Prevalence and predictors of potentially inappropriate medication use among older adults with heart failure: a 2023 beers criteria–based evaluation Authors: Marwan Sheikh-Taha Published: 2026 ABSTRACT BackgroundOlder adults with heart failure (HF) are particularly vulnerable to polypharmacy and potentially inappropriate medication (PIM) use due to multimorbidity, age-related physiological changes, and complex therapeutic regimens. The 2023 American Geriatrics Society (AGS) Beers Criteria provide updated guidance to identify medications that may increase the risk of adverse outcomes in older adults and can be applied to patients with HF. This study aimed to determine the prevalence and patterns of PIM use among older adults with HF and to identify clinical factors associated with increased PIM burden using the 2023 Beers Criteria.MethodsThis retrospective cross-sectional study included 524 adults aged ≥ 65 years admitted with HF to two hospitals within the Huntsville Hospital Health System between January 2020 and May 2025. Home medications at admission were evaluated according to the 2023 AGS Beers Criteria. PIMs were categorized as generally inappropriate, disease-specific, use with caution, high-risk drug–drug interactions, or medications requiring avoidance or dose adjustment based on renal function. Multivariable negative binomial regression was used to identify predictors of PIM burden defined as the total number of PIMs per patient.ResultsOverall, 472 patients (90.1%) were prescribed at least one PIM, and 26.2% of all home medications met Beers Criteria. Among 562 PIMs classified as potentially inappropriate, gastrointestinal (33.8%), cardiovascular (22.6%), and central nervous system agents (17.1%) were the most frequent. Clinically significant drug–drug interactions (n = 165) and renal function–related PIMs (n = 118) were also common. In multivariable analysis, higher PIM burden was independently associated with chronic kidney disease (IRR 1.38, 95% CI 1.20–1.59), depression (IRR 1.29, 95% CI 1.05–1.58), benign prostatic hyperplasia (IRR 1.26, 95% CI 1.03–1.53), and increasing number of comorbidities (IRR 1.04 per additional condition, 95% CI 1.00–1.07).ConclusionsPIM use is highly prevalent among older adults with HF. Targeted medication review—particularly in patients with renal impairment or multiple comorbidities—followed by deprescribing or safer substitutions, represents a key quality-of-care intervention to reduce preventable medication-related harm in this vulnerable population. Background The global trend of population aging is accompanied by a rising prevalence of multimorbidity, affecting approximately 62% of adults aged 65 years and older [1]. Managing multiple chronic conditions often necessitates the use of several medications, making polypharmacy a hallmark of geriatric care. While necessary in many cases, polypharmacy increases the risk of adverse outcomes, including drug–drug interactions (DDIs), adverse drug reactions, falls, cognitive decline, and higher healthcare utilization such as hospitalizations and mortality [2–4]. Adverse drug events account for nearly one in five unplanned hospitalizations among older adults, underscoring the importance of regular medication review and deprescribing to reduce preventable harm [5, 6]. Age-related physiological changes, such as declines in renal and hepatic function and alterations in body composition, further modify drug pharmacokinetics and pharmacodynamics, heightening susceptibility to medication-related complications. Consequently, older adults require individualized dosing strategies and close clinical monitoring to balance therapeutic benefits with safety. To guide safer prescribing in this population, the American Geriatrics Society (AGS) developed the Beers Criteria, an evidence-based tool identifying potentially inappropriate medications (PIMs) in older adults. The 2023 update of the Beers Criteria classifies medications into categories including those that should generally be avoided in older adults, those inappropriate in specific diseases or syndromes, medications to be used with caution, combinations associated with high-risk DDIs, and medications requiring dose adjustment or avoidance based on renal function [7]. Collectively, these categories provide a framework for optimizing pharmacotherapy and minimizing preventable adverse events. Heart failure (HF) is among the most prevalent and therapeutically complex chronic conditions in older adults. These patients frequently receive multiple medications to manage HF and its associated comorbidities, rendering them especially vulnerable to polypharmacy and PIM use [8, 9]. Moreover, common features of HF, such as renal impairment, hemodynamic instability, and frailty, exacerbate the risk of drug-related harm. Despite the availability of evidence-based prescribing guidelines, inappropriate medication use remains common in this high-risk group, contributing to increased rates of hospitalization, functional decline, and mortality [10–12]. Despite growing literature on PIM use in older adults, data specifically evaluating PIM burden and predictors in hospitalized older patients with HF using the updated 2023 Beers Criteria remain limited. Accordingly, the present study aimed to evaluate the prevalence and patterns of PIM use among older adults with HF and to identify factors associated with the number of PIMs prescribed in this population. Understanding the extent and determinants of PIM use can inform targeted interventions, support