Medication Management Strategies at Hospital Admission: A Systematic Review and Meta‐Analysis Authors: Tara Emadi, Chin Hang Yiu, Shania Liu, Dihan Tang, Kingston Yeung, Melissa Baysari, Charley Budgeon, Stephen Carter, Deirdre Criddle, Cristen Fleming, Martina Francis, Charu Grover, Andrew Hawthorne, Deborah Hawthorne, Alexander Khlentzos, Jillian Khlentzos, Linda Krogh, Faye McMillan, Rebekah Moles, Kim Nguyen, Lachlan Oldfield, Anna Packer, Michelle Penm, Lily Pham, Katie Phillips, Simon Poon, Ganga Ratnanayagam, Debbie Rigby, Frank Sanfilippo, Sepehr Shakib, Edwin Tan, Bethany Van Dort, Paul Yates, Jerry Yik, Manya Angley, Rohan A Elliott, Jonathan Penm Published: 2026 ABSTRACT ABSTRACTHospital admission is high risk for medication harm due to transfer errors, delays, and charting mistakes. Up to 70% of admission medication lists contain errors, 59% potentially harmful, highlighting the need for safety strategies. The aim of the study is to systematically review the effectiveness of strategies for medication management at hospital admission. Following PRISMA guideline, three databases were searched for RCTs. Subgroup analyses assessed outcomes including mortality, readmissions/Emergency Department (ED) visits, length of stay (LOS), medication errors, and adverse drug events (ADEs), with meta‐analysis where possible. Nineteen RCTs were included. The strategies were classified into five categories: medication reconciliation, supported medication review, collaborative model of medication review, Partnered Pharmacist Medication Charting (PPMC), and real‐time health information exchange (HIE) tools. Medication reconciliation alone showed limited effectiveness in reducing medication errors and ADEs while combining with other strategies was more effective in minimizing ADEs. The meta‐analysis showed that PPMC can be considered a potential strategy to reduce medication errors and ADEs. Outcomes such as readmission, ED visits, mortality, and LOS were reported as secondary outcomes, and the strategies showed no significant effect in improving them. Overall admission strategies can reduce medication‐related outcomes such as medication errors and ADEs; however, stronger evidence is needed to confirm their impact on other outcomes. Future research should focus on integrating Clinical Decision Support (CDS) and stratification tools into electronic medical records to enhance the efficiency of these tools.Keywords: hospital admission, medication management, medication safety, transition of care Introduction The period of admission to hospital is high risk for medication errors and medication‐related harm [1]. Medication history errors have been reported to contribute to approximately 85% of all admission medication order errors [2]. Evidence comparing community pharmacy medication lists with hospitalized patients' lists has shown that 25% of prescription medicines in use at home were not recorded on the hospital admission record [3]. Medication discrepancies can cause patient harm and substantial clinical and economic consequences [4]. Medication errors arise from patient, clinician, and system factors. Patients may have limited knowledge, polypharmacy, or difficulty recalling medications [5, 6]. Clinicians may face incomplete information or communication issues, while system level problems include fragmented records and inefficient workflows [7, 8]. Effective and coordinated strategies are required for safe medication management at the time of hospital admission to prevent medication errors and improve patient clinical outcomes. While existing literature reviews have predominantly focused on the impact of medication reconciliation or medication review on reducing errors on admission to hospital [9, 10, 11, 12, 13, 14], the effectiveness of other strategies remains unclear. Examples of these strategies include partnered pharmacist medication charting (PPMC) [15] and the use of electronic medication management systems [16]. Therefore, this systematic review aimed to evaluate the effectiveness of all available strategies used to improve medication management in the period immediately following admission to hospital. This work extends a 2024 evidence briefing prepared for the Australian Commission on Safety and Quality in Health Care (ACSQHC) [17]. Method This systematic review was reported based on PRISMA guidelines [18]. The protocol was registered on PROSPERO (ID: CRD42024516488). This systematic review builds upon a prior rapid review and evidence briefing commissioned by ACSQHC, which was published in November 2024 [17]. The initial search for that review was conducted in 2023. The present review extends and updates that work by focusing exclusively on randomized controlled trials (RCTs) of medication management strategies at hospital admission. It includes updated database searches through to early 2026 and applies more rigorous methodological approaches, including formal risk of bias assessment and meta‐analysis. Search Strategy The literature search was conducted across three databases, PubMed, Embase, and Scopus, to ensure comprehensive coverage of relevant studies. PubMed and Embase were selected as core biomedical databases to capture clinical studies [19], whereas Scopus was included as a broader database to identify additional studies not indexed in the other sources [20]. This approach is supported by methodological evidence indicating that searching more than two databases reduces the risk of missing relevant studies, and that searching beyond PubMed can improve study identification. The details for the full search strategy are found in Table S1–S4 of the Figure S1. Inclusion and Exclusion Criteria This review included randomized controlled trials (RCTs) published between January 2013 and February 2026 that evaluated medication‐related strategies during hospital admission. Hospital admission was defined as the transition from home or residential care to a hospital inpatient setting or emergency department (ED) visit. Excluded were studies addressing both admission and discharge, non‐RCT designs, abstracts, case reports, reviews, and non‐English publications. Data Extraction Two authors (CHY and DT) screened articles independently by title and abstract for potentially eligible studies. After removing duplicates, studies were merged into Covidence (Veritas Health Innovation, Melbourne, Australia) [21]. These studies were then assessed in full text to confirm eligibility. Each article was reviewed by two