Inappropriate doses of direct oral anticoagulants in real-world clinical practice: prevalence and associated factors. A subanalysis of the FANTASIIA Registry

Abstract

Aims

To describe the prevalence and associated factors of inappropriate doses of direct oral anticoagulants (DOAC) in a national registry of patients of real clinical practice.

Methods and results 

Five hundred and thirty outpatients with atrial fibrillation treated with DOAC were included in a prospective, national, multicentre study. The appropriateness of the doses of DOAC was defined according to the recommendations of the European Heart Rhythm Association. Mean age was 73 ± 9 years, with a 46% of women. Two hundred and sixty-seven patients were prescribed dabigatran, 190 rivaroxaban, and 73 apixaban. A total of 172 patients (32%) did not receive the appropriate dose: 93 patients received a lower dose (18%) and 79 patients a higher dose (15%). In the comparisons among the subgroups of inappropriately low, appropriate, and inappropriately high dose, we observed significant trends to older age (69 ± 8 years vs. 73 ± 10 years vs. 77 ± 6 years), more frequent female sex (37% vs. 46% vs. 59%), antiplatelet drugs (5% vs. 8% vs. 25%), rivaroxaban (14% vs. 38% vs. 53%), and apixaban use (5% vs. 15% vs. 19%), higher CHAD₂DS₂-VASc (3.00 ± 1.38 vs. 3.58 ± 1.67 vs. 4.59 ± 1.44) and HAS-BLED scores (1.83 ± 0.87 vs. 1.92 ± 1.07 vs. 2.47 ± 1.13), lower body mass index (30 ± 6 kg/m² vs. 29 ± 4 kg/m² vs. 28 ± 4 kg/m²) and glomerular filtration rate (74 ± 27 mL/min vs. 70 ± 22 mL/min vs. 63 ± 16 mL/min), and lower frequency of dabigatran use (81% vs. 47% vs. 28%) (all comparisons P ≤ 0.01).

Conclusion

In this real-life study, 32% of patients received an inappropriate dose of DOAC. Several clinical factors can identify patients at risk of this situation.

Introduction

The direct oral anticoagulants (DOAC, dabigatran, rivaroxaban, apixaban, and edoxaban) have been developed to overcome some of the limitations of vitamin-K antagonists (VKA), such as an unpredictable anticoagulant response, the need for regular therapeutic monitoring and dose adjustments, as well as drug or food interactions. The direct oral anticoagulants have demonstrated similar or higher efficacy in thrombo-embolic prevention in patients with atrial fibrillation (AF) as well as improved safety.^1–4 The advantages of DOAC over VKA are a fixed dose regimen, the absence of drug monitoring, and fewer drug–drug and drug–food interactions. However, determining the appropriate dose for a DOAC is dependent upon several patient-specific factors such as age, baseline renal and hepatic function, weight, bleeding risk, and concurrent medications. The recommended dose for each DOAC is settled by the summary of product characteristics (SmPCs), approved by European Medicines Agency.^5–7 Deviations from the recommended use have been reported in patients with AF.^8–13

The European Heart Rhythm Association (EHRA) issued a practical guide on the use of DOAC in patients with AF.¹⁴ This document recommends the reduction of DOAC dose in some specific situations that include, but are not limited to, those described in the SmPCs. However, to the best of our knowledge, the application of these recommendations has not been evaluated so far in daily clinical practice.

So, the aim of the present study was to describe the prevalence of inappropriate doses of DOAC, evaluated according to the EHRA recommendations, in a national registry of real clinical practice, and to investigate the factors associated with this condition.

