Introduction

Polypharmacy, often defined as 5 or more regular medications, has become an increasingly prevalent concern in healthcare, particularly among the aging population1,2. Older people often have multiple chronic conditions, leading to a higher likelihood of being prescribed multiple medications3. However, the unintended consequences of polypharmacy, including adverse drug reactions, drug-drug interactions, and decreased medication adherence, pose significant risks to the well-being of older people. The complexities of managing numerous medications from multiple prescribers, sometimes in different health systems, further exacerbate the problem and demand a comprehensive approach to mitigate potential harms4,5,6.

Pharmacist-led ambulatory clinics have the potential to improve quality of care and health outcomes for patients with a multitude of conditions7,8,9. These clinics leverage the specialized expertise of pharmacists to provide comprehensive care that goes beyond medication dispensation7. In the realm of diabetes management, pharmacist-led clinics offer personalized guidance on medication regimens, blood glucose monitoring, and lifestyle modifications. Similarly, in hypertension management, pharmacists in clinics closely monitor blood pressure, adjust medications as necessary, and offer lifestyle recommendations10. Pharmacist-led diabetes and hypertension clinics have been shown to significantly improve clinical outcomes10,11. By tailoring interventions to individual patient needs and providing education and counseling to patients and caregiver(s), pharmacist-led clinics hold the potential to enhance treatment outcomes, foster medication adherence, and ultimately contribute to improved quality of life for older individuals grappling with multiple chronic conditions and polypharmacy.

A systematic approach to measuring the unintended consequences and complexities of polypharmacy has not yet been established. It has been difficult to demonstrate a direct association between specific drug-related problems (DRPs) and clinical outcomes, let alone classifying and measuring the impact that multiple DRPs have on a frail patient with multiple comorbidities1,5. However, examining medication-related outcomes established in literature can demonstrate gaps in the quality of care provided to older adults. The first step in identifying polypharmacy issues is to verify how a patient is taking their medications compared to the list documented in the electronic health record (EHR) through an interactive medication reconciliation process12. An inaccurate medication list increases the potential for significant medication errors and adverse outcomes. Improving the accuracy of a patient’s medication list is critical to managing olypharmacy13. Once the patient’s medication list is reconciled, an evaluation of the appropriateness of each medication is warranted. During this assessment, each medication should be reviewed for its indications, benefits, side effects, and drug and disease interactions specific to the individual patient’s goals. The American Geriatrics Society (AGS) periodically publishes a list of Potentially Inappropriate Medications (PIMs) for older adults referred to as the Beers Criteria14. Although this is a standardized list of PIMs to objectively measure medication-related outcomes, some PIMs may be appropriate in certain specific situations. Conversely, some medications that are not on the AGS list of PIMs may be potentially inappropriate for certain patients. Assessing the appropriateness of the medications requires a thorough review of the chart, in addition to a conversation with the patient and/or their caregiver(s). Finally, once the pharmacist completes the clinical assessment, the pharmacist formulates interventions focusing on deprescribing, improving adherence, and educating prescribers, patients, and caregivers. The pharmacist may schedule a follow-up polypharmacy consult if there are medication changes or if the medication list is particularly complex.

Cedars-Sinai Medical Network (CSMN), the outpatient arm of the Cedars-Sinai Health System, a large, multi-hospital health system in Los Angeles, California, implemented a polypharmacy clinic with the objective to improve medication management among its population of older patients. In this paper, we describe the design of the clinic and offer a descriptive pre-post evaluation of process outcomes of patients attending the clinic from October-November 2022.

Objective

The objective of this pharmacist-led polypharmacy program evaluation was to demonstrate the value and role of clinical pharmacists in these polypharmacy consults for older patients. We aimed to define and describe the impact of polypharmacy visits in the CSMN Geriatrics Department and to identify improvement opportunities and reportable outcomes.

Setting

Cedars-Sinai Medical Network includes more than 1,000 physicians and 330,000 unique patients in multiple specialties throughout Los Angeles County. The CSMN Pharmacy Services Department supports primary care and specialty physicians through drug therapy management programs, as well as transitions of care, refills, and drug utilization and distribution services.

