https://orcid.org/0000-0003-4841-0861
Abstract
Although pharmacist-led interventions in anticoagulant (AC) therapy are widely accepted, there is a lack of evidence comparing their effectiveness with usual care in terms of AC therapy appropriateness and clinical outcomes. We aimed to estimate the comparative effectiveness of pharmacist-led interventions on the appropriateness and clinical outcomes of AC therapy.
Adhering to the PRISMA guidelines, we searched PubMed, EMBASE, and Scopus databases to identify randomized controlled trials and quasi-experimental and cohort studies published between 2010 and 2023. A random-effects model was used to calculate pooled intervention effects. We assessed heterogeneity (using Higgins’ I2 and Cochran's Q) and publication bias (using Egger's test, the trim-and-fill method, and visualization of the funnel plot). In total, 35 studies involving 10 374 patients in the intervention groups and 11 840 in the control groups were included. The pharmacist-led interventions significantly improved the appropriateness of AC therapy [odds ratio (OR): 3.43, 95% confidence interval (CI): 2.33–5.06, P < 0.01]. They significantly decreased total bleeding [relative risk (RR): 0.75, 95% CI: 0.58–0.96, P = 0.03) and hospitalization or readmission (RR: 0.64, 95% CI: 0.41–0.99, P = 0.04). However, the impact of the pharmacist-led interventions on thromboembolic events (RR: 0.69, 95% CI: 0.46–1.02, P = 0.07) and mortality (RR: 0.76, 95% CI: 0.51–1.13, P = 0.17) was not significant.
Pharmacist-led interventions demonstrated superior outcomes in optimizing AC therapy compared with usual care. Further research is needed to evaluate pharmacist-led interventions’ cost-effectiveness and long-term sustainability.
PROSPERO registration number: CRD42023487362.
What is already known
Anticoagulants are high-risk medications associated with serious adverse drug events, even when adhering to clinical guidelines.
What this study adds
Our systematic review and meta-analysis, the first of its kind, evaluate pharmacy-led interventions’ effectiveness on anticoagulant therapy appropriateness. Our findings show that patients under pharmacist-led interventions were over three times more likely to receive appropriate anticoagulant therapy.
In our meta-analysis, we found pharmacist-led interventions were also significantly associated with reduced total bleeding and hospitalizations or readmission.
This study provides evidence for the potential implementation of pharmacist-led interventions for improving the appropriateness of anticoagulant therapy and mitigating the risk of bleeding and hospitalizations.
Introduction
Anticoagulant (AC) therapy is commonly used in patients with chronic atrial fibrillation (AF) or at risk of or experiencing venous thromboembolism (VTE).1–4 Despite the clear benefits of AC therapy in reducing and treating thrombotic events, ACs are classified as high-risk medications.5,6 This is because they are associated with serious adverse drug events (ADEs), even when used in accordance with clinical practice guidelines.7–9 Bleeding is the main ADE with AC treatment10 and is also a cause of AC discontinuation in up to 50% of patients.11 Patients taking oral anticoagulants (OACs) have up to an eight-fold increase in the risk of intracranial bleeding.12,13
Previous reviews have found that 8–29% of patients with indications for AC therapy receive inappropriate direct-acting oral anticoagulant (DOAC) prescriptions.14,15 Moreover, the prescribing of OACs for stroke prevention in patients with AF often does not comply with guidelines, resulting in underuse in high-risk patients and overuse in low-risk ones.16–19 The complexity of AC therapy has led to the development of pharmacist-led services to enhance patient outcomes through monitoring, dosage adjustment, and early identification of bleeding and VTE risk factors.20,21 Pharmacist-led interventions may improve medication use outcomes and patients’ quality of life22–24 and prevent medication-related harm.22,25,26
However, to our knowledge, no recent comprehensive reviews have evaluated the impact of pharmacist-led interventions on AC therapy appropriateness. Moreover, previous reviews on bleeding and thromboembolic events have been inconsistent in their findings and based on studies that predominantly included patients treated with warfarin.27,28 For example, Zhou et al.’s28 review revealed no significant difference in bleeding and thromboembolic events between pharmacist-managed care and usual care; however, reviews by Hou et al.27 and Lee et al.25 found that pharmacist-led AC therapy compared with usual care reduced bleeding risk and mortality, respectively. Notably, these reviews had important limitations, such as the small number of participants in most included studies.25 Furthermore, their findings predominantly reflected the pre-DOAC era, with some post-DOAC studies that also focused solely on patients managed with warfarin,27,28 despite warfarin being largely replaced by DOACs for stroke and VTE prevention.29 Therefore, this systematic review aimed to evaluate the effect of pharmacist-led interventions on the appropriateness and clinical outcomes of AC therapy, focusing especially on the period following the introduction of DOACs into practice.
Methods
Protocol
The protocol is available at PROSPERO (CRD42023487362).30 The findings are reported following the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement.31
Data sources and search strategy
We searched PubMed, EMBASE, and Scopus for articles utilizing data since 2010 and published between 1 January 2010 and 24 November 2023 (updated on 23 December 2023). This time frame (2010 onwards) was chosen to focus on the post-DOAC era. The search strategy (developed in consultation with a research librarian) combined MeSH terms and keywords related to ACs, pharmacist-led intervention, AF, and VTE, as outlined in Supplement 1. We performed citation analysis by searching Google Scholar and the reference lists of the included studies. Study authors were contacted for additional data where necessary.
Study selection criteria
All identified records were imported into Covidence software, where duplicates were removed. Two authors (B.K. and W.B.) independently reviewed each title and abstract; discrepancies were resolved with a third author (G.P.) by discussion and adjudication. All studies marked for possible inclusion underwent independent full-text review by two reviewers (B.K. reviewed all articles and W.B. and G.P. half each). Exclusion reasons were documented.
Inclusion criteria
We included randomized controlled trials (RCTs), quasi-experimental (pre/post-intervention), and cohort studies that examined the effectiveness of pharmacist-led interventions compared with usual or physician-led care (both hereafter referred to as UC) in adults (aged ≥18 years) with AF or at risk of or having experienced VTE. Pharmacist-led interventions were defined as those in which pharmacists led an intervention to enhance AC therapy appropriateness and safety.
The primary outcomes of interest were as follows: (i) AC therapy appropriateness, evaluated based on evidence-based guidelines [e.g. American College of Clinical Pharmacy,32,33 summaries of product characteristics,34–36 country-specific guidelines,37,38 pharmacy-driven local protocol/guidelines39–44], and (ii) clinical outcomes, including thromboembolic events [systemic embolism, stroke, transient ischaemic attack, and myocardial infarction], bleeding events (major bleeding, minor or clinically significant non-major bleeding), and mortality. Hospitalization or readmission to hospital was a secondary outcome of interest.
Exclusion criteria
We excluded studies focused on pharmacists’ perceptions, attitudes, views, knowledge, and awareness towards AC therapy; studies not available in English; studies with interventions not predominantly driven by a pharmacist; and theses, editorials, commentaries, conference abstracts, opinions, study protocols, guidelines, and qualitative studies.
Quality assessment
We used the JBI Critical Appraisal Checklists to assess the quality of RCTs, quasi-experimental studies,45 and cohort studies.46 Two authors performed the appraisal (B.K. appraised all studies, and W.B., G.P., and C.M. appraised one-third each), and a third author resolved conflicts. The percentage of positively (yes) answered questions was calculated for each included study. The quality of evidence was ranked as high (score >70%), moderate (score 50–70%), or low (score <50%).47
Data extraction
One author (B.K.) extracted data (using a standardized data extraction form), which was independently checked by a second author (W.B.); discrepancies were resolved via consensus. We extracted the author's name, country, publication year, study design and setting, participant characteristics, a description of the pharmacists’ intervention, outcome measures (e.g. number of patients with appropriate AC therapy), the type of ACs used and their indications, and a summary of key findings.
Data synthesis and statistical analysis
Characteristics of patients and studies were summarized using descriptive statistics. We performed a meta-analysis for all outcomes using a random-effects model48 with an odds ratio (OR) or relative risk (RR) with 95% confidence intervals (CIs). Effect estimates were reported as ORs to compare the pooled appropriateness of AC therapy and as RRs to compare pooled events of clinical outcomes (bleeding, thromboembolic, mortality, and hospitalization or readmission) between the pharmacy-led and control groups. Han et al.44 reported two separate outcomes: the appropriateness of (i) AC discontinuation before endoscopy and (ii) AC bridging therapy. In our meta-analysis, we handled these as two independent studies. Any data not suitable for meta-analysis were narratively described.
Heterogeneity among studies was assessed using Higgins’ I2 and Cochran's Q.49 Using the I2 scale, heterogeneity is rated as low (0–25%), moderate (25–50%), substantial (50–75%), or considerable (>75%).50 Publication bias was assessed for all outcomes except mortality (as the number of studies was < 10) using Egger's test, the trim-and-fill method, and visualization of the funnel plot.51,52 We assessed the potential sources of heterogeneity using subgroup analyses (based on study design, region, and severity of bleeding) and leave-one-out analyses. The leave-one-out sensitivity analyses were performed by sequentially excluding each study to check its effect on the overall pooled OR or RR estimates for each outcome (AC appropriateness, bleeding events, thromboembolic events, and hospitalization or readmission). The statistical significance cut-off was P < 0.05. Analysis was conducted with STATA (Version 18.0; StataCorp, College Station, TX, USA).
Results
Study selection
We identified 2255 articles from the databases, of which 1833 unique studies were screened. Among these, 35 studies were included in the qualitative synthesis and meta-analysis. Figure 1 summarizes the search results and reasons for exclusion.
PRISMA flowchart of the study selection process and reasons for study exclusion.
Studies and participants’ characteristics
The 35 studies included 10 374 individual patients in the intervention group (range: 17–4939) and 11 840 in the control group (range: 17–6182). The study settings were mainly hospitals [29 (83%) studies],32,33,36–43,53–71 and the participants’ mean ages ranged from 42 to 86 years. Study designs were varied: 5 were RCTs,55,57,61,63,72 16 were quasi-experimental (pre/post-intervention),32–34,36–40,42,53,54,56,58–60,62 and 14 were cohort studies.35,41,43,44,64–71,73,74 Fourteen were conducted in Asia (5 in China,43,56,57,63,64 3 in Japan,65–67 2 in India,32,61 and one each in Taiwan,54 Malaysia,40 Saudi Arabia,69 and Lebanon55); 11 in North America (all in the USA)35,39,41,44,59,62,68,70,71,73,74; 6 in Europe (Belgium,60 Croatia,72 England,34 France,36 Malta,42 and Turkey33); 3 in Australia37,38,58; and 1 in Africa (Sudan).53 ACs were prescribed mainly for AF, VTE treatment, and VTE prophylaxis in patients undergoing various types of surgeries (see Supplementary material online, Table S1). Eleven studies34,36,38,41,42,55–57,60,61,71 included patients prescribed DOACs or warfarin, 10 studies53,54,58,63,65,66,68–70,72 warfarin alone, 6 studies35,39,62,64,73,74 DOACs, 4 studies32,37,40,43 parenteral ACs, and 4 studies33,44,59,67 any antithrombotic (including AC and/or antiplatelet) medications (used to prevent post-surgical VTE in patients who underwent surgery and stroke and other cardiac complications in those with AF and ischaemic heart disease).