deprescribing initiatives, and ultimately improve patient safety and clinical outcomes. Methods This retrospective cross-sectional study included older adults admitted with HF to two hospitals within the Huntsville Hospital Health System (Alabama, USA). While both institutions operate within a shared system, there may be some differences in patient case-mix and prescribing practices. Data were extracted manually from electronic medical records (EMRs) by trained study personnel using a standardized collection form to ensure accuracy and completeness for patients admitted between January 2020 and May 2025. A random sample of 524 patients was selected from the eligible population using a computer-generated random number sequence to ensure a representative sample and minimize selection bias. Eligible participants were adults aged 65 years and older with a documented diagnosis of HF. HF was identified using a predefined approach based on EMR documentation, including clinician diagnosis in admission notes or problem lists. ICD-10 codes for HF were used as supportive evidence when available but were not required, as reliance on codes alone may lead to misclassification or undercapture. To enhance diagnostic consistency, cases were included only when HF was clearly documented by the treating clinician in the medical record. Information on HF phenotype (HFrEF vs HFpEF) was not consistently available and was therefore not included in the analysis. Patients with incomplete medication records or missing key clinical data were excluded from the analysis, including those lacking a medication list at admission, documented comorbidities, or the serum creatinine level required for Beers Criteria assessment. For patients with multiple hospitalizations during the study period, only the first eligible admission was included to avoid duplication. Home medications at the time of hospital admission were reviewed and evaluated against the 2023 American Geriatrics Society Beers Criteria to identify PIMs. Home medications were defined as those documented in the EMR at admission as part of routine medication reconciliation, based on patient or caregiver report and verified by healthcare providers when possible. This included regularly scheduled, as-needed, and documented over-the-counter medications; dietary supplements were not included. For duration-dependent criteria (e.g., proton pump inhibitors (PPIs) > 8 weeks), duration was assessed when available; otherwise, these criteria were not applied. Disease- and syndrome-specific criteria were identified using documented diagnoses in the EMR (problem lists and clinician documentation). Medications could be classified into more than one Beers category; therefore, counts across categories are not mutually exclusive. Each medication was categorized according to the Beers framework into one or more of the following groups: (i) medications generally considered inappropriate in older adults; (ii) medications inappropriate in the presence of certain diseases or syndromes; (iii) medications that should be used with caution; (iv) DDIs that should be avoided; and (v) medications requiring renal dose adjustment or avoidance. Demographic and clinical data collected included age, sex, body weight, height, serum creatinine, and comorbidities. The number of home medications was recorded for each participant, and renal function was assessed using the Cockcroft–Gault equation, as recommended in drug dosing guidelines and the Beers Criteria for medication dose adjustment in older adults. Alternative equations, such as CKD-EPI, may classify renal function differently and could lead to variation in categorization. The study was reviewed by the Institutional Review Board of Huntsville Hospital and determined to be exempt on April 24, 2025. The requirement for informed consent was waived due to the retrospective design and use of de-identified data. Descriptive statistics were used to summarize baseline sociodemographic and clinical characteristics and patterns of PIM use. Continuous variables were reported as means with standard deviations, and categorical variables as frequencies and percentages. The primary outcome was the total number of PIMs per patient (PIM burden), analyzed as a count variable. Analyses were conducted at both the patient level (e.g., prevalence of PIM use) and medication level (e.g., distribution of PIM categories), as appropriate to the research question. Given evidence of overdispersion, negative binomial regression was used to examine factors associated with PIM burden [13], with results expressed as incidence rate ratios (IRRs) and 95% confidence intervals (CIs). Unadjusted bivariate associations between each covariate and PIM burden were assessed using a negative binomial regression model. Multivariable models were constructed a priori based on clinical relevance and existing literature. Covariates included age, sex, total number of comorbidities, and selected chronic conditions relevant to older adults with HF. These included coronary artery disease, cerebrovascular disease, depression, benign prostatic hyperplasia (BPH), chronic obstructive pulmonary disease/asthma, chronic kidney disease (CKD), and diabetes mellitus, which were entered as binary variables. Multicollinearity was assessed using variance inflation factors (VIFs), with values < 5 considered acceptable. As a sensitivity analysis, multivariable logistic regression was performed with PIM exposure dichotomized as present versus absent to assess the