authors during the title and abstract and full‐text screening process. Any discrepancies were discussed with a third author for consensus to be reached. Full texts were reviewed to confirm eligibility. Two authors (CHY and TE) screened titles, abstracts, and full texts, with a third resolving discrepancies. CHY and TE also extracted data on study characteristics, interventions, and outcomes, RE double checked the extracted data. Quality Assessment Two authors (CHY and TE) independently performed a quality assessment on all included RCTs. Version 2 of the Cochrane risk‐of‐bias tool for randomized trials (RoB 2) was used to assess quality of included studies [22]. Data Synthesis Meta‐aggregation was used to synthesize data on medication management strategies during hospital admission. Interventions and outcomes were categorized, heterogeneity assessed by study design and approach, and subgroup analyses conducted for each outcome. A meta‐analysis was conducted to examine the effect of interventions on medication errors. The included studies used different methods to report errors. Two key measures were identified: the total number of errors in medication orders and the number of patients who experienced one or more medication errors. Two studies were available for each outcome, and the meta‐analysis was performed using the R package “meta” [23]. Since the included studies had different populations, the random‐effects model was applied to adjust for variability while also considering the fixed‐effects model for comparison. Forest plots were created to visually display effect sizes and confidence intervals, and the I2 statistic was calculated to evaluate heterogeneity among the studies. Results The search resulted in 15,418 articles, and 14,624 articles were screened after removing duplicates. After screening by title and abstract, 1339 full‐text articles were assessed for eligibility based on the inclusion and exclusion criteria (Figure 1). In total, 19 studies were included in this review. PRISMA flow chart of the study selection. Study Characteristics All of the 19 included studies examined the transition from the community to either a hospital inpatient admission [24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38] (n = 16) or ED visit [39, 40, 41] (n = 3). The studies primarily focused on older adults (≥ 65 years) or those with polypharmacy. Australia had the highest number of studies (n = 4) [28, 30, 34, 40], followed by Denmark (n = 3) [24, 31, 33], with others conducted in Switzerland [26], Slovenia [29], Brazil [32], Ireland [27, 35], Korea [42], Colombia [39, 41], France [37], Pakistan [36], Norway [38] and the United States [25]. Sample sizes varied from small cohorts (n = 32) to larger populations (n = 1702). Among these studies, two pairs utilized the same population and methods but reported different outcomes. Becerra‐Camargo et al. (2013, 2015) [39, 41] involved the same 242 patients, and Gallagher et al. (2016) [27] and O'Sullivan et al. (2016) [35] studied 737 patients from the same sample. The characteristics of all included RCTs are summarized in Table 1. TABLE 1Summary of the medication management strategies at hospital admission; and characteristics and outcomes of included studies.Author, year, Country, sample size (n)Setting and populationDetails of interventionDetails of controlOutcomes and main findings Strategy: medication reconciliation only Ceschi et al. 2021 Switzerland n = 1702 [26]Patients ≥ 85 years, taking more than 10 medications at hospital admission.Pharmacist‐led medication reconciliation within 48 h of recruitment, based on a BPMH by pharmacy assistant from multiple sources including EMR.No pharmacist medication reconciliation. Home medications were recorded in an unstructured manner by a hospital physician. Primary outcome: Unplanned all‐cause ED visit in 90 days: No significant difference between the intervention and control group, (log‐rank p = 0.08). Unplanned all‐cause hospital readmissions in 90 days: No significant difference between the intervention and control group, (log‐rank p = 0.10). Secondary outcome: ADEs: No significant difference between the intervention and control group (p = 0.49). When only serious ADEs (p = 0.77). LOS: Median (IQR) was 8 (5–13) days in the intervention group and 8 (4–13) days in the control group (p = 0.23). Jost et al. 2022 Slovenia n = 120 [29]Patients admitted to general medical wards and taking at least one pre‐admission medication.Pharmacist‐led medication reconciliation between the 2nd and 5th day of hospitalization based on BPMH from multiple sources including EMR.Usual care (no pharmacist‐led medication reconciliation). Primary outcome: Medication discrepancies and ADEs: No significant difference between the intervention and control group in the number of pre‐admission medications with an unintentional discrepancy (15.2% vs. 17.9%) or that contributed to an in‐hospital ADE (19.7% vs. 21.0%)Mendes et al. 2016 Brazil n = 133 [32]Patients (≥ 18 years) newly admitted to the general wards.Pharmacist‐led medication reconciliation within 48 h of recruitment, based on BPMH following the Descriptive Elements of Pharmacist Interventions Characterization Tool (DEPICT).Medication history taken by nursing staff and medical residents with no medication reconciliation Primary outcome: LOS: The mean was 9 days (SD = 16) days in the intervention and 10 days (SD = 15) days in the control (p = 0.620). Strategy: Medication review supported by medication reconciliation, a risk stratification tool or a CDS Bonnerup et al. 2020 Denmark n = 369 [24]Patients (≥ 18 years) newly admitted to the general wards.Using a medication risk score (“MERIS”) to stratify patients into low and high‐risk groups; high risk patients received a medication review by a pharmacist or clinical pharmacologist on the day of admission, in addition to medication reconciliation. Although patient information obtained from EMR, MERIS was not incorporated to EMR.Usual care, including taking medication history and medication reconciliation by a physician. Primary outcome: Medication errors: No significant difference between the intervention and control group (p = 0.86). Secondary outcome: ED visits in 3 months: No significant difference between the intervention and control group (p = 0.38). All‐cause mortality in 3 months: No significant difference between the intervention and control group (p = 0.16) Briggs et al. 2015 Australia n = 1081 [40]Patients admitted to the ED