The types of pharmacist-led interventions most commonly used were patient education or counselling,32–34,37,38,41,53–55,57,61–63,69,72–74 medication chart or patient review,34,35,37,39,41,58,59,73,74 bleeding, stroke, or VTE risk assessment,34,40,43,56,59 the development and implementation of pharmacist-driven protocols or guidelines,32,39,40,42,44,56,60,62,73 medication reconciliation,56,59 staff training,36,56 and providing alternative AC therapy or dose recommendations.33,39,59,63,69 Their impact in improving the appropriateness of AC therapy was examined in 14 studies.32–44,60 Their impact on clinical outcomes was addressed as follows: bleeding in 19 studies,33,43,53–58,61,62,64–69,71,72,74 thromboembolic events in 14 studies,33,43,56–59,63,64,67–69,71,73,74 readmission or hospitalization in 10 studies,41,53,55,56,58,63,68–70,73 and mortality in 6 studies43,55,58,67,71,74 (Table 1).
Characteristics of the included studies in the review
| Study, year and country | Study design and settings | Comparison groups | Sample size | Follow-up time (months) | Drugs used | Specific intervention(s) given | Key findings (control vs. intervention) |
|---|---|---|---|---|---|---|---|
| Falamić et al. 2019,72 Croatia | RCT, Community pharmacy | Control group Intervention group | 66 65 | 6 | Warfarin | • Patient education • Optimizing warfarin dose and contacting physicians to avoid drug interactions | • ↓ Bleeding (from 85% to 29%; P < 0.05) |
| Karaoui et al. 2021,55 Lebanon | RCT, Hospital | Standard of care Pharmacist-counsel | 100 100 | 1 | Apixaban Dabigatran Rivaroxaban VKA | • Counselling and education of patients | • ↓ Bleeding (from 17% to 14%) • ↔ Readmission rates (7% vs. 7%) • ↓ Mortality (from 4% to 2%) |
| Lakshmi et al. 2013,61 India | RCT, Hospital | Control group Intervention group | 40 40 | 6 | OACs | • Patient counselling • Providing patient information booklets | • ↓ Bleeding (from 85% to 60%) |
| Liang et al. 2020,63 China | RCT, Hospital | UC PEFS | 75 77 | 43 | Warfarin | • Patient education • Providing dose recommendations | • ↓ TE events (9.3% vs. 6.5%; P > 0.05) • ↔ Hospitalization (12% vs. 11.7%; P > 0.05) |
| Liu et al. 2022,57 China | RCT, Hospital | Control group Intervention group | 64 61 | 6 | Warfarin Rivaroxaban | • Medication guidance and monitoring • Patient education and other pharmaceutical care services | • ↓ Bleeding (from 43.8% to 13.1%; P < 0.05) • ↓ Thrombosis events (from 7.8% to 0%; P < 0.05) |
| Ahmed et al. 2017,53 Sudan | Quasi-experimental, Hospital | Before intervention After intervention | 135 135 | 12 | Warfarin | • Patient education and providing written information • Dose adjustments | • ↓ Bleeding (from 39.2% to 27.4%; P < 0.05) • ↓ Hospitalization (from 10.4% to 3.7%; P < 0.001) |
| Chong et al. 2021,37 Australia | Quasi-experimental, Hospital | Pre-stewardship Post-stewardship | 400 411 | 6 | Enoxaparin Heparin Others | • Education for medical officers • Chart auditing • Gamification and health promotion • Clinician performance feedback | • ↑ AC prescription appropriateness (from 78% to 88%; P = 0.004) |
| Gauci et al. 2019,42 Malta | Quasi-experimental, Hospital | Pre-implementation of MAT-AF Post-implementation of MAT-AF | 150 150 | 7 | Warfarin DOAC | • Implementing MAT-AF as a clinical tool • Documentation | • ↑ AC therapy appropriateness (from 70% to 88%) |
| Hyland et al.a 2020,59 USA | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 694 533 | 6 | Not reported | • Patient review and VTE risk stratification • Recommending TP medication selection, dose, and duration • Managing perioperative chronic antithrombotic therapies • Comprehensive discharge medication reconciliation | • ↓ Postoperative readmission rate (from 4.8% to 1.3%; P = 0.002) • ↓ VTE (from 0.6% to 0.0%; P = 0.13) |
| Khalil et al. 2021,38 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 61 65 | 7 | Apixaban Rivaroxaban Dabigatran Warfarin | • Patient education • Providing a written AF brochure that contains treatment options | • ↑ AC therapy appropriateness (from 36% to 92%; P < 0.001) |
| Kiracı et al.a 2023,33 Turkey | Quasi-experimental, Hospital | Pre-education CP intervention | 340 191 | 2 | Any TP | • Providing education to surgeons and residents • Providing optimal TP recommendations to prescribers | • ↑ TP appropriateness (from 46.8% to 61.3%; P = 0.001) • ↓ VTE (from 1.2% to 0%) • ↓ Bleeding (from 1.8% to 1.0%) |
| Lachuer et al. 2021,36 France | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 58 89 | 2.4 | VKA DOAC | • Providing training for physicians | • ↑ Prescribing practices of VKA (from 74.4% to 77.9%) and DOAC (from 74% to 82.2%) • ↑ OAC prescription appropriateness (from 74.1% to 79.8%) |
| Lee et al. 2013,62 USA | Quasi-experimental, Hospital | UC Pharmacist-managed anticoagulation clinic | 48 20 | 3 | Dabigatran | • Patient education • Patient follow-up at 2 weeks, 1 and 3 months by means of telephone or face-to-face visits • Implementing pharmacy benefits management services guidelines | • ↔ Bleeding (4.2% vs. 10%; P = 0.148) |
| Miele et al. 2017,39 USA | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 50 85 | 3 | Apixaban Dabigatran Rivaroxaban Edoxaban | • Developing and implementing pharmacist-driven protocol • Patient profile reviews • Providing an appropriate medication alternative to the prescriber | • ↑ DOAC prescription appropriateness (from 60% to 70.6%) |
| Quintens et al. 2022,60 Belgium | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 466 485 | 24 | Enoxaparin Apixaban Rivaroxaban Edoxaban Dabigatran VKA | • Implementing a pharmacist-led CMA intervention | • ↓ AC-related residual potentially inappropriate prescriptions (from 74.9 to 22.5%) • ↓ Median proportion of residual PIPs (from 78.5% to 18.2%) |
| Sarika et al. 2021,32 India | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 45 45 | 5 | Heparin Enoxaparin Fondaparinux | • Implementing ACCP guideline • Patient counselling | • ↑ Appropriateness of AC therapy (from 28.9% to 62.22%) |
| Shang et al. 2021,56 China | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 240 337 | 18 | LMWH VKA Rivaroxaban | • Thrombosis and bleeding risk assessments • Consulting with physicians to formulate antithrombotic treatment protocols • Optimizing perioperative medication regimens • Evaluating and optimizing the feasibility of discharge prescriptions • Following up the thrombosis of TJA patients in the first and third months after surgery | • ↓ Deep vein thrombosis incidence (from 3.33% to 1.78%; P > 0.05) • ↔ Minor bleeding (6.67% vs. 5.93%; P = 0.720) • Improvement in TP use, administration timing, and treatment course at post-intervention (P < 0.001) |
| Sharma et al. 2024,34 England | Quasi-experimental, General practices | Pre-intervention Post-intervention | 470 498 | 14 | Apixaban Rivaroxaban Edoxaban Dabigatran Warfarin | • Medication review • Optimizing AF treatment • Patient counselling • Bleed risk assessment | • OAC prescriptions appropriateness (from 77% to 93.8%) • Successful transition from VKA to DOACs in 25.71% of patients |
| Tyedin et al. 2020,58 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 130 108 | 5 | Warfarin | • Medication review • Proactively charting and monitoring • Ordering INRs using electronic prescribing software, following discussion with doctors | • ↓ Re-hospitalization (from 3.1% to 0%; P = 0.18) • ↓ Mortality (from 4.6% to 0%) |
| Yap et al. 2019,40 Malaysia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 142 144 | 5 | Fondaparinux Enoxaparin Heparin | • Developing and implementing a pharmacist-driven DVT Risk Alert Tool • DVT and bleeding risk assessment | • ↑ DVT prophylaxis appropriateness (from 14.3% to 31.3%; P < 0.05) |
| Wu et al. 2022,54 Taiwan | Quasi-experimental, Hospital | Conventional Pharmacist-managed group | 39 33 | 24 | Warfarin | • Patient and caregiver education • Developing a monitoring plan for INR and ADRs • Managing drug–drug and drug–food interactions • Participating in the ward check rounds • | • ↓ Major bleeding (1 vs. 0, in conventional vs. pharmacist-managed group) • |
| An et al. 2017,65 Japan | Retrospective cohort, Hospital | UC group Intervention group | 32 25 | 28 | Warfarin | • Monitoring of bleeding and PT-INR • Proposing to physicians to change the dose of warfarin when appropriate • Checking interaction of warfarin and supplements or foods | • ↓ Major bleeding (from 12.5% to 8%) • ↔ Minor bleeding (28% vs. 28.1%) |
| Ashjian et al. 2017,35 USA | Retrospective cohort, Health system data | UC Pharmacist-led DOAC | 129 129 | 16 | Dabigatran Rivaroxaban Apixaban | • Medication review | • ↑ DOAC appropriateness (from 81.1% to 93.7%; P = 0.016) |
| Bakey et al. 2022,41 USA | Retrospective cohort, Hospital | No pharmacist involvement Pharmacist involvement | 44 14 | 12 | Apixaban Rivaroxaban Warfarin Enoxaparin | • Patient counselling • Chart review • Evaluating OAC options • Providing additional written discharge instructions to patients | • ↑ AC appropriateness (from 70.5% to 92.9%; P = 0.046) • ↓ Readmission rates (from 20.5% to 7.1%; P = 0.424) |