robustness of findings using a clinically interpretable binary outcome, and adjusted odds ratios (aORs) with 95% CIs were reported. All analyses were conducted using IBM SPSS Statistics version 25 (IBM Corp., Armonk, NY, USA). Tests were two-sided, and p-values < 0.05 were considered statistically significant. Based on an expected PIM prevalence of 60%, a minimum of 369 participants would be required to estimate prevalence with a 95% confidence level and 5% margin of error; the final sample of 524 patients therefore exceeded this requirement [14]. Results A total of 524 patients met the inclusion criteria. The cohort had a mean age of 76.5 ± 8.5 years, and 53.8% were male. Participants had an average of 6.2 ± 2.3 comorbidities, most commonly hypertension (76.7%), dyslipidemia (63.2%), and coronary artery disease (61.8%). The mean number of home medications per patient was 11.0 ± 3.6. Detailed demographic and clinical characteristics are summarized in Table 1.Table 1Baseline characteristics of the participantsCharacteristicTotal (N = 524)Gender, n (%) Males282 (53.8%) Females242 (46.2%)Age, years76.5 ± 8.5Number of comorbidities6.2 ± 2.3Number of home medications11.0 ± 3.6Comorbidities*, n (%) Hypertension402 (76.7%) Dyslipidemia331 (63.2%) Coronary artery disease324 (61.8%) Atrial fibrillation/flutter288 (55.0%) Diabetes215 (41.0%) Asthma/COPD146 (27.9%) Thyroid disorders118 (22.5%) GERD102 (19.5%) Arthritis95 (18.1%) Cerebrovascular accident/stroke82 (15.6%) DVT/PE / pulmonary embolism77 (14.7%) Cancer63 (12.0%) Depression60 (11.5%) Anxiety57 (10.9%) Peripheral vascular disease45 (8.6%)COPD Chronic obstructive pulmonary disease, DVT Deep venous thrombosis, GERD Gastroesophageal reflux disease, PE Pulmonary embolism*Comorbidities are listed only if present in > 5% of the study population Table 1Baseline characteristics of the participantsCharacteristicTotal (N = 524)Gender, n (%) Males282 (53.8%) Females242 (46.2%)Age, years76.5 ± 8.5Number of comorbidities6.2 ± 2.3Number of home medications11.0 ± 3.6Comorbidities*, n (%) Hypertension402 (76.7%) Dyslipidemia331 (63.2%) Coronary artery disease324 (61.8%) Atrial fibrillation/flutter288 (55.0%) Diabetes215 (41.0%) Asthma/COPD146 (27.9%) Thyroid disorders118 (22.5%) GERD102 (19.5%) Arthritis95 (18.1%) Cerebrovascular accident/stroke82 (15.6%) DVT/PE / pulmonary embolism77 (14.7%) Cancer63 (12.0%) Depression60 (11.5%) Anxiety57 (10.9%) Peripheral vascular disease45 (8.6%)COPD Chronic obstructive pulmonary disease, DVT Deep venous thrombosis, GERD Gastroesophageal reflux disease, PE Pulmonary embolism*Comorbidities are listed only if present in > 5% of the study population Baseline characteristics of the participants COPD Chronic obstructive pulmonary disease, DVT Deep venous thrombosis, GERD Gastroesophageal reflux disease, PE Pulmonary embolism *Comorbidities are listed only if present in > 5% of the study population Overall, 472 patients (90.1%) were prescribed at least one PIM, and among these patients, the number of PIMs per patient ranged from one to 16. Only 52 patients (9.9%) had no PIMs, whereas nearly two-thirds of patients (358/524, 68.3%) were prescribed two or more; the distribution of PIM counts per patient is summarized in Fig. 1. Among 5,759 total home medications, 1,508 unique medication entries (26.2%) met at least one Beers criterion, corresponding to approximately one in four medications.Fig. 1Distribution of patients by number of potentially inappropriate medications Distribution of patients by number of potentially inappropriate medications Medications classified as potentially inappropriate Among medications classified as generally inappropriate (n = 562), gastrointestinal (GI) agents were the most common (33.8%), followed by cardiovascular and antithrombotic agents (22.6%), central nervous system (CNS) medications (17.1%), and pain-related medications (11.0%) (Table 2).Table 2Medications classified as potentially inappropriate according to the 2023 AGS Beers Criteria (n = 562)Medication ClassMedication(s)n%GastrointestinalPPIs (mostly pantoprazole, omeprazole, & esomeprazole)18532.9Metoclopramide20.4Hyoscyamine20.4Dicyclomine10.2Subtotal19033.8Cardiovascular & antithromboticsAspirin for primary prevention5910.5Digoxin (first-line for AF/HF)285.0Rivaroxaban162.8Clonidine91.6α-1 blockers (doxazosin, terazosin)50.9Amiodarone50.9Warfarin30.5Dronedarone20.4Subtotal12722.6Central nervous systemBenzodiazepines396.9Antipsychotics295.2Antidepressants with strong anticholinergic activity (TCAs, paroxetine)142.5Z-drugs122.1Antiparkinson agent with strong anticholinergic activity (benztropine)10.2Barbiturates (primidone)10.2Subtotal9617.1Pain medicationsNSAID285.0Skeletal muscle relaxants (mostly methocarbamol & cyclobenzaprine)346.0Subtotal6211.0First-generation antihistaminesMostly diphenhydramine & meclizine437.7EndocrineSulfonylureas305.3Megestrol50.9Androgens (testosterone)40.7Estrogens (estradiol)40.7Subtotal437.7Anti-infectiveNitrofurantoin10.2Total562100Abbreviations: AF Atrial fibrillation, HF Heart failure, NSAIDs Nonsteroidal anti-inflammatory drugs, PPIs Proton pump inhibitors, TCAs Tricyclic antidepressants Table 2Medications classified as potentially inappropriate according to the 2023 AGS Beers Criteria (n = 562)Medication ClassMedication(s)n%GastrointestinalPPIs (mostly pantoprazole, omeprazole, & esomeprazole)18532.9Metoclopramide20.4Hyoscyamine20.4Dicyclomine10.2Subtotal19033.8Cardiovascular & antithromboticsAspirin for primary prevention5910.5Digoxin (first-line for