Methods

The FANTASIIA registry

The FANTASIIA Registry (Spanish acronym for Atrial fibrillation: Influence of anticoagulation level and type on stroke and bleeding event incidence) is a prospective, observational, national, multicentre study that included an adult population of Spanish patients with AF on anticoagulant treatment.^15,16 The main aim of the registry was to evaluate the incidence of thrombo-embolic and bleeding events in an unselected population of patients with AF over 3 years, specifically with reference to the use and type of antithrombotic agent, VKA and DOAC, as well as anticoagulation adjustment (in patients receiving VKA). Patients were recruited at 50 outpatient clinics by 81 investigators (81% cardiologists, 11% primary care physicians, and 8% internists) from 1 June 2013 to 15 October 2014. The clinics were chosen by the steering committee, to obtain a representative sample of the whole national territory, and were located throughout Spain, and investigators were selected among physicians with knowledge and special interest in AF. These criteria relied on the personal judgement of the members of the steering committee, and there were neither specific qualifying requisites nor a formal selection process. Inclusion criteria were patients older than 18 years with a diagnosis of AF (excluding patients with rheumatic mitral valve disease or valvular prostheses) who had been receiving anticoagulant therapy (by design, 80% VKA and 20% DOAC, i.e. dabigatran, rivaroxaban, or apixaban) for at least 6 months before enrolment. The patients’ clinical and laboratory data were collected in an electronic case report form. CHADS2, CHA2DS2-VASc, and HAS-BLED scores were calculated for each patient for assessing thrombotic and bleeding risk, as well as the Charlson index to evaluate associated comorbidities. This late score contains 19 issues including diabetes with diabetic complications, congestive heart failure, peripheral vascular disease, chronic pulmonary disease, mild and severe liver disease, hemiplegia, renal disease, leukaemia, lymphoma, metastatic tumour, and acquired immunodeficiency syndrome, each of which is weighted according to their potential influence on mortality. The patients were managed according to routine clinical practice. The FANTASIIA Registry complies with all the principles of the Declaration of Helsinki, and the study protocol was approved by the Clinical Research Ethics Committee at Hospital Universitario de Alicante and by all the local ethics committees. All study participants signed the informed consent.

Design of the study

This is a retrospective analysis of a prospectively collected database. For this analysis, cross-sectional baseline data from the FANTASIIA enrolment visit of patients treated with DOAC were used. Criteria proposed by the EHRA practical guide on the use of DOAC in patients with AF¹⁴ were used for the determination of the appropriate dose (Table 1). This document proposes three levels of alert for drug–drug interactions or other clinical factors that may affect DOAC plasma levels or effects: (i) ‘red’ interactions, precluding the use of a given DOAC in combination (i.e. ‘contraindication’ for use); (ii) ‘orange’ interactions, with the recommendation to reduce the DOAC dose, since they result in changes of the plasma levels or effect of DOAC that could potentially have a clinical impact; and (iii) ‘yellow’ interactions, with the recommendation to keep the original dose, unless two or more concomitant ‘yellow’ interactions are present. Two or more ‘yellow’ interactions need expert evaluation and may lead to the decision of not prescribing the drug (‘red’) or of adapting its dose (‘orange’). For this analysis, two or more ‘yellow’ interactions have been considered as an indication for reduced dose. Patients were classified as receiving a standard dose (i.e. dabigatran 150 mg/12 h, rivaroxaban 20 mg/24 h, or apixaban 5 mg/12 h) or a reduced dose (i.e. dabigatran 110 mg/12 h, rivaroxaban 15 mg/24 h, or apixaban 2.5 mg/12 h) of DOAC. Every patient was also categorized as requiring a standard dose of DOAC (no interactions or only one ‘yellow’ interaction) or a reduced dose (one ‘orange’ interaction or two or more ‘yellow’ interactions). Finally, those patients receiving the required dose of DOAC were classified as ‘appropriate dose’ and the rest of series as ‘inappropriate dose’: those requiring a standard dose but receiving a reduced dose were regarded as ‘inappropriately low dose’, and those requiring a reduced dose but receiving a standard dose were considered as ‘inappropriately high dose’.

Statistical analysis

Quantitative variables were described using the mean ± standard deviation or median (interquartile range), depending on whether they followed a normal distribution, which was tested with the Kolmogorov–Smirnov method. For between-group comparisons, the Student’s t-test, Mann–Whitney U test, one-way analysis of variance, and Kruskal–Wallis test for continuous variables and the χ² test for qualitative variables were used, as appropriate. Finally, multivariate logistic regression was used to investigate factors independently associated with inappropriately low or high doses. Statistical significance was defined as P < 0.05. Statistical analyses were performed with SPSS statistical package version 12.