| Derington et al. 2023,71 USA | Retrospective cohort, Hospital | UC AMS | 6182 4939 | 39 | Dabigatran Apixaban Rivaroxaban Warfarin | • Managing AC therapy • Evaluating DOAC appropriateness, dose-indication match, potential drug–drug interactions, needed laboratory measurements, and dose adjustments • Patient education • Ordering relevant laboratory tests | • ↔ Net clinical outcomes (major bleeding, TE, and mortality) (hazard ratio: 0.9, 95% CI: 0.7–1.0) • ↑ Major bleeding in patients with DOACs (hazard ratio: 1.5, 95% CI: 1.0–2.3) • ↓ Mortality in patients with DOAC (hazard ratio: 0.7, 95% CI: 0.6–0.9) • ↔ TE events (1.3% vs. 1.9%) • ↓ Mortality (from 13.3% to 11.6%) |
| DiRenzo et al. 2018,73 USA | Prospective cohort, Clot clinic | PCP Pharmacist-managed clinic | 17 17 | 6 | Rivaroxaban | • Patient counselling • Using the American College of Chest Physicians Guidelines (CHEST) • Reviewing patient profile • Identifying VTE symptoms and performing a physical examination | • ↔ TE events (6% vs. 6%; P = 1.000) • ↔ Hospitalization (6% vs. 6%; P = 1.000) |
| Han et al.a 2021,44 USA | Retrospective cohort, Academic health centre | Physician led CP led | 138 553 | 14 | Warfarin Apixaban Rivaroxaban Dabigatran Antiplatelet | • Implementing the best practice alert in addition to their pharmacotherapy training | • ↑ Appropriateness of OAC discontinuation during pre-procedure (from 91% to 97.4%; P = 0.001) • ↑ Appropriateness of bridging therapy (from 50% to 72.2%; P = 0.39) |
| Jones et al. 2020,74 USA | Retrospective cohort, Healthcare system data | Non-AMS group AMS | 370 90 | 42 | Apixaban Rivaroxaban Dabigatran | • Patient education followed by phone calls • Chart reviews | • ↑ Composite outcomes (bleeding, TE, mortality) (from 13.5% to 18.9%; P = 0.29) • ↑ Bleeding (from 10.5% to 18.9%; P = 0.03) • ↓ TE events (from 1.1% to 0%) • ↓ Mortality (from 1.9% to 0%) |
| Kose et al. 2018,66 Japan | Retrospective cohort, Hospital | UC Intervention group | 23 16 | 16 | Warfarin | • Monitoring of bleeding and PT-INR • Checking interaction of warfarin and supplements or foods • Advising physicians to adjust the dose of warfarin as needed | • ↔ Major bleeding (17.4% vs. 18.8%) • ↓ Minor bleeding (from 43.5% to 31.3%) |
| Kurimura et al.a 2023,67 Japan | Retrospective cohort, Hospital | Non-intervention group Intervention group | 264 132 | 48 | Warfarin Apixaban Rivaroxaban Dabigatran Edoxaban Antiplatelets | • Checking all prescription drugs from the hospital and other healthcare facilities • Pharmaceutical advice and suggesting prescription changes to physicians • Assessing adherence, adverse events, and discontinuation of unnecessary medications | • ↓ Bleeding (from 28.4% to 17.4%; P = 0.019) • ↓ TE (from 14% to 6.8%; P = 0.44) • ↑ Mortality (from 9% to 13.6%; P = 0.17) |
| Li et al. 2020,64 China | Prospective cohort, Hospital | UC PEFS | 202 179 | 6 | Rivaroxaban | • Observing potential interaction with rivaroxaban and managing bleeding and embolic complications • Evaluating patient medication adherence • Distributing paper-based medication education materials • Pharmacists kept in touch with the patients through WeChat or telephone weekly | • ↓ Bleeding (from 26.7% to 16.2%) • ↔ TE (9.4% vs. 7.8%; P = 0.675) |
| Manzoor et al. 2018,70 USA | Retrospective cohort, Hospital | NMAC PMAC | 100 100 | 15 | Warfarin | • Monitoring and managing patients receiving warfarin therapy | • ↑ Hospitalization or ED visits for patients in the NMAC group [eight times higher (OR: 7.68, 95% CI: 1.1–55.9, P < 0.05)] |
| Noor et al. 2021,69 Saudi Arabia | Retrospective cohort, Hospital | HMAC PMAC | 124 104 | 24 | Warfarin | • Patient counselling • Assessing INRs, warfarin therapy • Assessing ADRs, drug–drug interactions, or drug–food interactions • Documenting all therapeutic recommendations and prescribing a new warfarin prescription | • ↔ Bleeding (12.5% vs. 12.1%) • ↓ TE events (5.7% to 4%) • ↓ Hospitalization rates (from 9.6% to 4%) |
| Tarasiuk et al. 2018,68 USA | Retrospective cohort, Hospital | Nurse-managed group Pharmacist-managed group | 240 228 | 12 | Warfarin | • Managing patient dosing and monitoring autonomously utilizing clinical judgement | • ↓ Hospitalizations (OR: 0.29, P < 0.001) • ↓ Bleeding (from 2.1% to 1.3%) • ↓ TE (from 2.1% to 0.4%) • ↔ ED visits (39.2% vs. 43%; P = 0.402) |
| Zhang et al.b 2023,43 China | Retrospective cohort, Hospital | Control group CP services group | 162 176 | 21 | Heparin | • Assessing VTE and bleeding risk on admission and giving VTE prophylaxis recommendations • Providing medication consultation to physicians, nurses, and patients | • ↑ TP appropriateness (from 59% to 84%) • ↓ TE events (from 17% to 9%; P = 0.037) • ↓ Bleeding (from 11% to 5%; P = 0.042) • ↓ Mortality (from 28% to 14%; P = 0.001) |
| Study, year and country | Study design and settings | Comparison groups | Sample size | Follow-up time (months) | Drugs used | Specific intervention(s) given | Key findings (control vs. intervention) |
|---|---|---|---|---|---|---|---|
| Falamić et al. 2019,72 Croatia | RCT, Community pharmacy | Control group Intervention group | 66 65 | 6 | Warfarin | • Patient education • Optimizing warfarin dose and contacting physicians to avoid drug interactions | • ↓ Bleeding (from 85% to 29%; P < 0.05) |
| Karaoui et al. 2021,55 Lebanon | RCT, Hospital | Standard of care Pharmacist-counsel | 100 100 | 1 | Apixaban Dabigatran Rivaroxaban VKA | • Counselling and education of patients | • ↓ Bleeding (from 17% to 14%) • ↔ Readmission rates (7% vs. 7%) • ↓ Mortality (from 4% to 2%) |
| Lakshmi et al. 2013,61 India | RCT, Hospital | Control group Intervention group | 40 40 | 6 | OACs | • Patient counselling • Providing patient information booklets | • ↓ Bleeding (from 85% to 60%) |
| Liang et al. 2020,63 China | RCT, Hospital | UC PEFS | 75 77 | 43 | Warfarin | • Patient education • Providing dose recommendations | • ↓ TE events (9.3% vs. 6.5%; P > 0.05) • ↔ Hospitalization (12% vs. 11.7%; P > 0.05) |
| Liu et al. 2022,57 China | RCT, Hospital | Control group Intervention group | 64 61 | 6 | Warfarin Rivaroxaban | • Medication guidance and monitoring • Patient education and other pharmaceutical care services | • ↓ Bleeding (from 43.8% to 13.1%; P < 0.05) • ↓ Thrombosis events (from 7.8% to 0%; P < 0.05) |
| Ahmed et al. 2017,53 Sudan | Quasi-experimental, Hospital | Before intervention After intervention | 135 135 | 12 | Warfarin | • Patient education and providing written information • Dose adjustments | • ↓ Bleeding (from 39.2% to 27.4%; P < 0.05) • ↓ Hospitalization (from 10.4% to 3.7%; P < 0.001) |
| Chong et al. 2021,37 Australia | Quasi-experimental, Hospital | Pre-stewardship Post-stewardship | 400 411 | 6 | Enoxaparin Heparin Others | • Education for medical officers • Chart auditing • Gamification and health promotion • Clinician performance feedback | • ↑ AC prescription appropriateness (from 78% to 88%; P = 0.004) |
| Gauci et al. 2019,42 Malta | Quasi-experimental, Hospital | Pre-implementation of MAT-AF Post-implementation of MAT-AF | 150 150 | 7 | Warfarin DOAC | • Implementing MAT-AF as a clinical tool • Documentation | • ↑ AC therapy appropriateness (from 70% to 88%) |
| Hyland et al.a 2020,59 USA | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 694 533 | 6 | Not reported | • Patient review and VTE risk stratification • Recommending TP medication selection, dose, and duration • Managing perioperative chronic antithrombotic therapies • Comprehensive discharge medication reconciliation | • ↓ Postoperative readmission rate (from 4.8% to 1.3%; P = 0.002) • ↓ VTE (from 0.6% to 0.0%; P = 0.13) |
| Khalil et al. 2021,38 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 61 65 | 7 | Apixaban Rivaroxaban Dabigatran Warfarin | • Patient education • Providing a written AF brochure that contains treatment options | • ↑ AC therapy appropriateness (from 36% to 92%; P < 0.001) |
| Kiracı et al.a 2023,33 Turkey | Quasi-experimental, Hospital | Pre-education CP intervention | 340 191 | 2 | Any TP | • Providing education to surgeons and residents • Providing optimal TP recommendations to prescribers | • ↑ TP appropriateness (from 46.8% to 61.3%; P = 0.001) • ↓ VTE (from 1.2% to 0%) • ↓ Bleeding (from 1.8% to 1.0%) |
| Lachuer et al. 2021,36 France | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 58 89 | 2.4 | VKA DOAC | • Providing training for physicians | • ↑ Prescribing practices of VKA (from 74.4% to 77.9%) and DOAC (from 74% to 82.2%) • ↑ OAC prescription appropriateness (from 74.1% to 79.8%) |
| Lee et al. 2013,62 USA | Quasi-experimental, Hospital | UC Pharmacist-managed anticoagulation clinic | 48 20 | 3 | Dabigatran | • Patient education • Patient follow-up at 2 weeks, 1 and 3 months by means of telephone or face-to-face visits • Implementing pharmacy benefits management services guidelines | • ↔ Bleeding (4.2% vs. 10%; P = 0.148) |
| Miele et al. 2017,39 USA | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 50 85 | 3 | Apixaban Dabigatran Rivaroxaban Edoxaban | • Developing and implementing pharmacist-driven protocol • Patient profile reviews • Providing an appropriate medication alternative to the prescriber | • ↑ DOAC prescription appropriateness (from 60% to 70.6%) |
| Quintens et al. 2022,60 Belgium | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 466 485 | 24 | Enoxaparin Apixaban Rivaroxaban Edoxaban Dabigatran VKA | • Implementing a pharmacist-led CMA intervention | • ↓ AC-related residual potentially inappropriate prescriptions (from 74.9 to 22.5%) • ↓ Median proportion of residual PIPs (from 78.5% to 18.2%) |