AF/HF)285.0Rivaroxaban162.8Clonidine91.6α-1 blockers (doxazosin, terazosin)50.9Amiodarone50.9Warfarin30.5Dronedarone20.4Subtotal12722.6Central nervous systemBenzodiazepines396.9Antipsychotics295.2Antidepressants with strong anticholinergic activity (TCAs, paroxetine)142.5Z-drugs122.1Antiparkinson agent with strong anticholinergic activity (benztropine)10.2Barbiturates (primidone)10.2Subtotal9617.1Pain medicationsNSAID285.0Skeletal muscle relaxants (mostly methocarbamol & cyclobenzaprine)346.0Subtotal6211.0First-generation antihistaminesMostly diphenhydramine & meclizine437.7EndocrineSulfonylureas305.3Megestrol50.9Androgens (testosterone)40.7Estrogens (estradiol)40.7Subtotal437.7Anti-infectiveNitrofurantoin10.2Total562100Abbreviations: AF Atrial fibrillation, HF Heart failure, NSAIDs Nonsteroidal anti-inflammatory drugs, PPIs Proton pump inhibitors, TCAs Tricyclic antidepressants Medications classified as potentially inappropriate according to the 2023 AGS Beers Criteria (n = 562) Abbreviations: AF Atrial fibrillation, HF Heart failure, NSAIDs Nonsteroidal anti-inflammatory drugs, PPIs Proton pump inhibitors, TCAs Tricyclic antidepressants Medications inappropriate in the presence of certain diseases or syndromes A total of 126 inappropriate medications used in the presence of certain diseases or syndromes were identified, nearly half involving cardiovascular conditions, predominantly HF (49.2%). Inappropriate medications use in patients with dementia or cognitive impairment accounted for 27.8% of cases (Table 3).Table 3Medications inappropriate in the presence of certain diseases or syndromes according to the 2023 AGS Beers Criteria (n = 126)Disease or SyndromeDrug(s)n%Cardiovascular Heart failureDiltiazem2721.4NSAIDs and COX-2 inhibitors (mostly meloxicam, ibuprofen, celecoxib, naproxen, diclofenac)2519.8Cilostazol64.8Dronedarone21.6Pioglitazone21.6Subtotal6249.2Central Nervous System Dementia or cognitive impairmentAntipsychotics (mostly quetiapine, ziprasidone)1511.9Anticholinergics (mostly cyclobenzaprine, oxybutynin)118.7Benzodiazepines (mostly lorazepam, alprazolam)86.3Z drugs (zolpidem)10.8Subtotal3527.8Kidney/urinary tract Lower urinary tract symptoms / BPHAnticholinergics (mostly meclizine, cyclobenzaprine, promethazine, diphenhydramine, hydroxyzine)1511.9 Urinary incontinence in womenNon-selective peripheral α-1 blockers (terazosin)21.6Subtotal1713.5Gastrointestinal History of gastric or duodenal ulcerAspirin129.5Total126100Abbreviations: BPH Benign prostatic hyperplasia, COX-2 Cyclooxygenase-2, NSAIDs Nonsteroidal anti-inflammatory drugs Table 3Medications inappropriate in the presence of certain diseases or syndromes according to the 2023 AGS Beers Criteria (n = 126)Disease or SyndromeDrug(s)n%Cardiovascular Heart failureDiltiazem2721.4NSAIDs and COX-2 inhibitors (mostly meloxicam, ibuprofen, celecoxib, naproxen, diclofenac)2519.8Cilostazol64.8Dronedarone21.6Pioglitazone21.6Subtotal6249.2Central Nervous System Dementia or cognitive impairmentAntipsychotics (mostly quetiapine, ziprasidone)1511.9Anticholinergics (mostly cyclobenzaprine, oxybutynin)118.7Benzodiazepines (mostly lorazepam, alprazolam)86.3Z drugs (zolpidem)10.8Subtotal3527.8Kidney/urinary tract Lower urinary tract symptoms / BPHAnticholinergics (mostly meclizine, cyclobenzaprine, promethazine, diphenhydramine, hydroxyzine)1511.9 Urinary incontinence in womenNon-selective peripheral α-1 blockers (terazosin)21.6Subtotal1713.5Gastrointestinal History of gastric or duodenal ulcerAspirin129.5Total126100Abbreviations: BPH Benign prostatic hyperplasia, COX-2 Cyclooxygenase-2, NSAIDs Nonsteroidal anti-inflammatory drugs Medications inappropriate in the presence of certain diseases or syndromes according to the 2023 AGS Beers Criteria (n = 126) Abbreviations: BPH Benign prostatic hyperplasia, COX-2 Cyclooxygenase-2, NSAIDs Nonsteroidal anti-inflammatory drugs Medications classified as “use with caution” Medications designated as “use with caution” comprised 537 prescriptions, the majority of which were agents associated with hyponatremia or syndrome of inappropriate antidiuretic hormone secretion (SIADH) (88.8%) (Table 4).Table 4Medications classified as “use with caution” according to the 2023 AGS Beers Criteria (n = 537)Medication ClassSpecific Medication(s)n%Drugs that may exacerbate or cause SIADH or hyponatremiaDiuretics (furosemide, torsemide, bumetanide)27851.8Antidepressants (TCAs, SNRIs, SSRIs, mirtazapine)14026.1Antipsychotics (mostly quetiapine, aripiprazole)305.6Tramadol295.4Subtotal47788.8SGLT2 inhibitorsDapagliflozin (only agent used)397.3Subtotal397.3Prasugrel or ticagrelor in patients ≥ 75 yearsPrasugrel91.7Ticagrelor81.5Subtotal173.2TMP-SMX used with ACEI/ ARB/ARNITMP–SMX combination40.7Total537100Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARB Angiotensin receptor blocker, ARNI Angiotensin receptor-neprilysin inhibitor, SGLT2 Sodium-glucose cotransporter-2, SIADH Syndrome of inappropriate antidiuretic hormone, SNRIs Serotonin-norepinephrine reuptake inhibitor, SSRIs Selective serotonin reuptake inhibitors, TCAs Tricyclic antidepressants, TMP SMX Trimethoprim–sulfamethoxazole Table 4Medications classified as “use with caution” according to the 2023 AGS Beers Criteria (n = 537)Medication ClassSpecific Medication(s)n%Drugs that may exacerbate or cause SIADH or hyponatremiaDiuretics (furosemide, torsemide, bumetanide)27851.8Antidepressants (TCAs, SNRIs, SSRIs, mirtazapine)14026.1Antipsychotics (mostly quetiapine, aripiprazole)305.6Tramadol295.4Subtotal47788.8SGLT2 