who had been taking more than five medications daily.Using ISAR to stratify patients at risk of ED visits or readmission; ISAR positive patients received Medication review performed by pharmacist within the ED in terms of the Anticholinergic Burden Scale and the Beers criteria, in addition to medication reconciliation based on BPMH. However, the tools were not incorporated into EMR.Usual care with no medication review. Primary outcome: Admission from ED to ward: The intervention reduced the admission rate from the ED to an inpatient ward (53% vs. 62%, p = 0.003). Secondary outcome: LOS: The mean for both groups were 6 days (p = 0.87). ED visits in 3 months: No significant difference between the intervention and control group (p = 0.43) Admission to aged care home within 4 months: No significant difference between the intervention and control group (5.9% vs. 5% p = 0.37) O'Sullivan et al. 2016 Ireland n = 737 [35]Patients (≥ 65 years) admitted to the medical or surgical services.Pharmacist‐led medication reconciliation in 72 h of admission, followed by structured‐pharmacist review of medication, supported by START/STOPP CDS incorporated to EMR with written recommendations for the medical team.Usual care, including medication reconciliation and unstructured pharmacist medication review, usually within 72 h of admission. Primary outcome: Number of ADEs: The intervention significantly reduced the proportion of patients experiencing at least one ADE (13.9% vs. 20.7%; p = 0.02), with an ARR of 6.8% (95% CI: 1.5%–12.3%, p = 0.02) and an RRR of 33.3% (95% CI: 7.7%–51.7%). Secondary outcome: LOS: Median LOS was 9 days in the control group and 8 days in the intervention group, with no significant difference (p = 0.444). Mortality: 4.5% control, 4.7% intervention, no significant difference. Gallagher et al. 2016 Ireland n = 737 [27]Patients (≥ 65 years) admitted to the medical or surgical services.Pharmacist‐led medication reconciliation in 72 h of admission, followed by structured‐pharmacist review of medication, supported by START/STOP CDS incorporated to EMR with written recommendations for the medical team.Usual care, including medication reconciliation and unstructured pharmacist medication review, usually within 72 h of admission. Primary outcome: Number of ADEs: non‐significant decrease in the mean number of ADE events per patient of −0.064 (95% CI‐0.135 to 0.008; p = 0.081). Cost: Not cost‐effective Lee et al. 2023 Korea n = 32 [43]Patients (≥ 65 years) admitted to the Department of Hospital Medicine, who were taking at least five medications.Medication reconciliation by a clinical pharmacist within 24 h of admission, plus daily medication review based on Beers criteria and MRCI‐K criteria throughout the patient's hospital‐stay. Tools were not incorporated to EMR.Usual care without clinical pharmacy services. Primary outcome: ADEs in 30 days: the intervention led to a lower rate of ADEs (0% vs. 38.5%, p = 0.039). Secondary outcome: Medication appropriateness: No significant difference between the intervention and control group in change in PIMs score. Lisby et al. 2018 Denmark n = 108 [31]Patients (≥ 65 years) admitted who were non‐elective and treated with at least four medications at the time of hospital admission.Pharmacist‐led medication reconciliation within 24–48 h of admission and documentation on EMR, followed by medication review by the pharmacists and a clinical pharmacologist, with recommendations provided to the orthopedic physician.Usual care, without clinical pharmacy service. Primary outcome: Time to the first unplanned contact to a physician: The mean was 14.9 days in the intervention group vs. 27.3 days in the control group (log‐rank test, 0.05). Secondary outcome: Readmission, mortality in 3 months and LOS: No significant difference between the intervention and control group.Zafar et al. 2025, Pakistan n = 100 [36]Patients (≥ 18 years) with renal disease and admitted at the nephrology unit of hospital.Medication reconciliation and pharmacist review of all medications, including dose adjustments and drug interaction checks, conducted in a session of 15–20 min.Usual care, without clinical pharmacy service. Primary outcome: ADEs: 46.52% reduction comparing baseline and endpoint interventions. QoL: 58.64 ± 9.10 at the baseline to 74.48 ± 10.11 at the endpoint, p‐value < 0.001. Strategy: Collaborative medication review Leguelinel‐Blache et al. 2025 France n = 510 [37]Patients (≥ 65 years) admitted to an acute medical care unit.Pharmacist performs medication reconciliation and medication reviews and have a collaborative meeting with both the prescriber and the nurse.Usual care, medication without clinical pharmacy service. Primary outcome: Medication error: (patients with at least one error after intervention) control = 88.9% and intervention = 29.2% (p < 0.0001). The rate of patients with a 2–3 critical errors level decreased from 70.8% to 12.0% in the control vs. intervention periods respectively (OR: 18.4 [7.7;43.9], p < 0.0001).Santervas et al. 2026 Norway n = 137 [38] Patients (≥ 70 years) Using ≥ 6 regular medications admitted to an acute medical care unit. Specialists in general practice and geriatrics performed a medication review and discussed the findings with the patient's general practitioner.Usual care, medication review by one physician Primary outcome: QoL: no improvement, mean difference 0.009; 95% CI, −0.063 to 0.079;p = 0.8 Secondary outcome: Readmission, mortality in 16 weeks: No significant difference between the intervention and control group. Strategy: Partnered pharmacist medication charting and prescribing Becerra‐Camargo et al. 2013 and 2015 Colombia n = 242 [39, 41]Patients (≥ 18 years) admitted to an ED who had been taking at least one prescription. medication before admission.Pharmacists obtained medication histories in the ED before physician attendance, documenting them on an admission medication order form of EMR during first 24 h of admission, which was later verified by the doctor.Usual care with doctors documenting medication histories and nurses reviewing appropriateness of orders. Primary outcome: 2013 study: Medication discrepancies: The intervention reduced the proportion of patients having at least one home medication‐related discrepancy (60.7% vs. 93.6%, p < 0.0001). 