| Sarika et al. 2021,32 India | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 45 45 | 5 | Heparin Enoxaparin Fondaparinux | • Implementing ACCP guideline • Patient counselling | • ↑ Appropriateness of AC therapy (from 28.9% to 62.22%) |
| Shang et al. 2021,56 China | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 240 337 | 18 | LMWH VKA Rivaroxaban | • Thrombosis and bleeding risk assessments • Consulting with physicians to formulate antithrombotic treatment protocols • Optimizing perioperative medication regimens • Evaluating and optimizing the feasibility of discharge prescriptions • Following up the thrombosis of TJA patients in the first and third months after surgery | • ↓ Deep vein thrombosis incidence (from 3.33% to 1.78%; P > 0.05) • ↔ Minor bleeding (6.67% vs. 5.93%; P = 0.720) • Improvement in TP use, administration timing, and treatment course at post-intervention (P < 0.001) |
| Sharma et al. 2024,34 England | Quasi-experimental, General practices | Pre-intervention Post-intervention | 470 498 | 14 | Apixaban Rivaroxaban Edoxaban Dabigatran Warfarin | • Medication review • Optimizing AF treatment • Patient counselling • Bleed risk assessment | • OAC prescriptions appropriateness (from 77% to 93.8%) • Successful transition from VKA to DOACs in 25.71% of patients |
| Tyedin et al. 2020,58 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 130 108 | 5 | Warfarin | • Medication review • Proactively charting and monitoring • Ordering INRs using electronic prescribing software, following discussion with doctors | • ↓ Re-hospitalization (from 3.1% to 0%; P = 0.18) • ↓ Mortality (from 4.6% to 0%) |
| Yap et al. 2019,40 Malaysia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 142 144 | 5 | Fondaparinux Enoxaparin Heparin | • Developing and implementing a pharmacist-driven DVT Risk Alert Tool • DVT and bleeding risk assessment | • ↑ DVT prophylaxis appropriateness (from 14.3% to 31.3%; P < 0.05) |
| Wu et al. 2022,54 Taiwan | Quasi-experimental, Hospital | Conventional Pharmacist-managed group | 39 33 | 24 | Warfarin | • Patient and caregiver education • Developing a monitoring plan for INR and ADRs • Managing drug–drug and drug–food interactions • Participating in the ward check rounds • | • ↓ Major bleeding (1 vs. 0, in conventional vs. pharmacist-managed group) • |
| An et al. 2017,65 Japan | Retrospective cohort, Hospital | UC group Intervention group | 32 25 | 28 | Warfarin | • Monitoring of bleeding and PT-INR • Proposing to physicians to change the dose of warfarin when appropriate • Checking interaction of warfarin and supplements or foods | • ↓ Major bleeding (from 12.5% to 8%) • ↔ Minor bleeding (28% vs. 28.1%) |
| Ashjian et al. 2017,35 USA | Retrospective cohort, Health system data | UC Pharmacist-led DOAC | 129 129 | 16 | Dabigatran Rivaroxaban Apixaban | • Medication review | • ↑ DOAC appropriateness (from 81.1% to 93.7%; P = 0.016) |
| Bakey et al. 2022,41 USA | Retrospective cohort, Hospital | No pharmacist involvement Pharmacist involvement | 44 14 | 12 | Apixaban Rivaroxaban Warfarin Enoxaparin | • Patient counselling • Chart review • Evaluating OAC options • Providing additional written discharge instructions to patients | • ↑ AC appropriateness (from 70.5% to 92.9%; P = 0.046) • ↓ Readmission rates (from 20.5% to 7.1%; P = 0.424) |
| Derington et al. 2023,71 USA | Retrospective cohort, Hospital | UC AMS | 6182 4939 | 39 | Dabigatran Apixaban Rivaroxaban Warfarin | • Managing AC therapy • Evaluating DOAC appropriateness, dose-indication match, potential drug–drug interactions, needed laboratory measurements, and dose adjustments • Patient education • Ordering relevant laboratory tests | • ↔ Net clinical outcomes (major bleeding, TE, and mortality) (hazard ratio: 0.9, 95% CI: 0.7–1.0) • ↑ Major bleeding in patients with DOACs (hazard ratio: 1.5, 95% CI: 1.0–2.3) • ↓ Mortality in patients with DOAC (hazard ratio: 0.7, 95% CI: 0.6–0.9) • ↔ TE events (1.3% vs. 1.9%) • ↓ Mortality (from 13.3% to 11.6%) |
| DiRenzo et al. 2018,73 USA | Prospective cohort, Clot clinic | PCP Pharmacist-managed clinic | 17 17 | 6 | Rivaroxaban | • Patient counselling • Using the American College of Chest Physicians Guidelines (CHEST) • Reviewing patient profile • Identifying VTE symptoms and performing a physical examination | • ↔ TE events (6% vs. 6%; P = 1.000) • ↔ Hospitalization (6% vs. 6%; P = 1.000) |
| Han et al.a 2021,44 USA | Retrospective cohort, Academic health centre | Physician led CP led | 138 553 | 14 | Warfarin Apixaban Rivaroxaban Dabigatran Antiplatelet | • Implementing the best practice alert in addition to their pharmacotherapy training | • ↑ Appropriateness of OAC discontinuation during pre-procedure (from 91% to 97.4%; P = 0.001) • ↑ Appropriateness of bridging therapy (from 50% to 72.2%; P = 0.39) |
| Jones et al. 2020,74 USA | Retrospective cohort, Healthcare system data | Non-AMS group AMS | 370 90 | 42 | Apixaban Rivaroxaban Dabigatran | • Patient education followed by phone calls • Chart reviews | • ↑ Composite outcomes (bleeding, TE, mortality) (from 13.5% to 18.9%; P = 0.29) • ↑ Bleeding (from 10.5% to 18.9%; P = 0.03) • ↓ TE events (from 1.1% to 0%) • ↓ Mortality (from 1.9% to 0%) |
| Kose et al. 2018,66 Japan | Retrospective cohort, Hospital | UC Intervention group | 23 16 | 16 | Warfarin | • Monitoring of bleeding and PT-INR • Checking interaction of warfarin and supplements or foods • Advising physicians to adjust the dose of warfarin as needed | • ↔ Major bleeding (17.4% vs. 18.8%) • ↓ Minor bleeding (from 43.5% to 31.3%) |
| Kurimura et al.a 2023,67 Japan | Retrospective cohort, Hospital | Non-intervention group Intervention group | 264 132 | 48 | Warfarin Apixaban Rivaroxaban Dabigatran Edoxaban Antiplatelets | • Checking all prescription drugs from the hospital and other healthcare facilities • Pharmaceutical advice and suggesting prescription changes to physicians • Assessing adherence, adverse events, and discontinuation of unnecessary medications | • ↓ Bleeding (from 28.4% to 17.4%; P = 0.019) • ↓ TE (from 14% to 6.8%; P = 0.44) • ↑ Mortality (from 9% to 13.6%; P = 0.17) |
| Li et al. 2020,64 China | Prospective cohort, Hospital | UC PEFS | 202 179 | 6 | Rivaroxaban | • Observing potential interaction with rivaroxaban and managing bleeding and embolic complications • Evaluating patient medication adherence • Distributing paper-based medication education materials • Pharmacists kept in touch with the patients through WeChat or telephone weekly | • ↓ Bleeding (from 26.7% to 16.2%) • ↔ TE (9.4% vs. 7.8%; P = 0.675) |
| Manzoor et al. 2018,70 USA | Retrospective cohort, Hospital | NMAC PMAC | 100 100 | 15 | Warfarin | • Monitoring and managing patients receiving warfarin therapy | • ↑ Hospitalization or ED visits for patients in the NMAC group [eight times higher (OR: 7.68, 95% CI: 1.1–55.9, P < 0.05)] |
| Noor et al. 2021,69 Saudi Arabia | Retrospective cohort, Hospital | HMAC PMAC | 124 104 | 24 | Warfarin | • Patient counselling • Assessing INRs, warfarin therapy • Assessing ADRs, drug–drug interactions, or drug–food interactions • Documenting all therapeutic recommendations and prescribing a new warfarin prescription | • ↔ Bleeding (12.5% vs. 12.1%) • ↓ TE events (5.7% to 4%) • ↓ Hospitalization rates (from 9.6% to 4%) |
| Tarasiuk et al. 2018,68 USA | Retrospective cohort, Hospital | Nurse-managed group Pharmacist-managed group | 240 228 | 12 | Warfarin | • Managing patient dosing and monitoring autonomously utilizing clinical judgement | • ↓ Hospitalizations (OR: 0.29, P < 0.001) • ↓ Bleeding (from 2.1% to 1.3%) • ↓ TE (from 2.1% to 0.4%) • ↔ ED visits (39.2% vs. 43%; P = 0.402) |
| Zhang et al.b 2023,43 China | Retrospective cohort, Hospital | Control group CP services group | 162 176 | 21 | Heparin | • Assessing VTE and bleeding risk on admission and giving VTE prophylaxis recommendations • Providing medication consultation to physicians, nurses, and patients | • ↑ TP appropriateness (from 59% to 84%) • ↓ TE events (from 17% to 9%; P = 0.037) • ↓ Bleeding (from 11% to 5%; P = 0.042) • ↓ Mortality (from 28% to 14%; P = 0.001) |
Abbreviations: AC, anticoagulant; ACCP, American College of Clinical Pharmacy; ADR, adverse drug reaction; AF, atrial fibrillation; AMS, anticoagulation management service; CI, confidence interval; CMA, Check of Medication Appropriateness; CP, clinical pharmacist; DOAC, direct-acting oral anticoagulant; ED, emergency department; HMAC, haematologist-managed anticoagulation clinic; INR, international normalized ratio; LMWH, low-molecular-weight heparin; MAT-AF, medication assessment tool of atrial fibrillation; MUR, medication use review; NMAC, nurse-managed anticoagulation clinic; OAC, oral anticoagulant; OR, odds ratio; PEFS, pharmacist-led education and follow-up service; PMAC, pharmacist-managed anticoagulation clinic; PT-INR, prothrombin time–international normalized ratio; RCT, randomized controlled trial; TE, thromboembolic; TJA, total joint arthroplasty; TP, thromboprophylaxis; UC, usual care; VKA, vitamin K antagonist; VTE, venous thromboembolism.
aIncluded patients treated with antithrombotic medications (anticoagulants and/or antiplatelets).
bIncluded patients aged ≥16 years.
