 2025; 48: 1141–1147.

45.

Bhatkhande

Choudhry

Mahesri

, et al. Disentangling drug contributions: anticholinergic burden in older adults linked to individual medications: a cross-sectional population-based study. BMC Geriatr 2024; 24: 44.

46.

Ibrahim

Cox

Stevenson

, et al. A systematic review of the evidence for deprescribing interventions among older people living with frailty. BMC Geriatr 2021; 21: 258.

47.

Reeve

Farrell

Thompson

, et al. Evidence-based clinical practice guideline for deprescribing cholinesterase inhibitors and memantine: recommendations. University of Sydney, http://sydney.edu.au/medicine/cdpc/resources/deprescribing-guidelines.php (2018, accessed 2 January 2025).

48.

Liu

Yang

, et al. Risk factors for gastrointestinal complications during glucocorticoid therapy in internal medicine inpatients: a real-world retrospective analysis. BMC Pharmacol Toxicol 2025; 26: 37.

49.

Spence

Grosser

FitzGerald

GA.

Acetaminophen, nonsteroidal anti-inflammatory drugs, and hypertension. Hypertension 2022; 79: 1922–1926.

50.

Boateng

Pascual

Grassby

, et al. The impact of polypharmacy on health outcomes in the aged: a retrospective cohort study. PLoS One 2025; 20: e0317907.

51.

Pazan

Wehling

Polypharmacy in older adults: a narrative review of definitions, epidemiology and consequences. Eur Geriatr Med 2021; 12: 443–452.

52.

Alshoabi

Hamid

Gameraddin

, et al. Risks of khat chewing on the cardiovascular, nervous, gastrointestinal, and genitourinary systems: a narrative review. J Family Med Prim care 2022; 11: 32–36.

53.

Ngcobo

NN.

Influence of ageing on the pharmacodynamics and pharmacokinetics of chronically administered medicines in geriatric patients: a review. Clin Pharmacokinet 2025; 64: 335–367.

54.

Ruiz

DiCristina

Absorption to excretion: the aging body’s take on drugs—a review of pharmacokinetic changes and their impact on medication management. Curr Pharmacol Rep 2025; 11: 42.

55.

Maust

Bohnert

ASB

Strominger

, et al. Prescription characteristics associated with fall-related injury risk among older adults prescribed benzodiazepines: a cohort study. BMC Geriatr 2022; 22: 824.

56.

Salkar

Ramachandran

Bentley

, et al. Do formulation and dose of long-term opioid therapy contribute to risk of adverse events among older adults? J Gen Intern Med 2022; 37: 367–374.

57.

Wang

GHM

Hincapie-Castillo

Gellad

, et al. Association between opioid-benzodiazepine trajectories and injurious fall risk among US Medicare Beneficiaries. J Clin Med 2024; 13: 3376.

58.

Wang

GH-M

Goodin

Reise

, et al. Longitudinal patterns of antidepressant and benzodiazepine use associated with injurious falls in older adults with depression: a retrospective cohort study. BMC Med 2025; 23: 487.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.02 MB