2015 study: Potential adverse drug events (PADEs): Class 1 PADEs (unlikely to cause harm) (45% vs. 64%, p < 0.05), Class 2 PADEs (potential to cause moderate harm) (38% vs. 74%, p < 0.05), Class 3 PADEs (likely to result in severe harm) (31% vs. 56%, p < 0.05). Hale et al. 2013 Australia n = 384 [28]Patients admitted for elective surgery within a tertiary hospital.Pharmacist‐obtained BPMH and prescribing medications (regular medications and VTE prophylaxis) on the electronic medication chart in addition to duties as per usual care.Standard medication charting by medical officers with subsequent medication reconciliation performed by clinical pharmacist. Primary outcome: Medication errors: the intervention led to lower errors (1.2% vs. 31.5%, p < 0.001). Secondary outcome: VTE prophylaxis appropriateness: no significant difference. Khalil et al. 2016 Australia n = 110General medical patients (adult) admitted to the study ward.Pharmacist‐obtained BPMH and charting on the electronic medication administration record (MAR) after a medication management plan was signed off by the referring senior medical officer.Standard medication charting by medical officers with subsequent medication reconciliation performed by clinical pharmacist. Primary outcome: Number of medication errors per patient (0.52 vs. 4.41, p < 0.0001) Number of medication errors per medication (0.05 vs. 0.43, p < 0.05). Secondary outcome Total number of “Extreme and high” severity errors: 0 in intervention vs. 97 in control (p < 0.005). Number of VTE prophylaxis charted in the correct dose: 54 in intervention vs. 35 in control (p < 0.05) Nielsen et al. 2017 Denmark n = 593 [33]Medical patients (≥ 18 years) who were taking four or more medications daily.Pharmacist‐led reconciliation based on BPMH, review of pre‐admission medications, and entry of proposed prescriptions in the EMR, all before physician attendance.Standard hospital care with no clinical pharmacist involvement. Primary outcome: The proportion of patients with medication‐related harm: 10.8% vs. 17.1%, p = 0.05. Medication‐related harms per patient: 0.17 vs. 0.31, p = 0.02. Secondary outcome: Readmission, mortality and LOS: No significant difference between the intervention and control group. Tong et al. 2016 Australia n = 881 [34] Patients admitted to general medical units and emergency short‐stay units within a major referral hospital.Pharmacist‐obtained BPMH and charting of admission medications on patients' electronic medication charts after discussion of the medication management plan with the hospital medical officer.Standard medication charting by medical officers with subsequent medication reconciliation performed by clinical pharmacist within 24 h of admission. Primary outcome: Number of medication errors (15 vs. 372, p < 0.001). Number of patients who had zero errors (393 vs. 101, p < 0.001). Secondary outcome: Number of patients who had error classified as “High risk” (1 vs. 150, p = 0.01). Number (%) of patients who had error classified as “extreme risk” (0 vs. 25, p = 0.01) Strategy: Medication reconciliation using Health Information Exchange Boockvar et al. 2018 United States n = 381 [25]Patients admitted to one of four inpatient units.Using information from a Health Information Exchange (HIE) system to enhance medication reconciliation using a structured protocol.Medication reconciliation using a structured protocol without the support from information obtained from HIE. Primary outcome: Number of patients with identified risk‐weighted medication discrepancies: (8.0 vs. 5.9, p = 0 0.038). Number of risk‐weighted discrepancies: (6.4 vs. 5.8, p = 0.452), Discrepancy‐associated ADEs: (0.102 vs. 0.092 per admission, p = 0.964). Note: Colors are used only to separate different strategies used in the studies: Gray: medication reconciliation only; Blue: Medication review supported by medication reconciliation, a risk stratification tool or CDS; Pink: Collaborative Medication review; Green: Partnered pharmacist medication charting and prescribing; Orange: medication reconciliation using Health Information Exchange.Abbreviations: ADEs, adverse drug events; ARR, absolute risk reduction; BPMH, best possible medication history; CDS, clinical decision support software; ED, emergency department; EMR, electronic medical record; HIE, health information exchange; ISAR, identification of seniors at risk; LOS, length of stay; OR, odds ratio; PADEs, potential adverse drug events (defined as potentially harmful, unintentional medication discrepancies); PIMs, potentially inappropriate medications; QoL, quality of life; RRR, relative risk reduction; VTE, venous thromboembolism. Summary of the medication management strategies at hospital admission; and characteristics and outcomes of included studies. Primary outcome: Unplanned all‐cause ED visit in 90 days: No significant difference between the intervention and control group, (log‐rank p = 0.08). Unplanned all‐cause hospital readmissions in 90 days: No significant difference between the intervention and control group, (log‐rank p = 0.10). Secondary outcome: ADEs: No significant difference between the intervention and control group (p = 0.49). When only serious ADEs (p = 0.77). LOS: Median (IQR) was 8 (5–13) days in the intervention group and 8 (4–13) days in the control group (p = 0.23). Primary outcome: Medication errors: No significant difference between the intervention and control group (p = 0.86). Secondary outcome: ED visits in 3 months: No significant difference between the intervention and control group (p = 0.38). All‐cause mortality in 3 months: No significant difference between the intervention and control group (p = 0.16) Primary outcome: Admission from ED to ward: The intervention reduced the admission rate from the ED to an inpatient ward (53% vs. 62%, p = 0.003). Secondary outcome: LOS: The mean for both groups were 6 days (p = 0.87). ED visits in 3 months: No significant difference between the intervention and control group (p = 0.43) Admission to aged care home within 4 months: No significant difference between the intervention and control group (5.9% vs. 5% p = 0.37) Primary outcome: Number of ADEs: The intervention significantly reduced the proportion of patients experiencing at least one ADE (13.9% vs. 20.7%; p = 0.02), with an ARR of 6.8% (95% CI: 1.5%–12.3%, p = 0.02) and an RRR of 33.3% (95% CI: 7.7%–51.7%). Secondary outcome: LOS: Median LOS was 9 days in the control group and 8 days in the intervention group, with no significant difference (p = 0.444). Mortality: 4.5% control, 4.7% intervention, no significant difference. Primary outcome: Number of ADEs: non‐significant decrease in the mean number of ADE events per patient of −0.064 (95% CI‐0.135 to 0.008; p = 0.081). Cost: Not cost‐effective Primary outcome: ADEs in 30 days: the intervention led to a lower rate of ADEs (0% vs. 38.5%, p = 0.039). Secondary outcome: Medication appropriateness: No significant difference between the intervention and control group in change in PIMs score. Patients (≥ 70 years) Using ≥ 6 regular medications admitted to an acute medical care unit. Primary outcome: 2013 study: Medication discrepancies: The intervention reduced the proportion of patients having at least one home medication‐related discrepancy (60.7% vs. 93.6%, p < 0.0001). 