Characteristics of the included studies in the review
| Study, year and country | Study design and settings | Comparison groups | Sample size | Follow-up time (months) | Drugs used | Specific intervention(s) given | Key findings (control vs. intervention) |
|---|---|---|---|---|---|---|---|
| Falamić et al. 2019,72 Croatia | RCT, Community pharmacy | Control group Intervention group | 66 65 | 6 | Warfarin | • Patient education • Optimizing warfarin dose and contacting physicians to avoid drug interactions | • ↓ Bleeding (from 85% to 29%; P < 0.05) |
| Karaoui et al. 2021,55 Lebanon | RCT, Hospital | Standard of care Pharmacist-counsel | 100 100 | 1 | Apixaban Dabigatran Rivaroxaban VKA | • Counselling and education of patients | • ↓ Bleeding (from 17% to 14%) • ↔ Readmission rates (7% vs. 7%) • ↓ Mortality (from 4% to 2%) |
| Lakshmi et al. 2013,61 India | RCT, Hospital | Control group Intervention group | 40 40 | 6 | OACs | • Patient counselling • Providing patient information booklets | • ↓ Bleeding (from 85% to 60%) |
| Liang et al. 2020,63 China | RCT, Hospital | UC PEFS | 75 77 | 43 | Warfarin | • Patient education • Providing dose recommendations | • ↓ TE events (9.3% vs. 6.5%; P > 0.05) • ↔ Hospitalization (12% vs. 11.7%; P > 0.05) |
| Liu et al. 2022,57 China | RCT, Hospital | Control group Intervention group | 64 61 | 6 | Warfarin Rivaroxaban | • Medication guidance and monitoring • Patient education and other pharmaceutical care services | • ↓ Bleeding (from 43.8% to 13.1%; P < 0.05) • ↓ Thrombosis events (from 7.8% to 0%; P < 0.05) |
| Ahmed et al. 2017,53 Sudan | Quasi-experimental, Hospital | Before intervention After intervention | 135 135 | 12 | Warfarin | • Patient education and providing written information • Dose adjustments | • ↓ Bleeding (from 39.2% to 27.4%; P < 0.05) • ↓ Hospitalization (from 10.4% to 3.7%; P < 0.001) |
| Chong et al. 2021,37 Australia | Quasi-experimental, Hospital | Pre-stewardship Post-stewardship | 400 411 | 6 | Enoxaparin Heparin Others | • Education for medical officers • Chart auditing • Gamification and health promotion • Clinician performance feedback | • ↑ AC prescription appropriateness (from 78% to 88%; P = 0.004) |
| Gauci et al. 2019,42 Malta | Quasi-experimental, Hospital | Pre-implementation of MAT-AF Post-implementation of MAT-AF | 150 150 | 7 | Warfarin DOAC | • Implementing MAT-AF as a clinical tool • Documentation | • ↑ AC therapy appropriateness (from 70% to 88%) |
| Hyland et al.a 2020,59 USA | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 694 533 | 6 | Not reported | • Patient review and VTE risk stratification • Recommending TP medication selection, dose, and duration • Managing perioperative chronic antithrombotic therapies • Comprehensive discharge medication reconciliation | • ↓ Postoperative readmission rate (from 4.8% to 1.3%; P = 0.002) • ↓ VTE (from 0.6% to 0.0%; P = 0.13) |
| Khalil et al. 2021,38 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 61 65 | 7 | Apixaban Rivaroxaban Dabigatran Warfarin | • Patient education • Providing a written AF brochure that contains treatment options | • ↑ AC therapy appropriateness (from 36% to 92%; P < 0.001) |
| Kiracı et al.a 2023,33 Turkey | Quasi-experimental, Hospital | Pre-education CP intervention | 340 191 | 2 | Any TP | • Providing education to surgeons and residents • Providing optimal TP recommendations to prescribers | • ↑ TP appropriateness (from 46.8% to 61.3%; P = 0.001) • ↓ VTE (from 1.2% to 0%) • ↓ Bleeding (from 1.8% to 1.0%) |
| Lachuer et al. 2021,36 France | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 58 89 | 2.4 | VKA DOAC | • Providing training for physicians | • ↑ Prescribing practices of VKA (from 74.4% to 77.9%) and DOAC (from 74% to 82.2%) • ↑ OAC prescription appropriateness (from 74.1% to 79.8%) |
| Lee et al. 2013,62 USA | Quasi-experimental, Hospital | UC Pharmacist-managed anticoagulation clinic | 48 20 | 3 | Dabigatran | • Patient education • Patient follow-up at 2 weeks, 1 and 3 months by means of telephone or face-to-face visits • Implementing pharmacy benefits management services guidelines | • ↔ Bleeding (4.2% vs. 10%; P = 0.148) |
| Miele et al. 2017,39 USA | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 50 85 | 3 | Apixaban Dabigatran Rivaroxaban Edoxaban | • Developing and implementing pharmacist-driven protocol • Patient profile reviews • Providing an appropriate medication alternative to the prescriber | • ↑ DOAC prescription appropriateness (from 60% to 70.6%) |
| Quintens et al. 2022,60 Belgium | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 466 485 | 24 | Enoxaparin Apixaban Rivaroxaban Edoxaban Dabigatran VKA | • Implementing a pharmacist-led CMA intervention | • ↓ AC-related residual potentially inappropriate prescriptions (from 74.9 to 22.5%) • ↓ Median proportion of residual PIPs (from 78.5% to 18.2%) |
| Sarika et al. 2021,32 India | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 45 45 | 5 | Heparin Enoxaparin Fondaparinux | • Implementing ACCP guideline • Patient counselling | • ↑ Appropriateness of AC therapy (from 28.9% to 62.22%) |
| Shang et al. 2021,56 China | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 240 337 | 18 | LMWH VKA Rivaroxaban | • Thrombosis and bleeding risk assessments • Consulting with physicians to formulate antithrombotic treatment protocols • Optimizing perioperative medication regimens • Evaluating and optimizing the feasibility of discharge prescriptions • Following up the thrombosis of TJA patients in the first and third months after surgery | • ↓ Deep vein thrombosis incidence (from 3.33% to 1.78%; P > 0.05) • ↔ Minor bleeding (6.67% vs. 5.93%; P = 0.720) • Improvement in TP use, administration timing, and treatment course at post-intervention (P < 0.001) |
| Sharma et al. 2024,34 England | Quasi-experimental, General practices | Pre-intervention Post-intervention | 470 498 | 14 | Apixaban Rivaroxaban Edoxaban Dabigatran Warfarin | • Medication review • Optimizing AF treatment • Patient counselling • Bleed risk assessment | • OAC prescriptions appropriateness (from 77% to 93.8%) • Successful transition from VKA to DOACs in 25.71% of patients |
| Tyedin et al. 2020,58 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 130 108 | 5 | Warfarin | • Medication review • Proactively charting and monitoring • Ordering INRs using electronic prescribing software, following discussion with doctors | • ↓ Re-hospitalization (from 3.1% to 0%; P = 0.18) • ↓ Mortality (from 4.6% to 0%) |
| Yap et al. 2019,40 Malaysia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 142 144 | 5 | Fondaparinux Enoxaparin Heparin | • Developing and implementing a pharmacist-driven DVT Risk Alert Tool • DVT and bleeding risk assessment | • ↑ DVT prophylaxis appropriateness (from 14.3% to 31.3%; P < 0.05) |
| Wu et al. 2022,54 Taiwan | Quasi-experimental, Hospital | Conventional Pharmacist-managed group | 39 33 | 24 | Warfarin | • Patient and caregiver education • Developing a monitoring plan for INR and ADRs • Managing drug–drug and drug–food interactions • Participating in the ward check rounds • | • ↓ Major bleeding (1 vs. 0, in conventional vs. pharmacist-managed group) • |
| An et al. 2017,65 Japan | Retrospective cohort, Hospital | UC group Intervention group | 32 25 | 28 | Warfarin | • Monitoring of bleeding and PT-INR • Proposing to physicians to change the dose of warfarin when appropriate • Checking interaction of warfarin and supplements or foods | • ↓ Major bleeding (from 12.5% to 8%) • ↔ Minor bleeding (28% vs. 28.1%) |
| Ashjian et al. 2017,35 USA | Retrospective cohort, Health system data | UC Pharmacist-led DOAC | 129 129 | 16 | Dabigatran Rivaroxaban Apixaban | • Medication review | • ↑ DOAC appropriateness (from 81.1% to 93.7%; P = 0.016) |
| Bakey et al. 2022,41 USA | Retrospective cohort, Hospital | No pharmacist involvement Pharmacist involvement | 44 14 | 12 | Apixaban Rivaroxaban Warfarin Enoxaparin | • Patient counselling • Chart review • Evaluating OAC options • Providing additional written discharge instructions to patients | • ↑ AC appropriateness (from 70.5% to 92.9%; P = 0.046) • ↓ Readmission rates (from 20.5% to 7.1%; P = 0.424) |
| Derington et al. 2023,71 USA | Retrospective cohort, Hospital | UC AMS | 6182 4939 | 39 | Dabigatran Apixaban Rivaroxaban Warfarin | • Managing AC therapy • Evaluating DOAC appropriateness, dose-indication match, potential drug–drug interactions, needed laboratory measurements, and dose adjustments • Patient education • Ordering relevant laboratory tests | • ↔ Net clinical outcomes (major bleeding, TE, and mortality) (hazard ratio: 0.9, 95% CI: 0.7–1.0) • ↑ Major bleeding in patients with DOACs (hazard ratio: 1.5, 95% CI: 1.0–2.3) • ↓ Mortality in patients with DOAC (hazard ratio: 0.7, 95% CI: 0.6–0.9) • ↔ TE events (1.3% vs. 1.9%) • ↓ Mortality (from 13.3% to 11.6%) |
| DiRenzo et al. 2018,73 USA | Prospective cohort, Clot clinic | PCP Pharmacist-managed clinic | 17 17 | 6 | Rivaroxaban | • Patient counselling • Using the American College of Chest Physicians Guidelines (CHEST) • Reviewing patient profile • Identifying VTE symptoms and performing a physical examination | • ↔ TE events (6% vs. 6%; P = 1.000) • ↔ Hospitalization (6% vs. 6%; P = 1.000) |
| Han et al.a 2021,44 USA | Retrospective cohort, Academic health centre | Physician led CP led | 138 553 | 14 | Warfarin Apixaban Rivaroxaban Dabigatran Antiplatelet | • Implementing the best practice alert in addition to their pharmacotherapy training | • ↑ Appropriateness of OAC discontinuation during pre-procedure (from 91% to 97.4%; P = 0.001) • ↑ Appropriateness of bridging therapy (from 50% to 72.2%; P = 0.39) |
| Jones et al. 2020,74 USA | Retrospective cohort, Healthcare system data | Non-AMS group AMS | 370 90 | 42 | Apixaban Rivaroxaban Dabigatran | • Patient education followed by phone calls • Chart reviews | • ↑ Composite outcomes (bleeding, TE, mortality) (from 13.5% to 18.9%; P = 0.29) • ↑ Bleeding (from 10.5% to 18.9%; P = 0.03) • ↓ TE events (from 1.1% to 0%) • ↓ Mortality (from 1.9% to 0%) |
| Kose et al. 2018,66 Japan | Retrospective cohort, Hospital | UC Intervention group | 23 16 | 16 | Warfarin | • Monitoring of bleeding and PT-INR • Checking interaction of warfarin and supplements or foods • Advising physicians to adjust the dose of warfarin as needed | • ↔ Major bleeding (17.4% vs. 18.8%) • ↓ Minor bleeding (from 43.5% to 31.3%) |
| Kurimura et al.a 2023,67 Japan | Retrospective cohort, Hospital | Non-intervention group Intervention group | 264 132 | 48 | Warfarin Apixaban Rivaroxaban Dabigatran Edoxaban Antiplatelets | • Checking all prescription drugs from the hospital and other healthcare facilities • Pharmaceutical advice and suggesting prescription changes to physicians • Assessing adherence, adverse events, and discontinuation of unnecessary medications | • ↓ Bleeding (from 28.4% to 17.4%; P = 0.019) • ↓ TE (from 14% to 6.8%; P = 0.44) • ↑ Mortality (from 9% to 13.6%; P = 0.17) |