inhibitorsDapagliflozin (only agent used)397.3Subtotal397.3Prasugrel or ticagrelor in patients ≥ 75 yearsPrasugrel91.7Ticagrelor81.5Subtotal173.2TMP-SMX used with ACEI/ ARB/ARNITMP–SMX combination40.7Total537100Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARB Angiotensin receptor blocker, ARNI Angiotensin receptor-neprilysin inhibitor, SGLT2 Sodium-glucose cotransporter-2, SIADH Syndrome of inappropriate antidiuretic hormone, SNRIs Serotonin-norepinephrine reuptake inhibitor, SSRIs Selective serotonin reuptake inhibitors, TCAs Tricyclic antidepressants, TMP SMX Trimethoprim–sulfamethoxazole Medications classified as “use with caution” according to the 2023 AGS Beers Criteria (n = 537) Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARB Angiotensin receptor blocker, ARNI Angiotensin receptor-neprilysin inhibitor, SGLT2 Sodium-glucose cotransporter-2, SIADH Syndrome of inappropriate antidiuretic hormone, SNRIs Serotonin-norepinephrine reuptake inhibitor, SSRIs Selective serotonin reuptake inhibitors, TCAs Tricyclic antidepressants, TMP SMX Trimethoprim–sulfamethoxazole DDIs that should be avoided A total of 165 clinically important DDIs were identified, most commonly involving the concurrent use of three or more CNS-active medications, α-1 blockers with loop diuretics, or opioids with gabapentinoids (Table 5).Table 5Drug–drug interactions that should be avoided according to the 2023 AGS Beers Criteria (n = 165)Drug-Drug Interactionn%Combination of ≥ 3 of CNS-active drugs(Mostly gabapentin, pregabalin, hydrocodone, tramadol, diazepam, temazepam, citalopram, nortriptyline, zolpidem)4929.7Non-selective peripheral α-1 blockers + loop diuretics4728.5Opioids + gabapentin/pregabalin4627.9Opioids + benzodiazepines148.5Anticholinergic + anticholinergic(Mostly promethazine, hyoscyamine, prochlorperazine, cyclobenzaprine, meclizine, hydroxyzine)84.8ACEI + ARNI10.6Total165100Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARNI Angiotensin receptor-neprilysin inhibitor, CNS Central nervous system Table 5Drug–drug interactions that should be avoided according to the 2023 AGS Beers Criteria (n = 165)Drug-Drug Interactionn%Combination of ≥ 3 of CNS-active drugs(Mostly gabapentin, pregabalin, hydrocodone, tramadol, diazepam, temazepam, citalopram, nortriptyline, zolpidem)4929.7Non-selective peripheral α-1 blockers + loop diuretics4728.5Opioids + gabapentin/pregabalin4627.9Opioids + benzodiazepines148.5Anticholinergic + anticholinergic(Mostly promethazine, hyoscyamine, prochlorperazine, cyclobenzaprine, meclizine, hydroxyzine)84.8ACEI + ARNI10.6Total165100Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARNI Angiotensin receptor-neprilysin inhibitor, CNS Central nervous system Drug–drug interactions that should be avoided according to the 2023 AGS Beers Criteria (n = 165) Abbreviations: ACEI Angiotensin-converting enzyme inhibitor, ARNI Angiotensin receptor-neprilysin inhibitor, CNS Central nervous system Medications requiring avoidance or dose adjustment in impaired kidney function One hundred eighteen medications required avoidance or dose adjustment due to impaired kidney function, with CNS and analgesic agents comprising nearly half (49.1%) and cardiovascular medications one third (33.9%) (Table 6).Table 6Medications to be avoided or dose-adjusted in patients with impaired kidney function according to the 2023 AGS Beers Criteria (n = 118)Medication ClassMedication(s)n%Central nervous system and analgesicsGabapentin/pregabalin3126.3Tramadol119.3NSAIDs (mainly nabumetone, meloxicam, celecoxib)86.8Duloxetine75.9Baclofen10.8Subtotal5849.1Cardiovascular & antithromboticsSpironolactone3731.4Rivaroxaban21.7Triamterene10.8Subtotal4033.9GastrointestinalFamotidine1613.6HyperuricemiaColchicine21.7Anti-infectiveCiprofloxacin10.8Nitrofurantoin10.8Subtotal21.7Total118100Abbreviations: NSAIDs Nonsteroidal anti-inflammatory drugs Table 6Medications to be avoided or dose-adjusted in patients with impaired kidney function according to the 2023 AGS Beers Criteria (n = 118)Medication ClassMedication(s)n%Central nervous system and analgesicsGabapentin/pregabalin3126.3Tramadol119.3NSAIDs (mainly nabumetone, meloxicam, celecoxib)86.8Duloxetine75.9Baclofen10.8Subtotal5849.1Cardiovascular & antithromboticsSpironolactone3731.4Rivaroxaban21.7Triamterene10.8Subtotal4033.9GastrointestinalFamotidine1613.6HyperuricemiaColchicine21.7Anti-infectiveCiprofloxacin10.8Nitrofurantoin10.8Subtotal21.7Total118100Abbreviations: NSAIDs Nonsteroidal anti-inflammatory drugs Medications to be avoided or dose-adjusted in patients with impaired kidney function according to the 2023 AGS Beers Criteria (n = 118) Abbreviations: NSAIDs Nonsteroidal anti-inflammatory drugs In multivariable negative binomial regression analysis, higher PIM burden was independently associated with CKD (IRR 1.38, 95% CI 1.20–1.59), depression (IRR 1.29, 95% CI 1.05–1.58), BPH (IRR 1.26, 95% CI 1.03–1.53), and increasing number of comorbidities (IRR 1.04 per additional condition, 95% CI 1.00–1.07); all p < 0.05. No significant multicollinearity was detected among covariates (all VIFs < 3.0). In sensitivity analysis using multivariable logistic regression, no covariates were significantly associated with the presence of at least one PIM (all p > 0.05). This likely reflects reduced sensitivity associated with dichotomizing a count outcome. Discussion This study highlights a substantial and clinically relevant burden of inappropriate prescribing among older adults with HF. Notably, the vast majority of patients were exposed