2015 study: Potential adverse drug events (PADEs): Class 1 PADEs (unlikely to cause harm) (45% vs. 64%, p < 0.05), Class 2 PADEs (potential to cause moderate harm) (38% vs. 74%, p < 0.05), Class 3 PADEs (likely to result in severe harm) (31% vs. 56%, p < 0.05). Primary outcome: Medication errors: the intervention led to lower errors (1.2% vs. 31.5%, p < 0.001). Secondary outcome: VTE prophylaxis appropriateness: no significant difference. Primary outcome: Number of medication errors per patient (0.52 vs. 4.41, p < 0.0001) Number of medication errors per medication (0.05 vs. 0.43, p < 0.05). Secondary outcome Total number of “Extreme and high” severity errors: 0 in intervention vs. 97 in control (p < 0.005). Number of VTE prophylaxis charted in the correct dose: 54 in intervention vs. 35 in control (p < 0.05) Primary outcome: The proportion of patients with medication‐related harm: 10.8% vs. 17.1%, p = 0.05. Medication‐related harms per patient: 0.17 vs. 0.31, p = 0.02. Secondary outcome: Readmission, mortality and LOS: No significant difference between the intervention and control group. Tong et al. 2016 Australia n = 881 [34] Primary outcome: Number of medication errors (15 vs. 372, p < 0.001). Number of patients who had zero errors (393 vs. 101, p < 0.001). Secondary outcome: Number of patients who had error classified as “High risk” (1 vs. 150, p = 0.01). Number (%) of patients who had error classified as “extreme risk” (0 vs. 25, p = 0.01) Primary outcome: Number of patients with identified risk‐weighted medication discrepancies: (8.0 vs. 5.9, p = 0 0.038). Number of risk‐weighted discrepancies: (6.4 vs. 5.8, p = 0.452), Discrepancy‐associated ADEs: (0.102 vs. 0.092 per admission, p = 0.964). Note: Colors are used only to separate different strategies used in the studies: Gray: medication reconciliation only; Blue: Medication review supported by medication reconciliation, a risk stratification tool or CDS; Pink: Collaborative Medication review; Green: Partnered pharmacist medication charting and prescribing; Orange: medication reconciliation using Health Information Exchange. Abbreviations: ADEs, adverse drug events; ARR, absolute risk reduction; BPMH, best possible medication history; CDS, clinical decision support software; ED, emergency department; EMR, electronic medical record; HIE, health information exchange; ISAR, identification of seniors at risk; LOS, length of stay; OR, odds ratio; PADEs, potential adverse drug events (defined as potentially harmful, unintentional medication discrepancies); PIMs, potentially inappropriate medications; QoL, quality of life; RRR, relative risk reduction; VTE, venous thromboembolism. Medication Management Strategies at Admission The identified strategies were classified into five categories: 1. Medication reconciliation (n = 3) [26, 29, 32]: this strategy involved obtaining, verifying, and documenting the best possible medication history (BPMH) at hospital admission, after inpatient medication orders had been charted by a medical officer. 2. Supported Medication review (n = 7) [24, 27, 31, 35, 36, 40, 42]: this strategy involved performing a medication review that was supported by another medication management strategy, either medication reconciliation, CDS, or risk stratification tools. CDS for medication safety: Four CDS tools were used in included studies to optimize medication reviews by providing evidence‐based recommendations: Beers Criteria: a tool that identifies potentially inappropriate medications (PIMs) in older adults to minimize adverse drug events and improve medication safety [40, 42]. Beers Criteria: a tool that identifies potentially inappropriate medications (PIMs) in older adults to minimize adverse drug events and improve medication safety [40, 42]. Medication Regimen Complexity Index Korean version (MRCI‐K): it evaluates regimen complexity at admission, with a focus on variables such as the number of medications, antibiotic use, and duration [42]. Medication Regimen Complexity Index Korean version (MRCI‐K): it evaluates regimen complexity at admission, with a focus on variables such as the number of medications, antibiotic use, and duration [42]. Anticholinergic Burden Assessment: this evaluates the cumulative anticholinergic effects of prescribed medications, which are particularly relevant for older adults due to their association with cognitive impairment [40]. Anticholinergic Burden Assessment: this evaluates the cumulative anticholinergic effects of prescribed medications, which are particularly relevant for older adults due to their association with cognitive impairment [40]. STOPP/START Criteria: STOPP Criteria identify potentially harmful medications to avoid in older adults, while START Criteria highlight beneficial medications that are under‐prescribed [27, 35]. STOPP/START Criteria: STOPP Criteria identify potentially harmful medications to avoid in older adults, while START Criteria highlight beneficial medications that are under‐prescribed [27, 35]. Risk Stratification Tools to identify high‐risk patients: Two tools were used: Identification of Seniors at Risk (ISAR): a screening tool to identify older adults at risk of poor outcomes afterhospitalization, including readmissions, emergency visits, or functional decline, by assessing hospitalizations, medications, cognition, function, and support needs [40]. Identification of Seniors at Risk (ISAR): a screening tool to identify older adults at risk of poor outcomes afterhospitalization, including readmissions, emergency visits, or functional decline, by assessing hospitalizations, medications, cognition, function, and support needs [40]. Medication Error Risk Identification System (MERIS): a tool to identify patients at risk of medication errors by incorporating drug toxicity, potential for drug–drug interactions, the number of medications, and renal function [24]. Medication Error Risk Identification System (MERIS): a tool to identify patients at risk of medication errors by incorporating drug toxicity, potential for drug–drug interactions, the number of medications, and renal function [24]. 