| Li et al. 2020,64 China | Prospective cohort, Hospital | UC PEFS | 202 179 | 6 | Rivaroxaban | • Observing potential interaction with rivaroxaban and managing bleeding and embolic complications • Evaluating patient medication adherence • Distributing paper-based medication education materials • Pharmacists kept in touch with the patients through WeChat or telephone weekly | • ↓ Bleeding (from 26.7% to 16.2%) • ↔ TE (9.4% vs. 7.8%; P = 0.675) |
| Manzoor et al. 2018,70 USA | Retrospective cohort, Hospital | NMAC PMAC | 100 100 | 15 | Warfarin | • Monitoring and managing patients receiving warfarin therapy | • ↑ Hospitalization or ED visits for patients in the NMAC group [eight times higher (OR: 7.68, 95% CI: 1.1–55.9, P < 0.05)] |
| Noor et al. 2021,69 Saudi Arabia | Retrospective cohort, Hospital | HMAC PMAC | 124 104 | 24 | Warfarin | • Patient counselling • Assessing INRs, warfarin therapy • Assessing ADRs, drug–drug interactions, or drug–food interactions • Documenting all therapeutic recommendations and prescribing a new warfarin prescription | • ↔ Bleeding (12.5% vs. 12.1%) • ↓ TE events (5.7% to 4%) • ↓ Hospitalization rates (from 9.6% to 4%) |
| Tarasiuk et al. 2018,68 USA | Retrospective cohort, Hospital | Nurse-managed group Pharmacist-managed group | 240 228 | 12 | Warfarin | • Managing patient dosing and monitoring autonomously utilizing clinical judgement | • ↓ Hospitalizations (OR: 0.29, P < 0.001) • ↓ Bleeding (from 2.1% to 1.3%) • ↓ TE (from 2.1% to 0.4%) • ↔ ED visits (39.2% vs. 43%; P = 0.402) |
| Zhang et al.b 2023,43 China | Retrospective cohort, Hospital | Control group CP services group | 162 176 | 21 | Heparin | • Assessing VTE and bleeding risk on admission and giving VTE prophylaxis recommendations • Providing medication consultation to physicians, nurses, and patients | • ↑ TP appropriateness (from 59% to 84%) • ↓ TE events (from 17% to 9%; P = 0.037) • ↓ Bleeding (from 11% to 5%; P = 0.042) • ↓ Mortality (from 28% to 14%; P = 0.001) |
| Study, year and country | Study design and settings | Comparison groups | Sample size | Follow-up time (months) | Drugs used | Specific intervention(s) given | Key findings (control vs. intervention) |
|---|---|---|---|---|---|---|---|
| Falamić et al. 2019,72 Croatia | RCT, Community pharmacy | Control group Intervention group | 66 65 | 6 | Warfarin | • Patient education • Optimizing warfarin dose and contacting physicians to avoid drug interactions | • ↓ Bleeding (from 85% to 29%; P < 0.05) |
| Karaoui et al. 2021,55 Lebanon | RCT, Hospital | Standard of care Pharmacist-counsel | 100 100 | 1 | Apixaban Dabigatran Rivaroxaban VKA | • Counselling and education of patients | • ↓ Bleeding (from 17% to 14%) • ↔ Readmission rates (7% vs. 7%) • ↓ Mortality (from 4% to 2%) |
| Lakshmi et al. 2013,61 India | RCT, Hospital | Control group Intervention group | 40 40 | 6 | OACs | • Patient counselling • Providing patient information booklets | • ↓ Bleeding (from 85% to 60%) |
| Liang et al. 2020,63 China | RCT, Hospital | UC PEFS | 75 77 | 43 | Warfarin | • Patient education • Providing dose recommendations | • ↓ TE events (9.3% vs. 6.5%; P > 0.05) • ↔ Hospitalization (12% vs. 11.7%; P > 0.05) |
| Liu et al. 2022,57 China | RCT, Hospital | Control group Intervention group | 64 61 | 6 | Warfarin Rivaroxaban | • Medication guidance and monitoring • Patient education and other pharmaceutical care services | • ↓ Bleeding (from 43.8% to 13.1%; P < 0.05) • ↓ Thrombosis events (from 7.8% to 0%; P < 0.05) |
| Ahmed et al. 2017,53 Sudan | Quasi-experimental, Hospital | Before intervention After intervention | 135 135 | 12 | Warfarin | • Patient education and providing written information • Dose adjustments | • ↓ Bleeding (from 39.2% to 27.4%; P < 0.05) • ↓ Hospitalization (from 10.4% to 3.7%; P < 0.001) |
| Chong et al. 2021,37 Australia | Quasi-experimental, Hospital | Pre-stewardship Post-stewardship | 400 411 | 6 | Enoxaparin Heparin Others | • Education for medical officers • Chart auditing • Gamification and health promotion • Clinician performance feedback | • ↑ AC prescription appropriateness (from 78% to 88%; P = 0.004) |
| Gauci et al. 2019,42 Malta | Quasi-experimental, Hospital | Pre-implementation of MAT-AF Post-implementation of MAT-AF | 150 150 | 7 | Warfarin DOAC | • Implementing MAT-AF as a clinical tool • Documentation | • ↑ AC therapy appropriateness (from 70% to 88%) |
| Hyland et al.a 2020,59 USA | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 694 533 | 6 | Not reported | • Patient review and VTE risk stratification • Recommending TP medication selection, dose, and duration • Managing perioperative chronic antithrombotic therapies • Comprehensive discharge medication reconciliation | • ↓ Postoperative readmission rate (from 4.8% to 1.3%; P = 0.002) • ↓ VTE (from 0.6% to 0.0%; P = 0.13) |
| Khalil et al. 2021,38 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 61 65 | 7 | Apixaban Rivaroxaban Dabigatran Warfarin | • Patient education • Providing a written AF brochure that contains treatment options | • ↑ AC therapy appropriateness (from 36% to 92%; P < 0.001) |
| Kiracı et al.a 2023,33 Turkey | Quasi-experimental, Hospital | Pre-education CP intervention | 340 191 | 2 | Any TP | • Providing education to surgeons and residents • Providing optimal TP recommendations to prescribers | • ↑ TP appropriateness (from 46.8% to 61.3%; P = 0.001) • ↓ VTE (from 1.2% to 0%) • ↓ Bleeding (from 1.8% to 1.0%) |
| Lachuer et al. 2021,36 France | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 58 89 | 2.4 | VKA DOAC | • Providing training for physicians | • ↑ Prescribing practices of VKA (from 74.4% to 77.9%) and DOAC (from 74% to 82.2%) • ↑ OAC prescription appropriateness (from 74.1% to 79.8%) |
| Lee et al. 2013,62 USA | Quasi-experimental, Hospital | UC Pharmacist-managed anticoagulation clinic | 48 20 | 3 | Dabigatran | • Patient education • Patient follow-up at 2 weeks, 1 and 3 months by means of telephone or face-to-face visits • Implementing pharmacy benefits management services guidelines | • ↔ Bleeding (4.2% vs. 10%; P = 0.148) |
| Miele et al. 2017,39 USA | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 50 85 | 3 | Apixaban Dabigatran Rivaroxaban Edoxaban | • Developing and implementing pharmacist-driven protocol • Patient profile reviews • Providing an appropriate medication alternative to the prescriber | • ↑ DOAC prescription appropriateness (from 60% to 70.6%) |
| Quintens et al. 2022,60 Belgium | Quasi-experimental, Hospital | Pre-implementation Post-implementation | 466 485 | 24 | Enoxaparin Apixaban Rivaroxaban Edoxaban Dabigatran VKA | • Implementing a pharmacist-led CMA intervention | • ↓ AC-related residual potentially inappropriate prescriptions (from 74.9 to 22.5%) • ↓ Median proportion of residual PIPs (from 78.5% to 18.2%) |
| Sarika et al. 2021,32 India | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 45 45 | 5 | Heparin Enoxaparin Fondaparinux | • Implementing ACCP guideline • Patient counselling | • ↑ Appropriateness of AC therapy (from 28.9% to 62.22%) |
| Shang et al. 2021,56 China | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 240 337 | 18 | LMWH VKA Rivaroxaban | • Thrombosis and bleeding risk assessments • Consulting with physicians to formulate antithrombotic treatment protocols • Optimizing perioperative medication regimens • Evaluating and optimizing the feasibility of discharge prescriptions • Following up the thrombosis of TJA patients in the first and third months after surgery | • ↓ Deep vein thrombosis incidence (from 3.33% to 1.78%; P > 0.05) • ↔ Minor bleeding (6.67% vs. 5.93%; P = 0.720) • Improvement in TP use, administration timing, and treatment course at post-intervention (P < 0.001) |
| Sharma et al. 2024,34 England | Quasi-experimental, General practices | Pre-intervention Post-intervention | 470 498 | 14 | Apixaban Rivaroxaban Edoxaban Dabigatran Warfarin | • Medication review • Optimizing AF treatment • Patient counselling • Bleed risk assessment | • OAC prescriptions appropriateness (from 77% to 93.8%) • Successful transition from VKA to DOACs in 25.71% of patients |
| Tyedin et al. 2020,58 Australia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 130 108 | 5 | Warfarin | • Medication review • Proactively charting and monitoring • Ordering INRs using electronic prescribing software, following discussion with doctors | • ↓ Re-hospitalization (from 3.1% to 0%; P = 0.18) • ↓ Mortality (from 4.6% to 0%) |
| Yap et al. 2019,40 Malaysia | Quasi-experimental, Hospital | Pre-intervention Post-intervention | 142 144 | 5 | Fondaparinux Enoxaparin Heparin | • Developing and implementing a pharmacist-driven DVT Risk Alert Tool • DVT and bleeding risk assessment | • ↑ DVT prophylaxis appropriateness (from 14.3% to 31.3%; P < 0.05) |
| Wu et al. 2022,54 Taiwan | Quasi-experimental, Hospital | Conventional Pharmacist-managed group | 39 33 | 24 | Warfarin | • Patient and caregiver education • Developing a monitoring plan for INR and ADRs • Managing drug–drug and drug–food interactions • Participating in the ward check rounds • | • ↓ Major bleeding (1 vs. 0, in conventional vs. pharmacist-managed group) • |
| An et al. 2017,65 Japan | Retrospective cohort, Hospital | UC group Intervention group | 32 25 | 28 | Warfarin | • Monitoring of bleeding and PT-INR • Proposing to physicians to change the dose of warfarin when appropriate • Checking interaction of warfarin and supplements or foods | • ↓ Major bleeding (from 12.5% to 8%) • ↔ Minor bleeding (28% vs. 28.1%) |
| Ashjian et al. 2017,35 USA | Retrospective cohort, Health system data | UC Pharmacist-led DOAC | 129 129 | 16 | Dabigatran Rivaroxaban Apixaban | • Medication review | • ↑ DOAC appropriateness (from 81.1% to 93.7%; P = 0.016) |
| Bakey et al. 2022,41 USA | Retrospective cohort, Hospital | No pharmacist involvement Pharmacist involvement | 44 14 | 12 | Apixaban Rivaroxaban Warfarin Enoxaparin | • Patient counselling • Chart review • Evaluating OAC options • Providing additional written discharge instructions to patients | • ↑ AC appropriateness (from 70.5% to 92.9%; P = 0.046) • ↓ Readmission rates (from 20.5% to 7.1%; P = 0.424) |
| Derington et al. 2023,71 USA | Retrospective cohort, Hospital | UC AMS | 6182 4939 | 39 | Dabigatran Apixaban Rivaroxaban Warfarin | • Managing AC therapy • Evaluating DOAC appropriateness, dose-indication match, potential drug–drug interactions, needed laboratory measurements, and dose adjustments • Patient education • Ordering relevant laboratory tests | • ↔ Net clinical outcomes (major bleeding, TE, and mortality) (hazard ratio: 0.9, 95% CI: 0.7–1.0) • ↑ Major bleeding in patients with DOACs (hazard ratio: 1.5, 95% CI: 1.0–2.3) • ↓ Mortality in patients with DOAC (hazard ratio: 0.7, 95% CI: 0.6–0.9) • ↔ TE events (1.3% vs. 1.9%) • ↓ Mortality (from 13.3% to 11.6%) |