Clinic design and referral process

The polypharmacy clinic for older patients started in 2017 with one clinical pharmacist board certified in geriatrics who was staffed in the clinic at 0.1 FTE (half-day a week). The clinic was embedded in the Geriatrics Department with three geriatricians. The clinical pharmacist provided polypharmacy consults only. In 2021, with the expansion to five geriatricians, the polypharmacy clinic evolved to two clinical pharmacists staffing 0.7 FTE, providing polypharmacy consults and chronic disease management for diabetes, hypertension, and dyslipidemia per CSMN collaborative practice agreements. This study focuses on data collected only for the polypharmacy consults for older patients.

Prior to 2022, polypharmacy clinic patients were informally referred through email communication or EHR messages by providers, pharmacy team members, and case management team members. In 2022, a formal referral system was established in the EHR requiring electronic referrals by geriatricians, primary care providers, and clinical pharmacists. Referral reasons varied and included patients enrolled in a patient-centered clinic for highly complex patients, patients identified as having specific medication issues (e.g., medication adherence), patients with broad polypharmacy issues, and patients with insurance or pharmacy issues. Recently discharged patients identified as having poor medication adherence or medication literacy through the MedAL (Medication Adherence and Literacy) score15,16 were also referred to the polypharmacy clinic. The MedAL was initially developed in the inpatient setting of the health system by the hospital’s transitions of care team (i.e., from hospital to post-discharge settings) as a set of questions aimed at identifying, standardizing, and quantifying medication adherence and literacy13. The three questions used by the pharmacists during the post-discharge phone call included: (1) Do you ever forget to take your medications?, (2) How many times in the past week did you miss your scheduled medications?, (3) Are there any reasons you take your scheduled medications differently than prescribed? The pharmacist also estimates the degree of patient’s knowledge of the medication (name, dose, frequency of the medication) and notes whether the patient needs additional education. Each question gets one point, and higher scores indicate a higher need for subsequent follow up from a pharmacist.

Patients, participants

Geriatricians in the health system’s primary medical group are able to refer patients to the polypharmacy clinic from any health plan. Patients whose primary care provider is not a geriatrician but who are considered a “population health” patient (e.g., have a Medicare Advantage plan or are part of an Accountable Care Organization where the health system assumes some of the capitated risk for caring for the patient) are referred to the polypharmacy clinic based on their medication needs as determined by case management team, pharmacy team, and primary care team.

Interventions

Pharmacist services

Polypharmacy consults are conducted in-person, via video, or via telephone. Initial visits are scheduled for 60 minutes while follow-up visits are scheduled for 30 minutes and can be adjusted at 30-minute increments.

During the polypharmacy consult, the clinical pharmacist conducts a comprehensive medication review of all prescribed regular and as-needed medications, over-the-counter medications, herbal supplements, and vitamins. Medication adherence and medication literacy are assessed during the visit. Throughout the consult, the clinical pharmacist provides targeted education to the patient and caregiver(s) and simplifies the patient’s daily and weekly medication regimen when possible. The clinical pharmacist addresses specific medication-related issues based on the referral reason and identifies drug-related problems and deprescribing opportunities.

At the end of the visit, the clinical pharmacist provides patient recommendations through education and counseling. Since the polypharmacy clinic does not have a collaborative practice agreement given the wide scope of disease types of the patient population, clinical pharmacists’ recommendations to providers are discussed with referring physician, primary care providers, and specialists. Any changes in medications and monitoring frequency are documented in the EHR and communicated to the patient and caregiver(s). The medication list is updated, and if needed, the clinical pharmacist provides a medication chart that outlines timing of medication administration. Follow-up visits, which are often conducted via phone or by video visit, are scheduled at the discretion of the clinical pharmacist.

Design

Method

Our evaluation study design used a retrospective standardized chart review of polypharmacy consult visits between October 1 and November 30, 2022. We used a standardized abstraction spreadsheet to collect patient demographics, insurance plan, medication data, DRPs, and pharmacists’ interventions. Systematic data collection was completed by March 2023 over a three-month period.

For some data points (selected demographics, number of hospitalizations in the last 12 months, dementia diagnosis, medications at pre-, at-, and post-visit), we used an EHR data extract from the health system’s EHR data warehouse (Clarity) which contains clinical, medication, order, laboratory, imaging and other data for more than 2 million patient records for the health system. To determine whether patients were diagnosed with dementia, we used ICD-10 codes for dementia using the Bynum-Standard algorithm for dementia identification in administrative data which were included in the EHR problem list or used in billing17. We opted to examine patients with and without dementia given the increased risk of polypharmacy in older adults with dementia found in a recent large-scale study of persons with dementia in the U.S. (Odds Ratio 3.0; 95% CI: 2.1–4.3)18. Moreover, for clinicians, managing medications in patients with dementia requires additional considerations regarding PIMs, medication adherence, and caregivers education19.