3. Partnered Pharmacist medication Charting (PPMC) ( n = 6) [28, 30, 33, 34, 39, 41]: pharmacist obtains a BPMH and charts the inpatient medication orders onto the medication administration record (MAR). The pharmacist and medical officer discuss and agree on the medication plan before the pharmacist charts the orders. This process is mainly known as PPMC in Australia [28, 30, 34]. Three non‐Australian studies [33, 39, 41] followed the same process but did not explicitly refer to it as PPMC. 4. Collaborative medication review (n = 2) [37, 38]: This model includes conducting medication reviews in collaboration with other healthcare professionals, including specialists in geriatrics, nurses, and the patient's general practitioner, to ensure the most appropriate medication for patients. 5. Medication reconciliation using a real‐time health information exchange (HIE) tool ( n = 1) [25]: The HIE tool is designed to integrate and consolidate data from multiple sources, including healthcare facilities, providers, and insurance plans. It provides pharmacists with real‐time access to a patient's medication history and may support medication reconciliation. Digitalization: Most of the included studies used electronic medical records (EMRs) for medication reconciliation and review. Exceptions include the study by Tong, which used a paper‐based system [34]; the studies by Khalil and Hale [28, 30], which pre‐date the use of EMRs; and the studies by Mendes and Lee, which did not specify the use of EMRs [32, 42]. The stratification tools and CDSs used in the studies were not integrated into the EMRs, except for the STOPP/START CDS [27, 35], which was incorporated to support pharmacists in making clinical decisions. Effectiveness of the Strategies Per Outcome Data from RCTs on strategies to improve medication management at admission were synthesized by outcomes, including medication errors, adverse events, post‐discharge mortality, readmission, ED visits, length of stay, and cost. The effectiveness of strategies for each outcome is summarized in Table 2. TABLE 2Summary of effectiveness of medication management strategies at admission.StrategiesPrescribing/charting errors on admissionAdverse drug events/reactionsMortalityHospital readmission/ED visitAdmission from ED to wardTime to first unplanned GP visitLength of stayMedication reconciliation only (3 studies) Not significant (1 study) Not reportedNot reported Not significant (1 study) No reportNo report Not significant (1 study) Supported medication review (6 studies) Not significant (1 study) Significant (3 studies) Not significant (2 studies) Not significant (3 studies) Significant (1 study) Significant (1 study) Not significant (2 studies) Partnered pharmacist medication charting (5 studies) Significant (4 studies) Significant (1 study) Not significant Not significant Not reportedNot reportedNot significantCollaborative medication review (2) Significant (1 study) Not reported Not significant Not significant Not reportedNot reportedNot reportedMedication reconciliation using Health Information Exchange digital tool (1 study) Significant (1 study) Not significant (1 study) Not reportedNot reportedNot reportedNot reportedNot reported Note: Green: significant; Red: not significant. Summary of effectiveness of medication management strategies at admission. Note: Green: significant; Red: not significant. Medication Errors Medication errors are defined as medication discrepancies, inappropriate medications, doses, routes, dosing schedules, or interactions between medications. Nine RCTs investigated medication errors as the primary outcome [24, 28, 29, 30, 34, 37, 39, 41]. All studies that implemented the PPMC strategy and measured medication errors reported a reduction in medication errors [28, 30, 34, 39, 41] while studies using medication reconciliation alone [29] or medication review supported by medication reconciliation [24], showed no impact. We performed two meta‐analyses of the PPMC studies, because they measured medication errors differently. The first focused on “medication errors in total medication orders”, while the second examined the “number of patients experiencing medication errors”. Meta‐analysis of medication errors in total medication orders: The meta‐analysis included two studies [28, 30] with a total of 3087 participants (1499 in the intervention group and 1588 in the control group), reporting 1163 events. As shown in Figure 2, the common effect (fixed‐effects) model showed a risk ratio (RR) of 0.31 (95% CI: 0.28–0.35, p < 0.001), indicating a 69% lower risk of the event in the intervention group compared with the control group. The random‐effects model yielded a similar direction of effect but with a wider confidence interval, showing an RR of 0.21 (95% CI: 0.06–0.76, p = 0.017), reflecting a 79% lower risk of error. Substantial heterogeneity was observed, with I2 = 97.7% (95% CI: 94.3%–99.0%), tau2 = 0.814, and an H index of 6.54 (95% CI: 4.18–10.21). The Cochran's Q test (Q = 42.73, p < 0.001) confirmed significant heterogeneity. Sensitivity analysis further explored the influence of individual studies, showing that omitting the study by Hale et al. significantly reduced the pooled effect (RR = 0.11, 95% CI: 0.08–0.16, p < 0.001) [28], while omitting Khalil increased the pooled effect (RR = 0.39, 95% CI: 0.35–0.44, p < 0.00) [30]. The pooled estimate including both studies remained significant (RR = 0.31, 95% CI: 0.28–0.35, p < 0.001), but the substantial heterogeneity persisted. Both studies contributed meaningfully to the overall effect, as shown by their significant influence when excluded individually. Meta‐analysis results of Partnered pharmacist medication charting vs. usual medication charting for Medication errors in total medication orders. Meta‐analysis of the number of patients experiencing medication errors: The meta‐analysis included two studies [30, 34] with a total of 991 participants (464 in the intervention group and 527 in the control group), reporting 470 events overall. As shown in Figure 3, the common‐effect (fixed‐effects) model estimated a RR of 0.11 (95% CI: 0.08–0.15, z = −14.92, p < 0.001), indicating an 89% reduction in the risk of events in the intervention group compared with