This is the first report which shows that pharmacist intervention can reduce the risk of bleeding in cardiology outpatients by conducting patient interviews and advising prescription interventions in accordance with various guidelines for anticoagulant therapy.

The rate of bleeding defined as BARC criterion greater than type 2 were 17.4% and 28.4% in the pharmacist intervention and non-intervention groups, respectively, indicating a significant 39% lower bleeding rate by pharmacist intervention. For patients that received antithrombotic drugs during the observed period, the pharmacist recommended the physician to prescribe a shorter duration of multiple antithrombotic drugs and to change a single antithrombotic drug during the maintenance phase. In the WOEST study, the rates of the patients who haemorrhaged as defined by BARC as type 2 and 3 were 31.7% and 14.3% in the triple and double therapy groups, respectively [3], indicating a 55% reduction in bleeding rate by changing the antithrombotic therapy. The AFIRE trial showed that the rate of major bleeding, as defined by the criteria of the International Society on Thrombosis and Haemostasis, was 5.75% in the rivaroxaban plus antiplatelet drug and 4.14% in the rivaroxaban monotherapy [15,16,17], indicating a 28% reduction in the bleeding rate could be expected by switching to monotherapy during the early phase. Thus, the frequency of bleeding could be reduced by up to 28–55% by facilitating the switch according to the therapy guidelines. The reduction of bleeding rate in the pharmacist intervention group in the present study was comparable to the previously reported values. Furthermore, the pharmacist intervention in this study included not only switching to monotherapy in the early stages, but also dose reduction of anticoagulant drugs based on drug-drug interactions, discontinuation of antiplatelet drugs without appropriate indication, dose reduction due to reducing renal function, and changes to an antiplatelet drug with milder bleeding risk based on the incidence of subcutaneous bleeding. These collaborative medication adjustments with physicians likely contributed in part to the reduction in bleeding events.

Our analysis showed that both elderly and pharmacist non-intervention were significant bleeding risk factors, while renal failure, liver dysfunction and the use of multiple antithrombotic drugs at the initiation of observation did not show an association with bleeding, despite that they were included in the evaluation of the HAS-BLED score. The pharmacist intervention group had a higher ratio of prescription changes for antithrombotic drugs in this study, which might directly affect the reduced risk of bleeding because the use of multiple antithrombotic drugs at the initiation of observation was not associated with the bleeding risk. Renal failure in the HAS-BLED score was defined as a serum creatinine level greater than 2.26 mg/dL, which is present in patients with end-stage renal failure such as haemodialysis [18]. Eight (2.0%, data not shown) and five patients (1.3%, data not shown) among the total patients in this study had renal failure and liver dysfunction, respectively, and renal failure and liver dysfunction were not detected as bleeding risk factors due to the small number of applicable patients. Hypertension and concomitant alcohol intake were not assessed in this study. Although labile INR was also not evaluated in this study, the number of prescription changes of warfarin tended to be greater in the pharmacist intervention group. Changing the dose or discontinuation of warfarin may lead to better control of INR within the optimal range, which may affect bleeding events.

This study found no effect of pharmacist intervention on the mortality of eligible patients. A systematic review reported that pharmacist interventions in patients with heart failure reduced the heart failure hospitalisation rates, but had no effect on reducing mortality [19]. Furthermore, pharmacist intervention as part of a multidisciplinary team reduced the hospitalisation rates, but the pharmacist-only intervention did not lead to a reduction in the hospitalisation rates because the pharmacist-only intervention did not achieve the complete therapeutic management and education of the patient [19]. Since the pharmacist intervention evaluated in our study was not a multidisciplinary team-based intervention, the pharmaceutical outpatient clinic only could not reduce the rate of re-hospitalisation for heart failure. Although it is challenging to intervene with a medical team in an outpatient setting, comprehensive care in the community is important to improve patient outcomes, and future development in this area is expected.

Various intervention methods by pharmacists have been reported [20]. In our study, the pharmacist intervention was conducted by attending the physician consultations in the outpatient clinic. Potential obstacles to pharmacist-physician collaboration were reported as limited trust-building associated with less direct communication, limited communication methods for prescription suggestions and limited discussion time [20]. Previous reports demonstrated that the rate of prescription change was higher in the group where prescription issues were discussed directly between the physician and pharmacist in conference than in the group where the issues were discussed only using written communication [21]. Moreover, providing pharmaceutical information and instruction directly to the patients in a short period of time by the community pharmacists was reported to result in a decrease in HbA1c and improvement in systolic blood pressure [22]. Our pharmacist interviewed patients to check their adherence and the status of adverse events before the physician consultations and intervened in their prescriptions in collaboration with the physician as well as an educational approach for patients at high risk of bleeding [23], which may have led to the improvement of their bleeding event risk. In addition, although the Kaplan–Meier curves of the two groups were similar in the early part of the observation period, the pharmacist intervention group had a lower incidence rate trend in the later part (Fig. 3). The Japanese Circulation Society published a guideline focused on the update antithrombotic therapy in patients with coronary artery disease in 2020 [8] following the results of the AFIRE trial [17], near the third year of the present study observation period. This guideline had a major impact on the antithrombotic therapy including aggressively shortening the duration of multiple concomitant anticoagulation drugs. The pharmacist who provided the intervention in this study suggested both short-term multiple combination therapy and switching monotherapy as possible, considering the risk of bleeding and thrombosis. This change in clinical recommendations can be a turning point in significant changes in clinical outcomes. In addition, the pharmacist carefully adjusted the dosage of antithrombotic drugs according to the long-term fluctuations in renal function during the therapeutic period. The median of the pharmacist intervention period was 224 days, therefore, the interventions described above might result in long-term improvements in clinical outcomes corresponding to the result of the cumulative incidence of bleeding.

There are some limitations in this study. First, this is a small-scale, single-centre, retrospective study and the bias of the institution and pharmacist assigned to the study cannot be eliminated because the intervention was performed by a single specific pharmacist. The results of this study may have been influenced by the experience level of the pharmacist, because this pharmacist had many years of experience in charge of cardiology wards and had professional expertise in this field. And, the statistical power of the post-hoc analysis for the primary end-point was 62.7%. Therefore, the results of this study cannot be directly generalised to all clinical setting for pharmaceutical outpatient clinic of cardiovascular internal medicine. Second, bleeding events defined as BARC was determined from the descriptions in the electronic medical record by researchers. Prospective intervention studies are needed to verify these results. Third, events that occurred outside of the hospital may not have been fully evaluated. Outcomes for patients who no longer visited the hospital during the observation period were evaluated by the report records obtained from the external institutions. Forth, the number and duration of pharmacist intervention varied in this study, and the prescription changes in this study were only evaluated during the observation period in 2017, when the pharmacist intervention for outpatients started.