Results

Study population

A total of 530 patients were included in the study, with a mean age of 73 ± 9 years and 46% of women. General features of the population are described in Table 2. Most of them presented hypertension, a half hypercholesterolaemia, and more than a fourth diabetes mellitus. More prevalent comorbidities were cerebrovascular disease, chronic pulmonary obstructive disease, and chronic renal failure. Structural heart disease was present in nearly 40%, and most of them were in functional class I or II of the EHRA.

Doses of direct anticoagulants

A total of 267 patients were prescribed dabigatran, 190 rivaroxaban, and 73 apixaban. The standard dose of DOAC was prescribed to 296 patients (56%) and the reduced dose to 234 patients (44%). According to EHRA recommendations, 310 patients should have received a standard dose (58%) and 220 (42%) a reduced dose. A total of 358 patients received the appropriate dose (68%) and 172 (32%) an inappropriate dose: 93 (17%), an inappropriately low dose, and 79 (15%) an inappropriately high dose. The standard dose was prescribed less frequently than recommended for dabigatran and more frequently than recommended for rivaroxaban and apixaban (Figure 1); and the reduced dose was prescribed more frequently than recommended for dabigatran and less frequently than recommended for rivaroxaban and apixaban (Figure 2). The inappropriately low dose was more frequent than the inappropriately high dose for dabigatran, and the opposite was observed for rivaroxaban and dabigatran (Figure 3).

Factors associated to inappropriate dose of direct anticoagulants

Associations of baseline clinical features with subgroups of dose of DOACs are listed in Table 2. In the comparisons among the subgroups of inappropriately low, appropriate and inappropriately high dose, we found significant trends to older age, more frequent female sex, living with relatives, coronary disease, previous acute coronary syndrome, and aortic valvular disease. We also observed significant trends to higher values of CHADS₂, CHA₂DS₂-VASc, and HAS-BLED scores and lower values of weight, height, body mass index, haemoglobin and creatinine clearance in patients with inappropriately low, appropriate, and inappropriately high doses of DOAC. The percentage of prescription of antiplatelet agents, amiodarone, apixaban, and rivaroxaban was progressively and significantly higher in the groups of inappropriately low, appropriate, and inappropriately high doses of DOAC, and the opposite was observed for dabigatran use.

In multivariate analysis, the prescription of inappropriately low dose was associated with younger age, lower CHADS₂ score, higher body mass index, and dabigatran (Table 3). The use of inappropriately high dose was associated with lower Charlson comorbidity index, higher CHA₂DS₂-VASc score, the presence of hypertrophic cardiomyopathy, aortic valvular disease, aldosterone antagonists, antiplatelet agents, apixaban and rivaroxaban, and the absence of angiotensin receptor blockers (Table 3). Additional analysis was performed after excluding CHADS₂ and CHA₂DS₂-VASc scores, as those scores include age, a variable already included in the logistic regression model. The new models found the same variables as independent predictors for inappropriately low dose, but for inappropriately high dose, Charlson comorbidity index was excluded of the model for lack of statistical significance. Odds ratios were similar to the initial analysis.

Discussion

The main finding of our study was that nearly one-third of patients received an inappropriate dose of DOAC in this ‘real-life’ registry of daily clinical practice, according to EHRA recommendations, and that both over-treatment and under-treatment are common in this setting. This finding is even more striking if we take in consideration that investigators were selected by the steering committee as physicians with knowledge and special interest in AF. So, although DOAC have the theoretical advantage of simplified dosing vs. VKA, finding the appropriate dose is not always achieved in real-world daily clinical practice. Previous studies, either monocentric,^9–11 based in digitalized administrative registries,^8,12 or performed in national registries from the USA¹³ found frequencies of inappropriate dosing from 14% to 45%. Differences in the criteria for establishing appropriate dose, time from initial commercialization of the drugs in the specific countries, and so, physicians’ experience, and variability of practice among different centres could explain, as least partly, these differences. However, our results are in concordance with all these previous investigations in the sense that, regardless criteria and setting, inappropriate dosing is a relevant issue, which affects to a significant proportion of patients. As the standard by which the appropriate dosing was determined in this study was the EHRA guidelines,¹⁴ another possible explanation for our findings can be an insufficient knowledge or lack of confidence in these recommendations by the physicians. We did not specifically study the level of awareness of the investigators for these guidelines, so this issue requires further research.