the control group. However, the random‐effects model produced a RR of 0.16 (95% CI: 0.00–5.17, z = −1.04, p = 0.299), which was not statistically significant, reflecting the impact of high heterogeneity. The analysis revealed substantial heterogeneity with tau2 = 6.32, tau = 2.5140, I2 = 99.4% (95% CI: 98.8%–99.6%), and H = 12.53 (95% CI: 9.32–16.86). The heterogeneity test (Q = 157.07, p < 0.001) confirmed statistically significant differences between the studies. The sensitivity analysis under the common‐effect model further assessed the influence of individual studies by systematically omitting each study and recalculating the pooled effect estimate. Omitting Khalil [30] resulted in a pooled RR of 0.05 (95% CI: 0.03–0.08, p < 0.001), indicating a substantial reduction in the effect size and highlighting the strong influence of the Khalil 2016 study on the overall result. Conversely, omitting Tong 2019 increased the pooled RR to 0.52 (95% CI: 0.41–0.67, p < 0.001), suggesting that Tong 2019 contributed to lowering the overall pooled effect. The pooled estimate with both studies included remained significant (RR = 0.11, 95% CI: 0.08–0.15, p < 0.0001). Meta‐analysis results of Partnered pharmacist medication charting vs. usual medication charting for number of patients experiencing medication errors. Adverse Drug Events The adverse drug event (ADE) studies used varying terminologies, including medication‐related harm and adverse drug reactions. For the purpose of clearer analysis and comparison, all these terms are considered as ADEs in this review. Ten RCTs investigated ADEs as an outcome [25, 26, 27, 29, 33, 35, 36, 39, 41, 42]. All studies focused on ADE as a primary outcome, except the study by Ceschi et al. ADEs were measured either by counting the number of reactions or the percentage of patients who experienced at least one ADE. PPMC was successful in reducing ADEs in one study [39]. Medication reconciliation alone did not effectively reduce ADEs [26, 29], which may partly reflect the fact that Ceschi's study considered ADEs as a secondary outcome. Medication reconciliation using health information exchange (HIE) tools also failed to reduce ADEs [25]. Combining medication reconciliation with medication review showed some positive outcomes. Medication reconciliation within 24 h, combined with medication reviews using CDS tools (Beers criteria and MRCI‐K criteria), significantly reduced ADEs [42]. Two papers from one study evaluated strategies involving medication reconciliation within 72 h followed by medication review supported by STOPP/START CDS incorporated into EMR [27, 35]. Despite using the same dataset, these evaluations reported different findings due to variations in measures and statistical methods. In the primary paper, O'Sullivan et al. focused on clinical impact, measuring ADEs as the proportion of patients experiencing at least one ADE in the first 14 days ofhospitalization; this was the per‐protocol primary endpoint for the RCT. They reported a significant absolute risk reduction (ARR) of 6.8% (p = 0.02), a relative risk reduction of 33.3%, and a number‐needed‐to‐treat (NNT) of 15 [35]. Gallagher et al. evaluated cost‐effectiveness, measuring ADEs per patient. Fewer ADEs occurred in the intervention group (OR 0.655), but the reduction in mean ADEs (−0.064) was not statistically significant (p = 0.081) [27]. Mortality Rate Five RCTs examined post‐discharge mortality as a secondary outcome, measured at three and nine months after discharge [24, 31, 33, 35, 38]. The interventions included the use of the MERIS algorithm to prioritize medication reviews [24], pharmacist‐led medication reconciliation within 24 h of admission [31], pharmacist‐led medication reconciliation and review supported by STOPP/START CDS [35], collaborated medication review and PPMC [33]. None of the studies showed a significant difference in post‐discharge mortality. Hospital Readmission and ED Visit Six RCTs assessed hospital readmissions or ED visits as outcomes [24, 26, 31, 33, 38, 40]. The most effective intervention combined medication reconciliation with an ISAR stratification tool to prioritize patients for medication review and utilized CDS (Anticholinergic Burden Scale and the Beers criteria) for conducting the review. This approach significantly reduced the absolute hospital admission rate from the ED to an inpatient ward by 9% (p = 0.003). However, the primary outcome focused on whether patients reviewed by a pharmacist were admitted to a ward rather than hospital readmissions after discharge. There was no significant difference in subsequent ED visits within three months (p = 0.43) as the secondary outcome [40]. Only the study by Ceschi et al. [26] examined both hospital readmissions and ED visits as primary outcomes. This study used pharmacist‐led medication reconciliation within 48 h of admission but found no significant difference in 90‐day ED visits or readmissions. Other studies assessing these outcomes were likely underpowered [24, 31, 33]. The time to the first general practitioner (GP) contact: This outcome was investigated in only one study, Lisby et al. which utilized medication review supported by medication reconciliation [31]. The intervention significantly reduced the mean time to the first unplanned physician contact (14.9 days vs. 27.3 days; p = 0.05). This key outcome for pharmacist‐led interventions ensures timely follow‐up on medication changes, prevents adverse events, and improves continuity of care, especially in high‐risk patients [31]. Length of Stay (LOS) Five studies assessed length of stay (LOS) as an outcome, primarily as a secondary measure, with mixed findings. Two studies [32, 35] observed a one‐day reduction in mean and median LOS in the intervention groups compared with the control groups, although the differences were not statistically significant (p = 0.620 and p = 0.444, respectively) [32, 35]. Despite the lack of statistical significance, these results suggest a positive trend, indicating that pharmacist‐led medication reconciliation and reviews may contribute to slightly shorter hospital stays. In contrast, the other three studies [26, 31, 40] reported no meaningful differences in LOS between groups. Risk of Bias Assessment Of the 19 RCTs assessed, three studies were noted for consistently low risk of bias across all domains [36, 39, 41]. In three studies, there were “some concerns” observed primarily related to biases in randomization selection, concealment, adherence