For each visit, medication data was collected at three time points—“pre-visit”, “at-visit”, and “post-visit”. “Pre-visit” medications were listed in the EHR before the clinical pharmacist started the polypharmacy visit. The “pre-visit” data is defined as the time point at which medication data is obtained at the latest billable encounter visit prior to the pharmacy consult. This can range anywhere between 1-day to 12-months prior to the “at-visit” pharmacy consult. Throughout the interactive polypharmacy consult, the clinical pharmacist ascertains the patient’s verified medication regimen and reconciles the “pre-visit” medication list to produce the “at-visit” medication list. To account for delays from provider discussion and response time, “post-visit” medication data is collected at the earliest billable encounter 90 days after the polypharmacy consult, ranging from 90-days to 180-days after the consult. Obtained at subsequent provider’s visit, the “post-visit” medication list reflects the accepted recommendations to patient and providers after the polypharmacy consult.

The Cedars-Sinai Health System’s Institutional Review Board (CSHS IRB) does not require IRB submission for this project as it is not “human subject research” as defined by DHHS (45 CFR 46) or FDA regulations (21 CFR 50). Because this study design was part of a quality improvement program evaluation, the IRB deemed this study exempt from IRB (ethics) review and a waiver of informed consent was given for this work. All methods were carried out in accordance with relevant guidelines and regulations.

Outcomes

We collected numerous medication-related process outcomes, including the number of medications, number and types of identified DRPs, number of PIMs based on 2019 AGS Beers Criteria Tables 2 and 714.

Medication list accuracy is defined as the number of accurate medications listed on the medication list (with the correct name, dosage, route of administration, and frequency) divided by the sum of pre-visit medications and the number of additional medications identified at the visit that the patient is taking but which are not on the medication list13.

$$ \frac{{{\text{\# ~of~}}\;{\text{Accurate~}}\;{\text{Medications}}}}{{{\text{\# of Additional}}\;{\text{~Medications~}}\;{\text{not~}}\;{\text{on}}\;{\text{~Med~}}\;{\text{List + \# of PreVisit~}}\;{\text{Medications}}}} $$

“Pre-visit” and “at-visit” data were used to calculate medication list accuracy. “At-visit” and “post-visit” data were used to determine the changes in number of PIMs.

We modified the Institute of Medicine’s Medication-Use Process in Identifying and Preventing Medication Errors framework12 and the Pharmaceutical Care European Network Association’s Classification for Drug-Related Problem20 to produce a simple and applicable classification system for the ambulatory care setting. This consists of three categories of DRPs and their corresponding subcategories, including: (1) patient-related DRPs: medications which were identified as omissions, extraneous, or not taking as prescribed; (2) prescribing-related DRPs: which were identified as omissions, extraneous, medication allergies, drug-drug interactions, drug-disease interactions, duplicate therapies, wrong medications, wrong dose/frequency, wrong time of administration, incomplete order, no prescription given (e.g., the physician intended to make medication changes but physical/electronic new prescriptions were not provided/sent to pharmacy), monitoring error, adverse-drug reaction; (3) insurance/pharmacy-related DRPs: which were identified as non-formulary medications, and medications currently not in stock.

Analysis

We used descriptive statistics and examined means, standard deviations, and medians for all outcomes. We tested the normality of the medication outcome variables (number of medications, number of Beers Table 2 and Beers Table 7 medications) using the Shapiro-Wilk test. For non-normally distributed outcomes, we used the Wilcoxon rank-sum test to examine the differences between the means. We used Stata SE 18 (StataCorp. 2023. Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC) for statistical analyses.

Financing

Clinical pharmacists bill an “incident-to” service CPT code 99211 for in-office and virtual polypharmacy visits, and there is no charge for telephone polypharmacy visits.