| DiRenzo et al. 2018,73 USA | Prospective cohort, Clot clinic | PCP Pharmacist-managed clinic | 17 17 | 6 | Rivaroxaban | • Patient counselling • Using the American College of Chest Physicians Guidelines (CHEST) • Reviewing patient profile • Identifying VTE symptoms and performing a physical examination | • ↔ TE events (6% vs. 6%; P = 1.000) • ↔ Hospitalization (6% vs. 6%; P = 1.000) |
| Han et al.a 2021,44 USA | Retrospective cohort, Academic health centre | Physician led CP led | 138 553 | 14 | Warfarin Apixaban Rivaroxaban Dabigatran Antiplatelet | • Implementing the best practice alert in addition to their pharmacotherapy training | • ↑ Appropriateness of OAC discontinuation during pre-procedure (from 91% to 97.4%; P = 0.001) • ↑ Appropriateness of bridging therapy (from 50% to 72.2%; P = 0.39) |
| Jones et al. 2020,74 USA | Retrospective cohort, Healthcare system data | Non-AMS group AMS | 370 90 | 42 | Apixaban Rivaroxaban Dabigatran | • Patient education followed by phone calls • Chart reviews | • ↑ Composite outcomes (bleeding, TE, mortality) (from 13.5% to 18.9%; P = 0.29) • ↑ Bleeding (from 10.5% to 18.9%; P = 0.03) • ↓ TE events (from 1.1% to 0%) • ↓ Mortality (from 1.9% to 0%) |
| Kose et al. 2018,66 Japan | Retrospective cohort, Hospital | UC Intervention group | 23 16 | 16 | Warfarin | • Monitoring of bleeding and PT-INR • Checking interaction of warfarin and supplements or foods • Advising physicians to adjust the dose of warfarin as needed | • ↔ Major bleeding (17.4% vs. 18.8%) • ↓ Minor bleeding (from 43.5% to 31.3%) |
| Kurimura et al.a 2023,67 Japan | Retrospective cohort, Hospital | Non-intervention group Intervention group | 264 132 | 48 | Warfarin Apixaban Rivaroxaban Dabigatran Edoxaban Antiplatelets | • Checking all prescription drugs from the hospital and other healthcare facilities • Pharmaceutical advice and suggesting prescription changes to physicians • Assessing adherence, adverse events, and discontinuation of unnecessary medications | • ↓ Bleeding (from 28.4% to 17.4%; P = 0.019) • ↓ TE (from 14% to 6.8%; P = 0.44) • ↑ Mortality (from 9% to 13.6%; P = 0.17) |
| Li et al. 2020,64 China | Prospective cohort, Hospital | UC PEFS | 202 179 | 6 | Rivaroxaban | • Observing potential interaction with rivaroxaban and managing bleeding and embolic complications • Evaluating patient medication adherence • Distributing paper-based medication education materials • Pharmacists kept in touch with the patients through WeChat or telephone weekly | • ↓ Bleeding (from 26.7% to 16.2%) • ↔ TE (9.4% vs. 7.8%; P = 0.675) |
| Manzoor et al. 2018,70 USA | Retrospective cohort, Hospital | NMAC PMAC | 100 100 | 15 | Warfarin | • Monitoring and managing patients receiving warfarin therapy | • ↑ Hospitalization or ED visits for patients in the NMAC group [eight times higher (OR: 7.68, 95% CI: 1.1–55.9, P < 0.05)] |
| Noor et al. 2021,69 Saudi Arabia | Retrospective cohort, Hospital | HMAC PMAC | 124 104 | 24 | Warfarin | • Patient counselling • Assessing INRs, warfarin therapy • Assessing ADRs, drug–drug interactions, or drug–food interactions • Documenting all therapeutic recommendations and prescribing a new warfarin prescription | • ↔ Bleeding (12.5% vs. 12.1%) • ↓ TE events (5.7% to 4%) • ↓ Hospitalization rates (from 9.6% to 4%) |
| Tarasiuk et al. 2018,68 USA | Retrospective cohort, Hospital | Nurse-managed group Pharmacist-managed group | 240 228 | 12 | Warfarin | • Managing patient dosing and monitoring autonomously utilizing clinical judgement | • ↓ Hospitalizations (OR: 0.29, P < 0.001) • ↓ Bleeding (from 2.1% to 1.3%) • ↓ TE (from 2.1% to 0.4%) • ↔ ED visits (39.2% vs. 43%; P = 0.402) |
| Zhang et al.b 2023,43 China | Retrospective cohort, Hospital | Control group CP services group | 162 176 | 21 | Heparin | • Assessing VTE and bleeding risk on admission and giving VTE prophylaxis recommendations • Providing medication consultation to physicians, nurses, and patients | • ↑ TP appropriateness (from 59% to 84%) • ↓ TE events (from 17% to 9%; P = 0.037) • ↓ Bleeding (from 11% to 5%; P = 0.042) • ↓ Mortality (from 28% to 14%; P = 0.001) |
Abbreviations: AC, anticoagulant; ACCP, American College of Clinical Pharmacy; ADR, adverse drug reaction; AF, atrial fibrillation; AMS, anticoagulation management service; CI, confidence interval; CMA, Check of Medication Appropriateness; CP, clinical pharmacist; DOAC, direct-acting oral anticoagulant; ED, emergency department; HMAC, haematologist-managed anticoagulation clinic; INR, international normalized ratio; LMWH, low-molecular-weight heparin; MAT-AF, medication assessment tool of atrial fibrillation; MUR, medication use review; NMAC, nurse-managed anticoagulation clinic; OAC, oral anticoagulant; OR, odds ratio; PEFS, pharmacist-led education and follow-up service; PMAC, pharmacist-managed anticoagulation clinic; PT-INR, prothrombin time–international normalized ratio; RCT, randomized controlled trial; TE, thromboembolic; TJA, total joint arthroplasty; TP, thromboprophylaxis; UC, usual care; VKA, vitamin K antagonist; VTE, venous thromboembolism.
aIncluded patients treated with antithrombotic medications (anticoagulants and/or antiplatelets).
bIncluded patients aged ≥16 years.
Studies quality assessment
Based on JBI metrics, of the RCT studies, one63 was judged to be of high quality and four studies55,57,61,72 were rated as moderate quality (see Supplementary material online, Table S2). From the quasi-experimental studies, 14 studies32–34,36–40,42,53,56,59,60,62 were rated as high quality and 2 studies54,58 were judged to be of moderate quality (see Supplementary material online, Table S3). Most cohort studies35,41,43,44,64–70,73,74 were assessed as high quality (see Supplementary material online, Table S4); only the study by Derington et al.71 was judged to be of moderate quality. Blinding was a primary concern in RCTs, as either blinding status was unclear or participants, intervention providers, and outcome assessors were not blinded in four out of five studies.55,57,61,72 Quasi-experimental studies lacked a control group and showed baseline variations, while cohort studies were limited in addressing strategies to deal with confounding factors.
Meta-analysis
Appropriateness of anticoagulant therapy
Fourteen studies, 10 quasi-experimental studies,32–34,36–40,42,60 and 4 cohort studies35,41,43,44 reported the appropriateness of AC therapy. All studies except that by Quintens et al.,60 which reported the proportion of potentially inappropriate prescriptions that remained uncorrected after 48 h, were included in the meta-analysis. The meta-analysis (Figure 2) revealed an overall significant effect in favour of the pharmacist-led intervention on the appropriateness of AC therapy (OR: 3.43, 95% CI: 2.33–5.06, P < 0.01). This remained significant when subgroup analysis was performed based on study design: quasi-experimental (OR: 3.41, 95% CI: 1.94–5.99, P < 0.01) and cohort studies (OR: 3.62, 95% CI: 2.49–5.24, P < 0.01) (Figure 2). It also remained unchanged with subgroup analyses by medication types prescribed: all ACs (OR: 2.07, 95% CI: 1.42–3.01, P < 0.01), antithrombotics (including AC and/or antiplatelet) (OR: 2.32, 95% CI: 1.29–4.16, P < 0.01), DOACs (OR: 2.29, 95% CI: 1.08–4.85, P = 0.03), OACs (OR: 5.88, 95% CI: 1.96–17.64, P < 0.01), and parenteral ACs (OR: 3.52, 95% CI: 2.36–5.24, P < 0.01) (see Supplementary material online, Figure S1, Panel A). In the Quintens et al.60 study, the proportion of potentially inappropriate prescriptions that remained uncorrected after 48 h decreased from 78.5% to 18.2% by the pharmacist-led intervention.
Forest plot of the appropriateness of anticoagulant therapy between the pharmacist-led intervention group and the usual care group.
High heterogeneity was observed among the included studies (χ2 = 55.78, P < 0.01, I2 = 78.45%), which was verified in quasi-experimental studies (χ2 = 54.50, P < 0.01, I2 = 88.21%) (Figure 2), and studies in Australia (χ2 = 17.14, P < 0.01, I2 = 94.17%) and Europe (χ2 = 33.41, P < 0.01, I2 = 92.02%) (see Supplementary material online, Figure S1, Panel B). The leave-one-out analyses revealed that Gauci et al.42 contributed the highest heterogeneity, and its exclusion decreased I2 from 78.45% to 70.15% (see Supplementary material online, Figure S2, Panel A). Furthermore, excluding the Australian study by Khalil et al.38 decreased I2 to 49% (see Supplementary material online, Figure S2, Panel B). The impact of pharmacist-led interventions in improving AC therapy appropriateness remained significant in the leave-one-out sensitivity analyses (see Supplementary material online, Figure S3). The overall OR did not change by more than 0.38 points [ranging from 3.05 (95% CI: 2.14–4.34) to 3.68 (95% CI: 2.45–5.53)]. We did not find publication bias by visual analysis of the funnel plot and Egger's test (P = 0.50).
Bleeding events
The pooled analysis of nine cohort studies,43,64–69,71,74 six quasi-experimental studies,33,53,54,56,58,62 and four RCTs55,57,61,72 showed that the pharmacist-led interventions significantly decreased the risk of total bleeding events (RR: 0.75, 95% CI: 0.58–0.96, P = 0.03) (Figure 3). In the subgroup analyses, the bleeding risk remained significantly decreased in quasi-experimental (RR: 0.74, 95% CI: 0.55–0.99, P = 0.04) and RCT (RR: 0.50, 95% CI: 0.31–0.81, P < 0.01) studies, but not in cohort studies (RR: 0.90, 95% CI: 0.64–1.26, P = 0.55) (Figure 3). Based on medication types, pharmacist-led interventions significantly decreased bleeding risk for patients taking warfarin (RR: 0.65, 95% CI: 0.45–0.94, P = 0.02) and antithrombotic medications (RR: 0.61, 95% CI: 0.41–0.92, P = 0.02), but not for those on DOACs (RR: 1.17, 95% CI: 0.48–2.82, P = 0.73) or OACs (RR: 0.77, 95% CI: 0.39–1.52, P = 0.45) (see Supplementary material online, Figure S4, Panel A). Additionally, the bleeding risk was reduced in the pharmacist-led intervention groups in both Asian (RR: 0.68, 95% CI: 0.58–0.79, P < 0.01) and European studies (RR: 0.36, 95% CI: 0.24–0.52, P < 0.01) (see Supplementary material online, Figure S4, Panel B). Substantial heterogeneity (χ2 = 88.23, P < 0.01, I2 = 74.08%) was found among the included studies, specifically in cohort studies (χ2 = 40.42, P < 0.01, I2 = 74.38%) and RCTs (χ2 = 12.76, P = 0.01, I2 = 75.78%). Using leave-one-out analyses, we found that the study by Derington et al.71 contributed the highest heterogeneity. Its exclusion decreased I2 from 74.08% to 59.39% (see Supplementary material online, Figure S5). No publication bias was indicated by Egger's test (P = 0.52) or visual inspection of the funnel plot. In the leave-one-out sensitivity analyses, the overall RR of bleeding ranged from 0.69 (95% CI: 0.55–0.87) to 0.80 (95% CI: 0.63–1.02) and remained significant at all times except when the studies by Falamić et al.72 and Liu et al.57 were excluded (see Supplementary material online, Figure S6).
