Older adults in long-term hospital care were prescribed fewer medications after their hospital introduced a stepwise program of pharmacist-led polypharmacy management, according to a retrospective observational study published in the journal Drugs – Real World Outcomes. The average number of regularly prescribed oral medications fell from 6.44 in 2023 to 6.05 in 2025, while the share of patients taking six or more medications declined from 54.3 percent to 46.5 percent over the same period. Yet the reductions in medication counts did not translate into lower total medication costs, which remained statistically unchanged across the three years of the study.
The research was conducted at Omi Fureai Hospital, a chronic care hospital in Japan that provides long-term inpatient care to older adults. A team led by Takefumi Miyake analyzed routinely collected electronic medical record data from all inpatients hospitalized between January 2023 and December 2025, dividing the study period into three phases: a preintervention year in 2023, an implementation phase in 2024, and a comprehensive management phase in 2025. Approximately 320 patients were hospitalized each month throughout the study, and the population consisted primarily of older adults, with mean ages of roughly 79 years for men and 84 years for women and a slightly higher proportion of women than men.
The intervention unfolded in two stages. In January 2024, the hospital launched a multidisciplinary polypharmacy management program in which nurses, pharmacists, and physicians collaboratively reviewed medication regimens. Ward nurses identified patients receiving ten or more regularly prescribed medications during routine prescription checks and submitted medication review forms to the pharmacy department. Pharmacists then examined prescription records and medical charts, including diagnoses, medical history, laboratory results, and medications taken at admission, looking for duplicate prescriptions, unclear indications, drug interactions, and potential adverse effects. When changes were considered appropriate, physicians and pharmacists explained the proposed modifications to patients and their families before discontinuation, dose reduction, or substitution was carried out.
Beginning in January 2025, the program expanded into pharmacist-led comprehensive medication evaluation and management. Pharmacists no longer waited for nurse referrals; instead, they proactively identified eligible patients using prescription information and medical records, and continuously reassessed medication regimens during hospitalization, particularly when new medications were prescribed or a patient’s clinical condition changed. The ten-medication eligibility threshold remained unchanged between the two intervention phases, allowing the researchers to compare the effects of the two operating models against the 2023 baseline.
The primary outcome was the monthly average number of regularly prescribed oral medications, with topical medications, injectables, as-needed medications, and temporary prescriptions excluded from the analysis. The mean medication count declined from 6.44 plus or minus 3.37 in 2023 to 6.23 plus or minus 3.52 in 2024 and 6.05 plus or minus 3.47 in 2025, with the median falling from 6.0 to 5.0 by the final year. Because patients hospitalized across multiple months contributed repeated observations that were treated as independent in the primary analysis, the researchers also ran sensitivity analyses using linear mixed-effects models with patient identifier as a random effect. These models confirmed that the mean number of prescribed medications was significantly lower in both 2024 and 2025 than in the preintervention year, with P values of 0.002 and less than 0.001 respectively.
Prevalence measures told a consistent story. The proportion of patients receiving six or more medications fell from 54.3 percent in 2023 to 50.5 percent in 2024 and 46.5 percent in 2025, while the proportion receiving ten or more medications declined from 18.4 percent to 16.8 percent and then 15.5 percent. The six-medication threshold corresponds to the criterion for polypharmacy management in the Japanese national healthcare reimbursement system, whereas the ten-medication threshold reflects the hospital’s own institutional criterion for identifying patients eligible for pharmacist-led comprehensive medication review.
Despite the shrinking medication burden, total annual medication costs showed only small fluctuations: 29.4 million yen in 2023, 27.5 million yen in 2024, and 28.4 million yen in 2025, with no statistically significant differences across the three periods. Costs were calculated using the Japanese National Drug Price Standard issued by the Ministry of Health, Labour, and Welfare. The therapeutic-class analysis offered a partial explanation. Costs decreased for hypnotics and anxiolytics, nonsteroidal antiinflammatory drugs and paracetamol, antihypertensive drugs, and medications for constipation, while antithrombotic drug costs increased over the study period. The authors suggest that medications targeted for deprescribing were often relatively low-cost drugs such as sedatives and symptomatic treatments, that essential medications for chronic disease management were maintained or newly initiated when clinically indicated, and that shifts toward newer therapeutic agents may have offset the savings from deprescribing. The rising antithrombotic costs, they note, may reflect the increased use of direct oral anticoagulants in clinical practice.