Results

Patient demographics and clinical characteristics

Between October 1 to November 30, 2022, pharmacists completed a total of 84 polypharmacy consults. The average age of patients seen in the polypharmacy clinic was 80 years of age (SD of 11; median of 81). The demographics of the patient population, see Table 1, included: 70% of patients identified as White (n = 59), 19% of patients identified as Black (n = 16), 10% identified as Other (n = 8), and 1% identified as Asian (n = 1). Payor mix was comprised of 48% Medicare Advantage (n = 40), 26% Medicare Fee-For-Service (n = 22), 18% had other private health care plans (n = 15). 13% of patient-visits had a diagnosis of dementia (n = 11).

Table 1 Patient-visit characteristics of consults to the CSMN polypharmacy clinic.

Polypharmacy consult visits are characterized in Table 2. Of 84 polypharmacy consults completed during the study period, 42% were initial visits (n = 35) and 58% were follow-up visits (n = 49). Three-quarters of the polypharmacy patient-visits were referred for polypharmacy issues (n = 63), 18% of patients-visits were referred by high MedAL score for patients with poor medication adherence and poor literacy (n = 15), 7% were referred because patients were enrolled in complex patient clinic (n = 6).

Table 2 Polypharmacy consult characteristics.

During this study period, providers placed a total of 39 polypharmacy clinic referrals, some visits which occurred after the evaluation period. 61% of the referrals were placed by geriatricians (n = 24), 26% were placed by other pharmacists (n = 10), 8% were placed by the complex patient clinic team (n = 3), and 5% of the referrals placed by other providers (n = 2).

Medication-related outcomes

Overall, patients were on an average of 17.3 medications (range: 7–33) at the beginning of the polypharmacy consult visit (“at-visit”) (Table 3). This was defined as the number of medications on the patient’s medication list before the visit (“pre-visit”) and the number of additional medications identified at the visit that the patient was taking but that were not on the medication list. After the polypharmacy consult (“post-visit”), patients were on an average of 15.9 medications (range: 4–30), with an average of 1.4 medications deprescribed per polypharmacy consult visit (Table 4). The z-score for Wilcoxon rank-sum test for the differences between the at-visit and post-visit medications among all patients was 1.36 (p-value: 0.17). Among patients taking 7–16 at-visit medications, 0.9 medications were deprescribed (z-score: − 0.09, p-value: 0.93). Among patients taking 17–33 medications, an average of 2.6 medications were deprescribed per polypharmacy consult (z-score: 2.06, p-value: 0.04).

Table 3 Medication-related outcomes of polypharmacy consult visits.
Table 4 Change in Medication-related Outcomes of Polypharmacy Consult Visits.

The medication list accuracy increased by 6% at follow-up polypharmacy visits (72%) compared to at the initial polypharmacy visits (66%). Patients were on an average of 1.8 potentially inappropriate medications (PIMs) as defined by 2019 AGS Beers’ Criteria “Table 2 List of Potentially Inappropriate Medications”14 (Table 5). 44% of patients were on 1 or more PIMs. The number of patients on PIMs was reduced by 28% after the polypharmacy consult. On average, patients were on 1.3 strong anticholinergic medications as defined by 2019 AGS Beers’ Criteria “Table 7 List of Drugs with Strong Anticholinergic Properties” (DSAPs)14. 24% of patients were one 1 or more DSAPs. The number of patients on these DSAPs was reduced by 54% after the polypharmacy consult. The z-score for Wilcoxon rank-sum test for the differences between the at-visit and post-visit Beers Table 2 medications was 0.72 (p-value: 0.27). The z-score for the Beers Table 7 medications was 0.44 (p-value: 0.65).

Table 5 Medication-related outcomes of AGS’ Beers Criteria of high risk medications.

Clinical pharmacists identified 1.4 DRPs per patient. Overall, pharmacists identified 90 patient-related DRPs, 13 prescribing-related DRPs, and 11 insurance/pharmacy-related DRPs (Table 6).

Table 6 Classification of Drug-related Problems and Pharmacist intervention.

Patient-related drug related problems

The majority (39%, n = 35) of patient-DRPs were classified as “omissions” (i.e., patient not taking a prescribed medication) followed by “not taking as prescribed” (37%, n = 33), and “extraneous” (i.e., patient taking a medication that was not prescribed) (24%, n = 22) (Table 7). The clinical pharmacists intervened and provided 403 recommendations to the patient. Most of these interventions were classified as “patient education and counseling” (73%, n = 296), followed by “discontinue” (10%, n = 41), “adjust” (9%, n = 35), and initiate (8%, n = 31).