Forest plot of bleeding event between the pharmacist-led intervention group and the usual care group.
Two RCTs,55,61 three quasi-experimental studies,33,54,62 and six cohort studies57,65,66,69,71,74 reported major bleeding events. Their pooled estimate indicated no significant difference in decreasing major bleeding between the pharmacist-led interventions and UC (RR: 0.99, 95% CI: 0.59–1.66, P = 0.96). Furthermore, minor bleeding events from two quasi-experimental studies,56,62 three RCTs,55,57,61 and four cohort studies65,66,69,74 showed no significant difference between the pharmacist-led intervention and control groups (RR: 0.84, 95% CI: 0.61–1.17, P = 0.31) (see Supplementary material online, Figure S7).
Thromboembolic events
Fourteen studies (2 RCTs,57,63 4 quasi-experimental studies,33,56,58,59 and 8 cohort studies43,64,67–69,71,73,74) reported thromboembolic events. The overall pooled result showed a non-significant decrease in the risk of thromboembolic events in the pharmacist-led intervention groups (RR: 0.69, 95% CI: 0.46–1.02, P = 0.07) (Figure 4). Using subgroup analyses, pharmacist-led interventions did not significantly decrease thromboembolic events in the cohort (RR: 0.80, 95% CI: 0.51–1.25, P = 0.33) and RCT (RR: 0.41, 95% CI: 0.07–2.29, P = 0.31) studies, but did in quasi-experimental studies (RR: 0.33, 95% CI: 0.11–0.98, P = 0.05) (Figure 4). Additional subgroup analysis indicated that interventions led by pharmacists did not significantly reduce thromboembolic events in American studies (RR: 0.69, 95% CI: 0.24–2.01, P = 0.49) or in patients receiving DOACs (RR: 0.82, 95% CI: 0.44–1.53, P = 0.53) and warfarin (RR: 0.78, 95% CI: 0.38–1.62, P = 0.51); however, a significant reduction was observed in Asian studies (RR: 0.63, 95% CI: 0.45–0.87, P < 0.01) (see Supplementary material online, Figure S8). Egger's test (P = 0.03) and the trim-and-fill method (see Supplementary material online, Figure S9) indicated publication bias, supported by visual analysis funnel plots (Figure 5).
Forest plot of thromboembolic event between the pharmacist-led intervention group and the usual care group.
Funnel plot indicating publication bias in thromboembolic event analysis studies.
Moderate heterogeneity (χ2 = 24.37, P = 0.03, I2 = 43.22%) was found among the included studies, specifically in cohort studies (χ2 = 17.06, P = 0.02, I2 = 54.98%). In the leave-one-out analyses, we identified that the study by Derington et al.71 significantly influenced the overall pooled estimate, increased heterogeneity among included studies, and introduced publication bias. Excluding this study, pharmacist-led interventions significantly decreased thromboembolic events (RR: 0.59, 95% CI: 0.43–0.81, P < 0.01) with no detected publication bias (see Supplementary material online, Figure S10). In the leave-one-out sensitivity analyses, the overall RR remained non-significant except when the study by Derington et al.71 or Noor et al.69 was excluded and ranged from 0.59 (95% CI: 0.43–0.81) to 0.73 (95% CI: 0.47–1.12) (see Supplementary material online, Figure S11).
Mortality
Patient deaths were reported in one quasi-experimental study,58 one RCT,55 and four cohort studies.43,67,71,74 There was no significant reduction of all-cause mortality in the pharmacist-led intervention groups (RR: 0.73, 95% CI: 0.44–1.22, P = 0.23), with substantial heterogeneity (χ2 = 12.50, P = 0.03, I2 = 70.76%) among the studies (Figure 6).
Forest plot of mortality between the pharmacist-led intervention group and the usual care group.
Hospitalization or readmission
Hospitalization or readmission was reported in two RCTs,55,63 three quasi-experimental studies,53,56,58 and five cohort studies.41,68–70,73 The pharmacist-led intervention groups showed a significant reduction (RR: 0.64, 95% CI: 0.41–0.99, P = 0.04) in hospitalization or readmission, with low heterogeneity (χ2 = 13.40, P = 0.15, I2 = 30.40%) (Figure 7). Pharmacist-led interventions significantly decreased the risk of hospitalization or readmission in quasi-experimental studies (RR: 0.35, 95% CI: 0.15–0.81, P = 0.01), but not in RCTs (RR: 0.98, 95% CI: 0.51–1.90, P = 0.96) and cohort studies (RR: 0.66, 95% CI: 0.32–1.36, P = 0.26) (Figure 7). The overall RR ranged from 0.52 (95% CI: 0.37–0.73) to 0.72 (95% CI: 0.44–1.19) in sensitivity analyses, and when four studies (Karaoui et al.,55 Liang et al.,63 DiRenzo et al.,73 and Noor et al.69) were sequentially excluded, they remained consistent with the overall RR of the main analysis (see Supplementary material online, Figure S13). No publication bias was indicated by Egger's test (P = 0.6876) and visual analysis of the funnel plot.
Forest plot of hospitalization between the pharmacist-led intervention group and the usual care group.
Discussion
AC therapies are frequently underused and inappropriately prescribed for AF and VTE, despite the existence of evidence-based guidelines.17,19 To our knowledge, this review is the first to assess the impact of pharmacist-led interventions on improving the appropriateness of AC therapy. Pharmacists optimize AC therapy through various integrated care management interventions.75 This review highlights predominant pharmacist-led interventions: patient education or counselling, medication charts or patient reviews, and developing and implementing pharmacist-driven protocols or guidelines. Patients receiving pharmacist-led care were three times more likely to be on appropriate AC therapy. This finding was consistent across study design, country, indication, and AC medication type.
Regarding the impact of pharmacist-led interventions on clinical outcomes, the findings of this meta-analysis were mixed. Prior research has identified that AC therapy is associated with preventable ADEs in hospitalized patients, most notable of which is bleeding.10,11,76 A previous review of eight RCTs and nine cohort studies27 found that the risk of bleeding was reduced by 57% in pharmacist-managed anticoagulation clinics compared with other models in patients prescribed warfarin (RR: 0.43, P = 0.006). Our meta-analysis, which focused on the post-DOAC era, found that pharmacist-led interventions reduced the risk of total bleeding events by 25%, with this being driven by a 50% reduction in bleeding in the subgroup analysis of the RCTs. Our subgroup analysis revealed that pharmacist-led interventions did not significantly reduce bleeding in DOAC studies. However, it did reduce bleeding by 35% in warfarin studies. Pharmacist-led interventions were more effective in Asian studies, which were mainly conducted on warfarin, compared with the North American studies, which mainly examined DOACs. The higher bleeding risk associated with warfarin, as compared with DOACs, could potentially be mitigated significantly through close monitoring by pharmacists in Asia involved in the intervention.
While there was also a significant 36% reduction in hospitalizations or readmissions with pharmacist-led interventions compared with UC, there were no significant differences in thromboembolic events. This might be due to the inclusion of the study by Derington et al.,71 a disproportionately large cohort study that had dissimilarities between groups at baseline that were not accounted for, as well as the non-restricted involvement of pharmacists in the UC DOAC model. Notably, when this cohort study was excluded from the meta-analysis in sensitivity analyses, pharmacist-led interventions significantly reduced thromboembolic events by 41% compared with UC (RR: 0.59, P < 0.01).
The non-significant difference in mortality observed with the pharmacist interventions in comparison with UC may be due to the low event rate. Specifically, only one RCT and one quasi-experimental study, each with a small number of patients and a short follow-up duration, could be included in this analysis. Detecting the effect of pharmacist-led interventions on this outcome requires large studies and extended follow-up periods.
The strength of our inferences relied on adhering to PRISMA guidelines for a rigorous systematic review. This included following a predesigned protocol to address the research question, employing a meticulous method for identifying relevant studies, rigorously assessing the methodological quality of the included studies, exploring sources of heterogeneity, and quantitatively summarizing the evidence. However, this review also has some limitations. Firstly, the studies included in our review showed significant statistical heterogeneity in AC therapy appropriateness and bleeding events, requiring caution in interpretation. Heterogeneity may have arisen from varying definitions of AC therapy appropriateness; some studies defined it based on guideline compliance, others on summaries of product characteristics, and some on specific local criteria. However, in our appropriateness analysis, consistent estimates across studies (irrespective of study design and country) suggest the absence of clinical heterogeneity, indicating the high generalizability of pharmacist-led interventions to various circumstances. Heterogeneity in the meta-analysis of bleeding was apparently due to one study.71 As seen in our meta-analysis, the outlier study might have affected the I2, which tends to increase when studies with very large sample sizes are included.77 Secondly, most of the studies were conducted at single centres (hospital settings) with varied interventions, and we could not confirm the most effective intervention and did not assess the cost-effectiveness of pharmacist-led interventions. Future research is needed to identify which pharmacist-led interventions are most effective and to determine their cost-effectiveness, including in community settings. Also, most studies were conducted in high-income countries, potentially limiting the generalizability of the findings to developing countries.
Conclusion
This review has shown that pharmacist-led interventions improve AC therapy appropriateness and reduce total bleeding and hospitalizations or readmissions. However, the effect on thromboembolic events and mortality remains inconclusive. Given their proven benefits, policymakers and guideline developers should consider translating pharmacist-led interventions for optimizing AC therapy, particularly in hospitals. Further research is needed to evaluate pharmacist-led interventions’ cost-effectiveness and long-term sustainability.
Acknowledgements
The authors thank the University of Tasmania research librarians for their assistance in developing the search strategy. B.K. received support from a Tasmanian Graduate Research Scholarship, covering a living allowance and tuition fees, and a top-up Pharmacy Research Scholarship from the Pharmaceutical Society of Australia (Tasmanian Branch).
Funding
No funding was received for this work.
Conflict of interest
None declared.
Author contributions
B.K.: conceptualization, screening, data extraction, formal analysis, investigation, methodology, validation, visualization, and writing the original draft of the manuscript; W.M.B., G.M.P., and C.M.: conceptualization, screening, data extraction, methodology, supervision, validation, writing, review, and editing of the manuscript. All authors read and approved the final version of this manuscript.
Data Availability
All data are available in the manuscript or the supplemental files.