to at least one PIM and on average, more than one in four medications in a patient’s home regimen met criteria for potential inappropriateness, underscoring the pervasive nature of medication-related risk in this vulnerable population. The observed prevalence of PIM use in this cohort is concerning but not unexpected given the high degree of polypharmacy, with patients receiving an average of eleven home medications at admission. HF treatment frequently necessitates the concurrent use of multiple pharmacologic agents to address both cardiac and non-cardiac comorbidities, increasing the likelihood of prescribing cascades and inadvertent exposure to medications with unfavorable risk–benefit profiles in older adults [15, 16]. Furthermore, common features of HF, including renal dysfunction, frailty, and fluctuating volume status, may further amplify susceptibility to adverse drug events when PIMs are prescribed. Prior studies have reported substantial, though generally lower, rates of PIM use among older adults hospitalized with HF. In a large U.S. cohort of older patients admitted with HF, 61.1% had at least one PIM at hospital admission based on the 2019 Beers Criteria [14]. More recently, a study conducted in China among older adults hospitalized for HF using the updated 2023 Beers Criteria identified PIMs in 75.5% of patients [17]. Although this more contemporary estimate is closer to the prevalence observed in the present study, our findings demonstrate an even higher burden, with 90.1% of patients exposed to at least one PIM. Differences in patient complexity, increasing number of comorbidities, prescribing practices, and healthcare settings may partially explain this variation. However, methodological factors likely also contributed, as our definition of PIM exposure included medications classified as “use with caution,” DDIs, and those requiring renal dose adjustment or avoidance, with potential overlap across categories. Furthermore, the use of the updated 2023 AGS Beers Criteria may partly account for the higher prevalence compared with earlier studies based on the 2019 criteria, given the inclusion of expanded DDI warnings, more detailed renal adjustment recommendations, and refined drug–disease interaction classifications. Collectively, these factors suggest that the true prevalence of PIM use in older adults with HF, particularly in hospitalized populations, may be underestimated when earlier criteria or less comprehensive assessments are applied. While the AGS Beers Criteria provide a comprehensive, non–disease-specific framework for identifying PIMs in older adults, disease-specific tools such as the PIPHFrEF criteria offer complementary insights tailored to HF management. These tools may better capture medications that directly affect HF pathophysiology, including those that worsen cardiac function or promote fluid retention. For example, data from the Swedish HF Registry (2005–2020) reported that approximately 47% of patients with HFrEF were exposed to at least one potentially inappropriate prescription using HF-specific criteria [8]. Although direct comparison is limited by differences in methodology, populations, and definitions, these findings highlight how the choice of prescribing assessment tool can substantially influence estimated prevalence. Accordingly, reliance on the Beers Criteria alone may overestimate inappropriate prescribing in some contexts while underrepresenting HF–specific risks. This distinction is important when interpreting the high prevalence observed in the present study and should be considered a limitation, as the Beers Criteria do not fully reflect disease-specific prescribing considerations in patients with HF. Integrating both general and disease-specific approaches may provide a more comprehensive evaluation of prescribing quality in this population. Patients with HF represent a particularly vulnerable population in the context of potentially inappropriate prescribing. The coexistence of multimorbidity, renal dysfunction, and hemodynamic instability often necessitates complex pharmacotherapy while simultaneously increasing susceptibility to adverse drug events. In addition, several medications commonly encountered in the Beers Criteria, such as NSAIDs, certain calcium channel blockers, and drugs promoting sodium or fluid retention, may directly worsen HF pathophysiology by impairing myocardial function, promoting fluid retention, or exacerbating renal dysfunction. These factors highlight the importance of careful medication selection and regular reassessment of prescribing appropriateness in older adults with HF. Medications classified as potentially inappropriate Medications classified as potentially inappropriate according to the 2023 AGS Beers Criteria constituted a substantial proportion of the overall prescribing burden in this cohort. Agents affecting the GI tract, cardiovascular and antithrombotic systems, and the CNS predominated. Among GI agents, chronic use of PPIs constituted approximately 32% of medications classified as potentially inappropriate. Although PPIs are effective for the treatment and prevention of gastroesophageal reflux disease and peptic ulcer disease, their long-term use, unless prescribed for high-risk patients or after unsuccessful trials of dose reduction, discontinuation, or substitution, is considered potentially inappropriate in older adults. Prolonged PPI therapy has been associated with adverse outcomes, including Clostridioides difficile infection, pneumonia, bone