Table 7 Subcategories and interventions of patient-related Drug-Related Problems.

Prescribing-related drug related problems

Majority of prescribing-related DRPs were classified as “drug-drug interactions” (38%, n = 5), followed by “duplicate medications and extraneous” (25%, n = 3), and “omission and inconsistency between notes and medication list” (8%, n = 1) (Table 8). Clinical pharmacists intervened and provided 92 recommendations to providers; 65% of all recommendations (n = 60) were accepted. 75 out of the 84 visits required the pharmacist to reach out to the providers. Most of these interventions were classified as “monitor” (40%, n = 37), followed by “adjust” (36%, n = 33), “discontinue” (15%, n = 14), and “initiate” (9%, n = 8).

Table 8 Subcategories and interventions of prescribing-related Drug-Related Problems.

Insurance or pharmacy-related drug related problems

Insurance-related DRPs include formulary issues requiring a change in pharmaceutical therapy. Pharmacy-related DRPs may include drug coverage issues or medication organization tools (i.e. setting up multi-dose packaging) requiring interventions at the pharmacy level. The clinical pharmacist provided interventions on the 11 identified insurance/pharmacy-related DRPs.

Discussion

Our descriptive evaluation demonstrates that the CSMN Polypharmacy Program for older patients was effective in improving medication list accuracy, discontinuing and deprescribing PIMs and other high-risk medications in older individuals. The process of evaluating this program highlights the complexity of polypharmacy issues in older patients and advocates for targeted interventions and dedicated time to identify and resolve DRPs. With growing number of older individuals and the rise in polypharmacy-related issues, health systems can benefit from effectively identifying patients at risk and providing targeted interventions to prevent polypharmacy-related adverse events.

In our evaluation, we found that of 84 polypharmacy consults, 37 patient-visits had at least 1 PIM prescribed, which represents 44% of patient visits. Similarly, we found that 24% of patient-visits had at least 1 DSAP prescribed. These findings represent significant room for improvement in prescribing given the risks of sedative and anticholinergic medications in older adults.

We identified numerous issues during our program evaluation related to the broader field of identifying and managing polypharmacy. First, more guidance is needed to define and identify patients at risk of polypharmacy-related adverse events. The number of medications used to objectively define polypharmacy (> 4 medications) and excessive or hyper-polypharmacy (> 9 medications) may be insufficient to identify patients with polypharmacy-related issues21. While most patients in the program could be categorized into the excessive polypharmacy group, some patients had many PIMs or potentially high-risk medications. Moreover, patients with a smaller number of medications could still remain at high risk due to the risk profile of particular medications22 and the patient’s disease or functional state (e.g., frailty). The concepts of appropriate and inappropriate polypharmacy are noteworthy23. While the number of medications alone may be a simple indication of risk, the issue of appropriate verse inappropriate polypharmacy is multi-level24,25. If the benefits outweigh the risks, some PIMs may be appropriate in some patients. For example, clonazepam in a patient with a rapid eye movement (REM) sleep disorder who failed other non-PIM therapy may be considered appropriate. On the other hand, a non-PIM medication may be inappropriate for some patients. For example, the use of amlodipine in a patient on three antihypertensive agents who experiences hypotension and edema may be inappropriate. Furthermore, with the use of multiple single-disease guidelines, older individuals with multiple comorbidities are subjected to an increasing number of medications and increasingly complex medication regimens, leading to lower adherence and potentially to overprescribing due to a subtherapeutic or ineffective response1,5.

Subsequently, we found that demonstrating the value of a polypharmacy intervention was challenging, given the heterogeneity of the medication lists and comorbidities of the patients. Moreover, in a non-randomized study, it is challenging to causally connect the intervention to improvement or changes in clinical outcomes such as hospitalizations. Our evaluation focused on process outcomes, identifying improvement in medication list accuracy and deprescribing medications, but we did not examine clinical outcomes or hospitalizations. One of the challenges of untangling the effects of polypharmacy-related interventions on outcomes is separating the effects of polypharmacy from the patients’ underlying medical conditions1,5. For example, older adults are more prone to orthostatic hypotension because of impaired baroreceptor sensitivity. If a patient experiences orthostatic hypertension, it can be challenging to distinguish the cause, whether from older age, existing medical conditions, or adverse effects from multiple antihypertensive agents. Identifying appropriate or inappropriate polypharmacy requires clinical expertise that is more practical to implement in a clinic setting that allows for follow-up opportunities. This may require trials of dose decreases or drug discontinuation, or sometimes additions or up-titrations, and then follow up visits with the patient to monitor if the patient is stable or if symptoms have improved or worsened.

