Statistical analysis
Descriptive and statistical analyses were carried out with IBM SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA). The Kolmogorov–Smirnov test was used to test the normality of the continuous variables. Histograms were evaluated to assess normality. Categorical variables were expressed as percentages and continuous variables as medians with interquartile ranges (IQRs). Differences in the continuous variables among the four DOACs were evaluated using the Kruskal-Wallis test. The Chi-square test was used to compare categorical variables. A two-sided significance level of 0.05 was used. Binary logistic regression analysis was conducted to determine factors associated with the under- and overdosing of the DOACs. Factors that were significantly associated in the univariable analysis (p<0.1) were included in the multivariable logistic regression analysis. For the logistic regression models a corrected p-value <0.05 (Benjamini-Hochberg correction for multiple comparisons) was used, goodness of fit was assessed (R2, receiver operator characteristics curve) and residuals were reviewed. The odds ratios (OR) were reported with their 95% confidence intervals (CI).
Results
A total of 1688 consecutive and unique AF patients were included for which the characteristics per DOAC are shown in table 1. Apixaban was the most prescribed DOAC (34.7%) followed by edoxaban (32.7%), rivaroxaban (23.8%) and dabigatran (8.8%). Apixaban users were significantly older (median 81.0 years), weighed significantly less (median 70.1 kg) and had a statistically lower renal function (median 52.0 mL/min) compared to users of another DOAC. Moreover, apixaban users were diagnosed more often with arterial hypertension (73.0%) and heart failure (HF) (35.5%). Combination with at least one antiplatelet drug was observed in 23.6% and 21.8% of rivaroxaban and apixaban users, respectively. Dual antiplatelet therapy was the highest in the apixaban group (3.2%).
The overall inappropriate prescribing rate in this study was 16.9% (n=286) with the highest underdosing and overdosing rates seen in the apixaban (14.2%) and rivaroxaban (9.7%) group, respectively (Figure 1). Overall, underdosing, overdosing and contraindications were seen in 9.7%, 6.9% and 0.4% of patients, respectively.
Considering the four DOACs together (see Table 2), a body weight <60 kg (adj. OR 0.46, 95% CI 0.27-0.77), the use of edoxaban compared to rivaroxaban (adj. OR 0.42, 95% CI 0.24-0.74), undergoing surgery (adj. OR 0.57, 95% CI 0.37-0.87), and being DOAC naive (adj. OR 0.45, 95% CI 0.29-0.71) were associated with a significantly lower odds of underdosing. On the other hand, having a bleeding history (adj. OR 1.86, 95% CI 1.24-2.80) and the use of narcotic analgesics (adj. OR 1.67, 95% CI 1.13-2.46) were associated with a significantly higher odds of underdosing. Determinants associated with a higher odds of overdosing were renal impairment (adj. OR 11.29, 95% CI 6.23-20.45) and a body weight <60 kg (adj. OR 2.34, 95% CI 1.42-3.85), whereas the use of dabigatran (adj. OR 0.24, 95% CI 0.08-0.71) and apixaban (adj. OR 0.18, 95% CI 0.10-0.32) were associated with a lower odds of overdosing compared to rivaroxaban.
For rivaroxaban (adj. OR 100.95, 95% CI 23.23-438.70) and edoxaban (adj. OR 3.25, 95% CI 1.49-7.12) users, a decreased renal function was associated with a higher odds of overdosing compared to patients with a normal renal function. For rivaroxaban, we observed a lower odds of underdosing in patients with a CrCl <50 mL/min (adj. 0.17, 95% CI 0.06-0.48).
A lower body weight was a risk factor for underdosing in apixaban users (adj. OR 0.26, 95% CI 0.12-0.55), whereas this factor was associated with a higher odds of overdosing in the edoxaban group (adj. OR 4.16, 95% CI 1.97-8.77). Having a history of bleeding (adj. OR 2.14, 95% CI 1.22-3.75) in apixaban users was associated with a higher odds of underdosing.