In our study, the reduced dose of DOAC was prescribed to 44% of patients (57% for dabigatran, 34% for rivaroxaban, and 23% for apixaban). In the whole sample, this figure is similar to the 42% of patients (37% for dabigatran, 49% for rivaroxaban, and 37% for apixaban) who should have received a reduced dose according to EHRA recommendations, although we found significant discrepancies at individual patient level, and according to each DOAC drug. Data from Danish registries show that from 1 August 2011 up to 30 November 2015, 26 242 patients first prescribed a DOAC with a standard dose presented AF,¹⁷ and in a slightly longer period (up to 28 February 2016), 9275 patients first prescribed a DOAC with a reduced dose presented AF¹⁸; so, 28.5% of new prescriptions of DOACs for AF were reduced doses (27.3% for dabigatran, 20.4% for rivaroxaban, and 29.6% for apixaban), assuming a small supraestimation due to the slightly longer inclusion period for the reduced doses group. The higher percentages for reduced doses in our study (except for apixaban) could be explained for differences in study populations and for methodology issues: we did not include all patients prescribed DOAC in Spain, but a small sample, and patients of FANTASIIA registry should have received anticoagulant treatment for at least 6 months prior to inclusion and were not new users. However, Danish registries do not allow for accurate identification of patients with an EHRA indication for a reduced DOAC dose, as it has been the case in the present study.

Previous studies have found an association between advanced age and inappropriately high doses of dabigatran⁸ or inappropriate doses of all DOAC.¹¹ We have found a significant trend to higher age in the comparison of groups of under-treatment, appropriate dose, and over-treatment, and in the multivariate analysis, younger age was predictor of inappropriately reduced dose. In the respective SmPC, age ≥ 80 years is a criterion for reduced dose for dabigatran⁵ and one of the criteria to take into account for reducing apixaban dose.⁷ And, although not a criterion for reduced dose in the SmPC, EHRA recommendations include age ≥ 75 years as one of the ‘yellow’ criteria for reduced dose, which associated with other ‘yellow’ criteria, could imply a dose reduction.¹⁴ So, it is not surprising the association of age with inappropriate dosing. Younger patients usually should receive standard doses of DOAC as the indications for dose reduction are less frequent in these populations. So the risk of receiving a wrong prescription in this group rarely will be to receive a higher dose than recommended, and it make sense that the main risk of these patients will be to receive an inappropriately low dose.

Reduced creatinine clearance or kidney disease has been associated with inappropriate doses of DOAC in previous studies.^8,9 Creatinine clearance has shown a significant trend to lower values in the comparison of groups of under-treatment, appropriate dose, and over-treatment in the present study. Patients with high creatinine clearance should receive more frequently standard doses of DOAC, so the main risk of inappropriate prescription in these patients hardly will be to receive a higher dose than recommended, but just the opposite, as we have found in our study. For the assessment of renal function, we calculated creatinine clearance with the Cockroft–Gault formula, the method used in the pivotal trials of DOAC.^1–4 Differences with other methods of assessing renal function have been shown to have implications in dosing of DOACs,¹⁹ and could have had a role in the frequency of using inappropriate doses, as methods that not include patient weight, as modification of diet in renal disease, are widely used by clinical laboratories in Spain.