to intervention [28, 32, 38]. Thirteen studies were identified as having a “high risk of bias” in at least 1 domain, particularly due to issues concerning selection bias regarding adherence to intervention and reporting, along with detection and randomization biases [24, 25, 26, 27, 29, 30, 31, 33, 34, 37, 40, 43, 44]. Figure S1 shows the risk of bias summary, with detailed domain‐specific assessments using the RoB 2.0 tool for RCTs provided in Table S5. Discussion This review suggests that pharmacists' involvement in charting of medications at hospital admission, also known as PPMC, may be a more effective medication management strategy for reducing the risk of medication errors at hospital admission compared with medication reconciliation or supported medication review. PPMC appears to reduce medication errors and may also contribute to minimizing harm and adverse effects associated with these errors. The findings from identified RCTs [28, 30, 33, 34, 39, 41] align with findings from pre–post intervention studies [15, 45] describing the effect of PPMC in reducing medication‐related outcomes such as medication errors and adverse effects. Although included RCTs did not investigate LOS as a primary outcome, previous pre–post studies suggested that the PPMC model also reduces LOS as a primary outcome [15]. The term PPMC is used in Australia and the current evidence base for PPMC primarily comes from studies conducted in Australia [15, 28, 30, 34, 45]. Two other studies by Nielsen and Becerra also examined this strategy; however, they did not refer to it as PPMC [33, 41]. This highlights the need for broader international recognition of this approach as a possible effective medication management strategy at hospital admission. Supported medication review with CDS tools has been identified to have the potential to reduce ADEs [46]. Previous systematic reviews have found moderate certainty evidence that CDS can improve prescribing practices, particularly by reducing inappropriate medications and supporting deprescribing; however, their impact on clinical outcomes remains inconsistent. While these systems can improve process measures, challenges such as alert fatigue, poor usability, and limited interoperability may reduce their real‐world effectiveness [47, 48]. Stratification tools are also another form of CDS that facilitate clinical decision making by early risk detection, improve resource allocation and enhance patient outcomes [49]. Effective communication within multidisciplinary teams is essential to improve the process of medication review and to facilitate the implementation of recommended changes to patients' medications. Collaborative models of medication review, supported by regular multidisciplinary meetings, can enhance prescribing processes and reduce medication errors; however, their impact on clinical outcomes remains uncertain. Similar findings have been reported in multidisciplinary care models in chronic disease, where improvements in medication management do not consistently translate into reductions in hospitalizations or mortality [50]. This review has some limitations. Meta analysis was conducted with a small number of studies (n = 2) due to variation in study methodology and outcome measures. High heterogeneity was identified across the meta‐analyses. Although each PPMC study showed a reduction in outcomes, the pooled estimates should be interpreted with caution and should be regarded as exploratory and hypothesis‐generating, rather than conclusive. Also, PPMC is geographically concentrated, with most trials conducted in Australia. Although these studies offer important insights, their generalizability to other health systems and models of care may be limited. More studies are needed to improve the reliability of the estimates and reduce heterogeneity. Most included studies involved broad or mixed patient populations; however, one study focused exclusively on pre‐admission surgical patients. This may limit the extrapolation of findings to specific clinical settings, and further research is needed to evaluate the effectiveness of these strategies in more defined populations. Furthermore, a high proportion of the included randomized controlled trials were assessed as having high risk of bias, which reduces confidence in the overall findings and limits the strength of the conclusions. Also, mortality and readmission were secondary outcomes in studies and may have been underpowered, and the high risk of bias may have further affected these results. Furthermore, CDS and stratification tools were rarely integrated into EMRs, and most PPMC evidence came from Australia, limiting broader applicability. Author Contributions T.E., C.H.Y., S.L., D.T., K.Y., M.B., C.B., S.C., D.C., C.F., M.F., C.G., A.H., D.H., A.K., J.K., L.K., F.M., R.M., K.N., L.O., A.P., M.P., L.P., K.P., S.P., G.R., D.R., F.S., S.S., E.T., B.V.D., P.Y., J.Y., M.A., R.A.E., and J.P. wrote the manuscript. T.E., S.L., M.B., S.C., R.M., B.V.D., M.A., R.A.E., and J.P. designed the research; T.E., C.H.Y, D.T., K.Y., M.F., C.G., L.O., L.P., R.A.E. and J.P. performed the research; T.E., C.H.Y., and R.A.E. analyzed the data. Funding The authors have nothing to report. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: cts70586‐sup‐0001‐Supinfo.docx. Risk of bias assessment of the included studies using ROB2. Table S1: Keywords of search strategy. Table S2: PubMed Search strategy and keywords. Table S3: Scopus Search strategy and keywords. Table S4: Embase Search strategy and keywords. Table S5: Quality assessment summary of the included RCTs using ROB2 tool. Acknowledgments The authors acknowledge the use of OpenAI‘s ChatGPT‐5 (https://openai.com) for assistance with language editing and improving the clarity of the manuscript. Open access publishing facilitated by The University of Sydney, as part of the Wiley ‐ The University of Sydney agreement via the Council of Australasian University Librarians. Contributor Information Lachlan Oldfield, Email: l.oldfield@unsw.edu.au. Jonathan Penm, Email: jonathan.penm@sydney.edu.au. References Associated Data Supplementary Materials Figure S1: cts70586‐sup‐0001‐Supinfo.docx. Risk of bias assessment of the included studies using ROB2. Table S1: Keywords of search strategy. Table S2: PubMed Search strategy and keywords. Table S3: Scopus Search strategy and keywords. Table S4: Embase Search strategy and keywords. Table S5: Quality assessment summary of the included RCTs using ROB2 tool.