Of the 286 identified AF patients with an inappropriate DOAC dose, the physician was contacted by telephone in 226 (79.0%) cases of which 131 and 90 calls concerned underdosed and overdosed prescriptions, respectively. Five cases of dabigatran use and a CrCl <30 mL/min were classified as a contraindication. In addition, telephone calls were conducted for eight cases concerning drug interactions (not shown in figure 2) with rifampicin (n=5), carbamazepine (n=2) and cyclosporine (n=1). Physicians accepted the pharmacists’ advice in 179 cases (79.2%) which consisted of 92 (51.4%) advices for underdosing, 82 (45.8%) for overdosing and 5 (2.8%) for contraindicated dosages. Regarding the non-accepted interventions (n=47), it concerned 39 (83.0%) underdosed and 8 (17.0%) overdosed cases. The advices were effectively implemented, as evidenced by a correction of the prescription in the patient’s electronic medical record, for 75 out of 92 (81.5%) underdosed, 69 out of 82 (84.1%) overdosed and 4 out of 5 (80.0%) contraindicated cases, a total of 148 cases (65.5%). Non-implemented advices, for which the reasons are listed in figure 2, were observed for 15 underdosed, 11 overdosed and 1 contraindicated case. For 2 underdosed as well as 2 overdosed cases we were unable to track whether the intervention was implemented or not
Discussion
Despite their increased use and substantial research focusing on their effectiveness and safety in clinical practice, appropriate DOAC prescribing remains a problem as also evidenced by our study24, 25. Inappropriate first DOAC prescriptions were found in 16.9% of the patients, with underdosing being more prevalent than overdosing. This is in line with a previous study conducted at our institution 3 as well as other studies1, 2, 7-10. The dosing recommendations differ depending on the DOAC, with some of them requiring dose reductions based solely on renal function and others taking additional criteria into consideration. The high underdosing rate found in apixaban users seems clinically relevant as underdosing with this DOAC was reported to be associated with a nearly 5-fold increased stroke risk in AF patients26. On the other hand, it has been reported in a prospective study that hydrophilic drugs like DOACs in frail older patients with low muscle mass can lead to supratherapeutic DOAC plasma levels, placing them at higher risk for major bleeding complications27.
Among patients prescribed DOACs, those receiving apixaban had a lower renal function, were older and weighed significantly less compared to users of other DOACs. According to recent literature, including a systematic review in frail AF patients, apixaban was associated with a lower risk of major bleedings compared to rivaroxaban and dabigatran, although their effectiveness was comparable 28-31. As shown in a registry of rivaroxaban users, renal impairment was a risk factor for experiencing major bleeding events 32. The risk of major bleedings was comparable between apixaban and edoxaban29. Intra DOAC comparisons still need to be confirmed by findings from randomized controlled trials that are expected to be released in the upcoming years e.g., the DARING-AF trial and DANNOAC-AF trial 28, 29.
According to the literature, physicians may place more value on the avoidance of bleeding in high risk AF patients, especially in those with a high fall and bleeding risk 3, 33, 34. Results of previous studies suggest an increased risk of falls and fractures among older adults using opioids 35, 36. Walenga and colleagues describe that the relative rate of clinically relevant bleedings with the use of opioids was nearly twofold greater with low dose rivaroxaban compared with enoxaparin 37.
The higher OR for underdosing apixaban in patients with a lower weight may be related to the fact that weight is often used as a single criterion for dose reduction in apixaban, while this is only to be adapted in combination with an age ≥80 years and/or serum creatinine ≥1.5 mg/dL.
The association of weight with overdosing in edoxaban users is mainly due to younger patients with a weight <60 kg who were hospitalized to undergo an ablation.
The association between a reduced CrCl and the higher odds to receive a supratherapeutic dose as found for rivaroxaban and edoxaban users could be due to a fluctuating renal function or physicians being unaware of the CrCl threshold for dose adjustment 2.