loss, and fractures. When clinically appropriate, alternative strategies include nonpharmacologic interventions, step-down therapy, histamine-2 receptor antagonists, and alginic acid for breakthrough symptoms [18]. CNS medications classified as potentially inappropriate were also commonly prescribed, with benzodiazepines constituting the most frequently identified class, followed by antipsychotics, and antidepressants with strong anticholinergic activity. The use of these agents in older adults is discouraged due to well-established associations with adverse outcomes, including cognitive impairment, delirium, excessive sedation, falls, and fractures. The predominance of CNS PIMs underscores the continued exposure of older adults with HF to medications with unfavorable risk–benefit profiles, despite the availability of safer therapeutic alternatives and nonpharmacologic approaches. Benzodiazepines are commonly prescribed for the management of anxiety and insomnia; however, safer and more effective strategies exist for both conditions in older adults. First-line management of anxiety emphasizes nonpharmacologic interventions, with selective serotonin reuptake inhibitors (SSRIs) or serotonin–norepinephrine reuptake inhibitors (SNRIs) preferred when pharmacologic therapy is indicated [18]. Antipsychotics, the second most prevalent CNS PIM identified, carry substantial risks when used outside of well-defined indications, particularly in older adults with dementia. Their use has been associated with increased risks of cerebrovascular events and mortality. In the management of delirium, first-line treatment focuses on multicomponent nonpharmacologic strategies that address modifiable risk factors and support orientation, mobility, sleep, and hydration [19]. Antipsychotics should be reserved for short-term use only in cases of severe behavioral symptoms that pose an immediate risk to the patient or others and persist despite nonpharmacologic interventions [18]. Similarly, in patients with dementia who exhibit agitation or aggression, careful assessment and management of contributing factors, including underlying medical conditions, medication effects, and environmental stressors, are essential, with antipsychotics considered only when nonpharmacologic approaches fail and there is a substantial risk of harm [18]. Beyond medication safety, exposure to PIMs may contribute to important geriatric outcomes in older adults with HF, including falls, functional decline, loss of independence, institutionalization, and mortality. Several medication classes identified in this study, particularly CNS-active drugs and agents with anticholinergic effects, have been consistently associated with delirium, impaired mobility, and fall-related injuries in older adults. These findings highlight the importance of targeted interventions aimed at reducing inappropriate prescribing. Structured medication review during hospital admission and discharge, pharmacist-led deprescribing initiatives, and routine use of geriatric prescribing tools such as the Beers Criteria may help reduce medication-related harm and improve clinical outcomes in this vulnerable population. Aspirin prescribed for primary prevention was a commonly identified potentially inappropriate cardiovascular medication. Contemporary guidelines discourage routine aspirin use for primary prevention in older adults due to an unfavorable risk–benefit profile, with bleeding risk outweighing potential cardiovascular benefit [20]. The frequency observed in this cohort highlights the continued use of aspirin despite evolving evidence and underscores the need for regular medication review and deprescribing when appropriate. Medications inappropriate in the presence of certain diseases or syndromes Medications deemed inappropriate in the presence of specific diseases or syndromes represent a particularly important category of potentially inappropriate prescribing, as their use may directly exacerbate underlying conditions and worsen clinical outcomes. In patients with HF, certain medications are known to promote fluid retention, impair myocardial contractility, or increase the risk of hospitalization and mortality, making their avoidance a key principle of evidence-based HF management [21]. In the present study, nearly half of all disease-specific PIMs, most commonly diltiazem and nonsteroidal anti-inflammatory drugs (NSAIDs), were prescribed in the context of HF, underscoring the continued exposure of older adults to agents that may directly compromise HF stability. Medications classified as “use with caution” Medications classified as “use with caution” according to the 2023 AGS Beers Criteria accounted for a substantial proportion of identified PIMs, with agents associated with syndrome of inappropriate antidiuretic hormone secretion (SIADH) or hyponatremia predominating. Similar patterns have been reported in prior studies, where medications associated with hyponatremia, particularly diuretics and antidepressants, were among the most frequently identified “use with caution” agents in older adults [17]. Nearly nine out of ten medications in this category were drugs commonly used in HF and comorbid conditions, including loop diuretics, antidepressants, antipsychotics, and tramadol. Although many of these agents are often clinically necessary, their use in older adults with HF warrants careful consideration due to the heightened risk of electrolyte disturbances, which may negatively affect patient safety. These findings emphasize