We also have found a significant trend to lower values of weight, height, and body mass index, in the subgroups of patients with inappropriately low, appropriate, and inappropriately high dose of DOACs, and multivariate analysis identified higher body mass index as an independent predictor of the prescription of inappropriately low dose. Patients with high weight, height, and body mass index usually should receive standard doses of DOAC, so the main risk of wrong prescription in these populations probably will be to receive an inappropriately low dose and not a high dose. Low body weight (≤60 kg) is one of the criteria for reduced apixaban dose in the SmPC,⁷ and one of the EHRA ‘yellow’ criteria for reduced dose of DOAC,¹⁴ but it could possibly be not taken into account by many physicians when adjusting DOAC dose, although body weight can have significant impact in effectiveness and safety of DOAC.²⁰ Similar reasons could explain the association found between inappropriate dosing and antiplatelet use and higher HAS-BLED scores. According to the EHRA guideline,¹⁴ these conditions are ‘yellow’ criteria for reduced dose of ACOD. However, this is not stated in the SmPC^5–7 that only recommend ‘caution’ in these situations, and perhaps physicians do not fully bear in mind this guideline recommendation.

However, the association among DOAC drug and inappropriate dosing found in this study is intriguing. We observed that dabigatran was mainly associated with under-treatment, and apixaban and rivaroxaban, with over-treatment, either in univariate and multivariate analysis. Dabigatran was the first DOAC introduced in Spain, and the pivotal clinical trial of this drug fully studied both doses, with similar efficacy and higher safety than warfarin for the 110 mg twice daily dose.¹ Perhaps an excessively prudent strategy of dosing, looking for lower bleeding rates and fearing adverse events in a new drug, and therapeutic inertia in the follow-up, could explain the prevalence of under-treatment with this drug. It is possible that the wider criteria for reducing dose in the EHRA recommendations¹⁴ vs. the more restrictive in the SmPC for rivaroxaban⁶ and apixaban⁷ could explain the association of apixaban and rivaroxaban with over-treatment found in our study (e.g. many patients receiving standard dose classified as requiring reduced dose according to EHRA recommendations). This point could be a relevant issue, as real-world studies have found that apixaban 2.5 mg twice a day was associated with a trend towards higher rates of ischaemic stroke/systemic embolism compared with warfarin,¹⁸ and the authors suggest that this discrepancy between clinical trial outcomes² and data from routing care could be related to inappropriate prescribing patterns of reduced dose of apixaban in clinical practice. Actually, the pivotal trials of rivaroxaban² and apixaban³ did not include enough patients with the reduced doses for a full assessment of their efficacy and safety. Our data suggest that a strict application of EHRA guideline would have resulted in a wider prescription of the lower dose of apixaban, and the clinical implications of such strategy have not been properly studied.

Limitations

Our study has some limitations. Data on edoxaban, which was introduced in Spain after finishing the inclusion period, were not available in our study. The associations of hypertrophic cardiomyopathy and aortic valvular disease with the use of inappropriately high dose found in the present study should be taken cautiously because of the low number of patients with these conditions included in the sample. We also have not performed an analysis of appropriateness considering only the SmPC, as our main objective was specifically to assess the compliance with EHRA recommendations, which offer additional guidance in prescribing DOAC in those situations where the SmPC only advise caution, e.g. high-bleeding risk. Finally, our analysis is centred in the adherence to these recommendations, but we do not have data on embolic or thrombotic risk associated with using inappropriate doses of DOACs. Information about the prognostic impact of inappropriate dose of DOACs is very scarce, with a small study showing a trend toward a higher risk of adverse drug events in the group with inappropriate prescription of the drugs,¹⁰ and a large real-world observational study expressing concern about a non-significant trend to higher rate of embolic events in patients receiving a reduced dose of apixaban, but with an incomplete assessment of appropriateness of dosing,¹⁸ so more information is required regarding this issue.

Conclusion

In this real-life, prospective, nationwide registry, nearly one-third of patients with AF treated with DOAC received an inappropriate dose, according to EHRA recommendations. Several clinical factors can identify patients at risk of over-treatment or under-treatment with DOAC. This information could be of interest in order to initiate educational or management strategies directed to improve the correct use of these drugs.

Funding

This work was supported by an unrestricted investigational grant from Pfizer/Bristol-Myers Squibb and by grants from Institute of Health Carlos III FEDER (RD12/0042/0068, RD12/0042/0010, RD12/0042/0069, RD12/0042/0049, and RD12/0042/0063).

Conflict of interest: none declared.

References

Author notes