A high percentage of the advices by the clinical pharmacists was immediately accepted (79.2%) and effectively implemented (65.5%) by the physicians. The accepted advices consisted for 79.9% of prescriptions for DOACs that were initiated prior to the hospital admission. Physicians acknowledged that they did not consistently check the DOAC dose upon admission due to a lack of time. A history of bleeding and a fluctuating renal function were the most important factors why advices to adjust the dose were not accepted. Some physicians, less familiar with anticoagulation, preferred to discuss the proposed dosage adjustment with the GP/cardiologist. In some cases, as depicted in figure 2, renal function altered during hospitalization so no adjustment was needed anymore, even if the advice was already accepted. Other important reasons retrieved from the medical record for non-implementation of accepted advices were the patients’ bleeding risk (n=2) and concomitant use of antiplatelet drugs (n=2). Despite the initial acceptance, physicians chose to keep the reduced DOAC dose in these cases after consultation with the head of department. After implementation of the recommended dose adjustments, the total inappropriate prescribing rate according to the SmPC was 8.2% vs. the initial 16.9%, keeping in mind most of these off-label doses were intentional due to the abovementioned reasons.
Some institutions have an integrated pharmacist-led DOAC service or antithrombotic stewardship programs with the aim of improving the safety and efficacy of DOAC use through identification and resolution of dosing errors, patient education, improved patient follow-up and laboratory monitoring 16, 20, 22, 38, 39. Studies showed that such services increase appropriate DOAC dosing at baseline and follow-up, which is in line with the recommendations of organizations like EHRA and the American College of Cardiology (ACC)38, 40, 41. Clinical pharmacists are well positioned to provide recommendations regarding the DOAC choice and dose, presence of a contraindication as well as potential drug-drug and drug-disease interactions 20. They are also able to inform physicians of patients who may not be the best candidates for DOACs (e.g., extremes of body weight, severe renal impairment)16. In addition, they also provide motivational interviewing to promote patient understanding of DOAC therapy and emphasize the importance of DOAC adherence 38.
Involvement of clinical pharmacists should be more mainstream in the hospitals, especially for HRM such as DOACs. Quintens et al. highlighted the greater impact of advices given by telephone since half of the advices were not accepted/read if only an electronic note was left20. In addition, a study of Dreijer et al. conducted in patients using anticoagulants showed that implementation of a multidisciplinary antithrombotic team over time significantly reduces adverse drug reactions (ADRs) like bleedings or thrombotic events from hospitalization until three months after hospitalization and resulted in a lower all-cause mortality 42.
A useful recommendation for physicians could be to clearly document the rationale for deviating from the approved dose in the patient’s medical record to avoid unnecessary phone calls. Physicians should be aware of the most common mistakes concerning DOAC prescriptions14.
To our knowledge, this is the first study to assess determinants for DOAC under- and overdosing in AF patients, nevertheless there are limitations to our study. First, this was a single center study, therefore the results may not be generalizable to other populations. Due to the cross-sectional nature of our study, only the initial DOAC prescription of a patient was taken into account. However, the conditions necessitating a DOAC change or dose adaptation during hospitalization were only present in a minority of patients. We also did not examine the effect of inappropriate dosing and the dose adaptations on the thromboembolic/bleeding event rate. In addition, our results may be biased by inaccurate or incomplete information in the electronic medical records. Further, this study was neither directed to identify high risk AF patients who did not receive oral anticoagulants, nor to assess DOAC adherence.
Additional research regarding the long term effects of the clinical pharmacists’ interventions is needed. Lastly, it would be of interest for future research to focus on the interventions with the highest impact on patient outcomes and to investigate which patients would benefit most from the implementation of a multidisciplinary clinical DOAC service.
In conclusion, in this study where for the first time determinants for under- and overdosing of DOACs were assessed, inappropriate DOAC prescribing was found to be still common. Interventions by clinical pharmacists can reduce this burden. We recommend that clinical services led by pharmacists play a greater role in assisting physicians during the prescription process of high risk medications such as DOACs, in order to reduce the number of inadequate prescriptions.