We reviewed the literature on similar interventions, finding that other pharmacist-led polypharmacy clinics have achieved similar results. Varas-Doval et al., measured the effect of monthly polypharmacy reviews on the number of controlled health problems by community pharmacists. The authors demonstrated a significant reduction in the number of uncontrolled health problems in the intervention group (–0.72, 95% CI: –0.80, –0.65) compared to no change in the control group (–0.03, 95% CI: –0.10, 0.04) in a randomized clinical trial26. The authors reported 1561 DRPs for 688 patients, amounting to 2.3 DRPs per patient. Although we identified the wrong dose for a medication or the omission of a medication as DRPs, we did not specifically include uncontrolled health problem as a DRP, suggesting a potential clinical outcome that could be measured in the polypharmacy clinic in the future.

Studies have demonstrated that acceptance rates by prescribing clinicians of pharmacist recommendations range between 40–75% for medication therapy management services provided by community pharmacists27. Similar polypharmacy interventions have reported acceptance rates by prescribing clinicians of 20–58%26,28. We achieved an acceptance rate of 65%. High acceptance rates highlight the critical establishment of relationships between physicians and pharmacists, fluid integration and workflow of pharmacists as part of the outpatient clinic(s), and support from all stakeholders to successfully implement multidisciplinary approaches to address polypharmacy. Other important considerations include careful integration of the pharmacist into the clinic workflow29. In a prospective study by Clark et al., in which a community-based clinical pharmacist was incorporated as part of a Medicare Annual Wellness Visit to make deprescribing recommendations for PIMs in older individuals being seen in a family medicine patient-centered medical home, workflow challenges such as lack of clinic staff knowledge of the roles and responsibilities about the pharmacist, changes in medication reconciliation from the nurse to the pharmacist, lack of EHR access by the clinical pharmacist, and the inability of pharmacists to enter notes into the EHR with their recommendations all hampered implementation28. One of the advantages of our polypharmacy clinic is that the pharmacist is employed by the health system and has access to the EHR. Additionally, as pharmacists are regularly integrated into the health system’s other programs (e.g., diabetes, hypertension, weight management programs), many physicians are used to working with pharmacists29.

An important consideration for the implementation of a polypharmacy clinic is that the referring physicians are aware of the potential harms of polypharmacy and the benefits of deprescribing. Geriatricians, who have additional training and often see complex patients with numerous medications, may be more aware of the potential harms of PIMs and polypharmacy, while physicians in other specialties may be less aware of drug-drug interactions or drug-disease interactions. The value of a polypharmacy clinic is increased when physicians discuss the potential harms of PIMs with patients before referring to the clinic, thereby increasing adoption by the patient and their caregivers. One model which has been tested in the Veterans Affairs system is the GeriPACT team, where a clinical pharmacist is embedded into an inter-disciplinary team. A study by Ammerman et al., evaluating this model found that 26.8% of PIMs were deprescribed compared to 16% in usual care30. While our study did not use an interdisciplinary team approach, pharmacists have been integrated into the health system and work in various chronic disease pharmacist-led clinics29,31. Thus, as physicians continue to gain awareness of the polypharmacy clinic, they may be more likely to refer patients for PIM reduction or potentially harmful polypharmacy. When physicians prioritize and approach deprescribing as they would with initiating clinically appropriate medication, the prevalence of older individuals with PIMs may be reduced5.

Our findings regarding medication list accuracy underscore the importance of medication review and follow-up frequency. Older patients have multiple specialists that may adjust medications on a regular basis; while these may be updated on the medical record if the health systems share EHRs, not all specialists are on the same EHR, increasing care fragmentation6. Furthermore, patients often self-adjust prescription medications as well as add over-the-counter medications and herbal supplements. To optimize patient care and evaluation, clinicians would ideally have access to accurate medication list to aid in proper clinical assessment and making appropriate medication changes1. For patients with high levels of medical complexity, in the absence of a national EHR, medication list accuracy can be achieved by ongoing thorough polypharmacy consults and follow up appointments to assess for variations and adherence.