Prescription volume, measured as defined daily dose per 100 patients per day using the World Health Organization Anatomical Therapeutic Chemical/DDD classification system, remained generally stable for most therapeutic categories. Antihypertensive drugs showed the highest prescription volume at roughly 3.5, followed by constipation medications and antithrombotic agents at about 2.0, and blood glucose-lowering drugs at approximately 1.0. Analysis of variance comparing prescription volumes across the three study periods found no significant difference in 2024, but a significant difference emerged in 2025, with a P value of 0.009. The authors caution that this measure is derived from standardized daily doses and may not fully reflect actual prescribed doses in frail older hospitalized patients, where dose reductions due to renal impairment, advanced age, or deprescribing strategies can create discrepancies between prescribed doses and DDD-based estimates.
The study’s authors are careful about what the findings do and do not show. A decrease in medication count, they write, does not necessarily indicate improved prescribing appropriateness or higher-quality pharmacotherapy, because optimization depends on whether each medication is clinically indicated, effective, and safe for the individual patient. The retrospective, single-center design precludes definitive conclusions about causality, and potential confounding factors such as disease severity, comorbidities, and case-mix differences were not fully adjusted for. The study also did not assess clinical outcomes such as adverse drug events, hospitalization, falls, or mortality, and prescribing appropriateness was not evaluated with validated measures such as potentially inappropriate medications or the STOPP/START criteria. Pharmacist recommendations, acceptance rates, and the reasons for individual medication changes were not systematically evaluated, and the aggregated prescription data could not reveal which individual medications were most frequently discontinued or modified.
Nevertheless, the researchers argue that the results carry practical significance for chronic care settings, where patients often experience prolonged hospitalization and complex multimorbidity, and where medications initiated during acute care may be continued without sufficient reassessment. The class-level patterns suggest a selective approach to medication management rather than uniform discontinuation: drugs frequently associated with adverse effects in older adults, such as hypnotics and antiinflammatory agents, showed declining costs, while medications considered essential for cardiovascular risk management were largely maintained. The authors conclude that evaluating deprescribing strategies requires multiple indicators, including medication count, prescription volume, and therapeutic class, since reductions in the number of prescribed medications were not necessarily accompanied by a substantial decrease in overall medication exposure. They call for multicenter prospective studies incorporating validated measures of prescribing appropriateness, detailed assessment of pharmacist interventions, and clinically relevant outcomes such as adverse drug events, functional status, and healthcare utilization.
Subject of Research: Pharmacist-led polypharmacy management and prescribing patterns among older inpatients in a chronic care hospital
Article Title: Impact of Pharmacist-Led Polypharmacy Management on Prescribing Patterns in Older Inpatients: A Real-World Study in a Chronic Care Hospital
Article References: Miyake, T., Nakamura, Y., Masui, K., Iwasaki, H., Nakamura, F., Hashimoto, M., & Kadowaki, D. (2026). Impact of Pharmacist-Led Polypharmacy Management on Prescribing Patterns in Older Inpatients: A Real-World Study in a Chronic Care Hospital. Drugs – Real World Outcomes. https://doi.org/10.1007/s40801-026-00584-4
Image Credits: AI Generated
DOI: 10.1007/s40801-026-00584-4
Keywords: polypharmacy, deprescribing, pharmacist-led medication review, older inpatients, chronic care hospital, prescribing patterns, medication costs, defined daily dose, Japan, retrospective observational study, geriatric pharmacotherapy, medication burden
News Source: Beatrice Stafford. (October 28, 2026). Pharmacist-Led Medication Reviews Trim Prescriptions in Older Hospital Patients. Scienmag.