the importance of close laboratory monitoring, dose optimization, and periodic reassessment of ongoing need, rather than routine avoidance, when prescribing medications classified as “use with caution” in this population. Importantly, not all medications identified by the Beers Criteria as potentially inappropriate are inherently inappropriate in patients with HF. Several agents categorized as “use with caution,” including loop diuretics and sodium–glucose cotransporter-2 (SGLT2) inhibitors, are cornerstone therapies in contemporary HF management and confer substantial clinical benefit. Similarly, angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) remain foundational treatments, although certain combinations—such as with trimethoprim–sulfamethoxazole—are flagged due to an increased risk of adverse effects, particularly hyperkalemia. These findings highlight the importance of interpreting Beers Criteria within the appropriate clinical context, emphasizing individualized risk–benefit assessment and careful monitoring rather than automatic discontinuation. DDIs that should be avoided Potentially inappropriate DDIs were common in this cohort, reflecting the complexity of polypharmacy in older adults with HF. The most frequent interactions involved multiple CNS-active agents, combinations of α-1 blockers with loop diuretics, and opioids with gabapentinoids or benzodiazepines. This is consistent with previous literature, where CNS polypharmacy and opioid-related combinations are among the most commonly reported high-risk DDIs in older adults [14, 17]. These combinations increase the risk of sedation, falls, and hypotension, emphasizing the need for careful medication review and selection of safer alternatives when multiple high-risk drugs are prescribed concurrently. Medications requiring avoidance or dose adjustment in impaired kidney function Medications requiring avoidance or dose adjustment in patients with impaired kidney function most commonly involved CNS/analgesic agents and cardiovascular medications, accounting for nearly half and one-third of cases, respectively, with gabapentin/pregabalin and spironolactone being the most commonly identified agents, underscoring the importance of careful renal function assessment in older adults receiving these therapies. In this study, higher PIM burden was independently associated with CKD, depression, BPH, and increasing number of comorbidities. CKD was associated with a 38% higher expected number of PIMs, while each additional comorbidity increased PIM burden by approximately 4%. Patients with CKD had the greatest increase in PIM burden, reflecting the frequent need for renal dose adjustments and the continued use of renally cleared or nephrotoxic agents in older adults with HF. The association with depression likely reflects frequent use of antidepressants associated with hyponatremia, falls, or anticholinergic burden, which are highlighted in the Beers Criteria, particularly in older adults with HF. Importantly, each additional comorbidity was associated with a progressive increase in PIM burden, underscoring the cumulative impact of multimorbidity and therapeutic complexity on potentially inappropriate prescribing. This study has several strengths, including a randomly selected cohort of older adults with HF from two hospitals, with a sample size exceeding the minimum required for precise estimation of PIM prevalence, reducing selection bias within the eligible population, although findings may reflect prescribing practices specific to the study setting. The comprehensive evaluation of medications according to the 2023 AGS Beers Criteria allowed for detailed assessment of multiple PIM categories, including disease-specific, renal, and DDI considerations. Limitations include its retrospective, cross-sectional design, which precludes assessment of causality or longitudinal outcomes. Additionally, certain potentially relevant variables, including HF subtype and standardized comorbidity indices, were not consistently available and could not be included in the analysis. Although both hospitals belong to the same health system, hospital site was not included as a covariate; thus, a residual inter-site differences in patient case-mix or prescribing practices cannot be excluded. This study adds to the existing literature by providing a comprehensive evaluation of PIM use in older adults with HF using the most recent 2023 AGS Beers Criteria. Compared with prior studies using earlier criteria, this approach allows for a more contemporary assessment of prescribing practices, including updated medication classifications and DDI considerations. Additionally, the identification of specific clinical predictors of higher PIM burden may help inform targeted interventions in high-risk subgroups. Conclusions In older adults with HF, the high prevalence of PIM exposure at hospital admission represents a critical opportunity for intervention. Given that home medications at admission often reflect prescribing decisions made at prior hospital discharge, transitions of care provide an ideal setting to reassess medication appropriateness, optimize therapy according to geriatric prescribing principles, and implement deprescribing or safer alternatives. Targeted review is particularly important in patients with renal impairment, multiple comorbidities, or polypharmacy, and may help reduce preventable medication-related harm, improve clinical outcomes, and enhance overall quality of care in this vulnerable population.