We encountered several logistical and practical challenges. As a polypharmacy consult spans over multiple disease states, a collaborative practice agreement for this program has not been developed. All recommendations are communicated to primary care and referring provider(s) and any change in prescription therapy is communicated to the prescriber before implementation. Because of this process, there may be delays in medication changes or no response from prescriber that contributed to the 65% acceptance rate. However, the acceptance may reflect the fact that some of the recommendations to prescriber were to monitor for labs or symptoms that need to be addressed at follow-up visits. Some specialists are outside of the health system, increasing communication challenges. Time and resources need to be set up for appropriate follow up communications to these providers since these changes are not made at the time of polypharmacy visit.

We also identified a clear need for an efficient and systematic tool in our system to collect and track data for the polypharmacy program. Other clinical programs for specific disease states such as diabetes or hypertension have specific target goals such as hemoglobin A1c or blood pressure. However, there is little consensus in the literature on which medication-related outcomes are most valuable to track in the polypharmacy space. While tracking the number of medications, number of high-risk medications, number of drug-related problems and number of interventions is a start, reducing the number of medications may not the end-goal, or even clinically necessary. For example, it is difficult to demonstrate the benefit of tapering off a medication when the end result may not change in patient status, no adverse-drug events, or no falls. Bayless et al., from the U.S. Deprescribing Research Network have developed a set of outcome measurements to evaluate deprescribing studies which include meaningful medication outcomes, clinical and utilization outcomes resulting from medication discontinuation (adverse-drug events, minimizing fall risks, lowering healthcare costs, reducing hospital readmissions)32. These outcomes may be useful to track on a regular basis in a health system.

Policymakers in the U.S. have recently implemented measures to increase the detection and management of polypharmacy and PIMs, which strengthens the case for the implementation of polypharmacy clinics such as the one described in this study. The Centers for Medicare & Medicaid Services (CMS) published a final rule on April 12, 2023 which includes measures of polypharmacy, specifically the use of multiple anticholinergic and central nervous system-active medications in older adults, to the Medicare Advantage Star Ratings program33. CMS highlights the use of these high-risk medications in older adults as important areas of focus for the Medicare Advantage and Medicare Part D populations with the intent to encourage health plans and health systems to identify patients at high risk of adverse events and encourage appropriate prescribing33. To further support our polypharmacy-related work, we implemented an EHR-embedded polypharmacy note template to efficiently collect data instead of manual chart review. Additionally, we initiated an electronic build of a polypharmacy outcomes dashboard to establish our baseline performance, for example, tracking number of PIMs and DSAPs. We hope that the polypharmacy outcomes dashboard will identify potential gaps in care and guide our resource allocation to specific practice areas that may be in higher need for older adults across our organization.

We identified several limitations. First, this is a descriptive pre-post study examining medication outcomes. Patients were not randomly assigned to the intervention. A more rigorous and extensive study using random assignment may help to demonstrate a return on investment on an expensive and ongoing resource-intensive program such as a polypharmacy consult28. Second, we did not measure clinical outcomes such as falls, hospitalizations, and other medication-related adverse outcomes. We plan to assess such outcomes in the future. Furthermore, our evaluation period spanned only a few months due to the pragmatic approach of this evaluation, and thus resulted in a small sample size; future studies should track longer time frames and larger sample size to overcome any seasonal trends and biases. We were likely underpowered in our statistical analyses to see significant results. Finally, the results from our study may be more generalizable to health systems similar to those in the United States with already established physician and patient relationships with clinical ambulatory care pharmacists, and in which patients are mostly insured through Medicare.

Conclusion

Our descriptive evaluation demonstrates the value of a consult in improving medication list accuracy, identifying DRPs, and deprescribing PIMs. Evaluating polypharmacy in older patients requires clinical expertise and intensive ongoing efforts to be successful at not only identifying potentially appropriate and inappropriate medications based on general consensus, but also tailoring recommendations that are specific to the patient. While medication-related outcomes are descriptive surrogate indicators in evaluating pharmacist-led polypharmacy consults, more studies are needed to define clinical polypharmacy outcomes important to patients and providers.  . Establishing a causal connection between deprescribing and polypharmacy management and clinical outcomes and hospitalizations rates may be challenging to demonstrate; measuring the value of these polypharmacy consult services to physicians and patients may provide another perspective. For example, physicians may see reduced workloads and patients may benefit from having a greater understanding of their medications.