Article Text
PDF
Abstract
Objective Antithrombotic drugs are widely used in the prevention and treatment of cardiovascular diseases; yet, nationwide long-term usage trends remain unexplored. We examined long-term trends in the use of antithrombotic drugs in Denmark.
Methods Using nationwide prescription data, we obtained information on primary care use of antiplatelet drugs, vitamin K antagonists (VKA), non-vitamin K antagonist oral anticoagulants (NOAC), heparins and fondaparinux during 1999–2014.
Results During the 16-year period, the use of antithrombotic drugs per 1000 inhabitants/day increased from 64 to 96 defined daily doses (DDD), and the prevalence proportion of users doubled from 5.1% to 9.6% of the Danish population. From 1999 to 2014, there was an increased use of both antiplatelet drugs (from 60 to 79 DDD per 1000 inhabitants/day) and VKA (from 4 to 9 DDD per 1000 inhabitants/day). NOAC was marketed in 2008 and had an abrupt rise in use to 8 DDD per 1000 inhabitants/day in 2014. The use of heparins and fondaparinux increased slightly during the study period (from 0 to 0.6 DDD per 1000 inhabitants/day). Hospital use of antithrombotic drugs also increased during the study period, but constituted a minor part of the total use (4 DDD per 1000 inhabitants/day in 2014).
Conclusions Considerable changes have occurred in the use of antithrombotic drugs during the past 16 years, including the introduction of several new and increasingly used treatment modalities such as NOAC. The trends in use of individual drugs correlate well with the publication of landmark studies.
Statistics from Altmetric.com
Introduction
Antithrombotic drugs are the cornerstone in the management of patients with venous thromboembolism, ischaemic heart disease, ischaemic stroke, peripheral arterial disease, prosthetic heart valves and atrial fibrillation/flutter.1 Trends in use of antithrombotic drugs are influenced by disease prevalence, adherence to prescriptions and pharmaceutical advertising, patents and costs. A number of new drugs such as non-vitamin K antagonist oral anticoagulants (NOAC) and newer antiplatelet drugs have been approved and heavily marketed for clinical use in recent years, which may have impacted the usage trends of antithrombotic drugs substantially. Importantly, several clinical trials on antithrombotic drugs have been conducted during the recent decades.2 The pattern of drug use is also affected by expert recommendations in international clinical guidelines from large organisations such as the European Society of Cardiology, American Heart Association and the American College of Cardiology.3 It is unknown to what extent changes in scientific evidence on antithrombotic drugs are being implemented in clinical practice.
Despite the importance of understanding the patterns in use of antithrombotic drugs, no previous studies have examined these trends in a nationwide setting. We examined 16-year trends in antithrombotic drug use in Denmark.
Methods
Setting
We conducted this study in Denmark from 1 January 1999 to 31 December 2014. Denmark has a population of approximately 5.6 million inhabitants, and all inhabitants have equal and free access to universal tax-supported healthcare, including services at general practitioners, hospitals and partial reimbursement for prescribed medications, including antithrombotic drugs. Antithrombotic drug sales in the primary healthcare sector comprise purchases of prescription drugs and pharmacy-only over-the-counter sale of aspirin. Of note, the vast majority of low-dose aspirin is obtained by prescription due to coprescribing with other drugs and cost reduction by reimbursement.4 Drugs prescribed by physicians at discharge from hospitals or at ambulatory visits are dispensed from community pharmacies and thus part of the primary sector sales.
Data source
We retrieved data on use of antithrombotic drugs according to the Anatomical Therapeutic Chemical (ATC) classification system from the publicly accessible MEDical STATistics (Medstat).5 Medstat has provided on-demand aggregate statistics on drug usage from the Danish primary sector since 1996, divided by age and sex from 1999 onwards and hospital sector since 1997.6 It is based on the Register of Medical Product Statistics.6 Although the pharmacy sale data in Medstat are only provided in aggregated format, it is based on individual-level prescription fillings. The registration includes drug sale in defined daily doses (DDD). The DDD is a WHO-defined statistical measure of drug consumption, representing the assumed average maintenance dose required by an adult when the drug is used for its main indication. Reporting data in DDD allows for comparison of trends in drug use, independent of varying prices and pack sizes. If the drug is later used for another indication and in another dose, the WHO can decide to change the DDD value. Importantly, Medstat updates each year all data (including data from previous years) according to the most recent WHO-defined DDD for each drug, which allows for comparison of drug use over time. The registry also includes the annual prevalence of drug users in the primary healthcare sector. As the actual denominator used in the calculations of the prevalence of users is not provided directly in Medstat, we obtained information on the size of the Danish population during 1999–2014 according to age groups and sex from Statistics Denmark.7
Statistical analysis
We compiled data on the primary care use of antithrombotic drug (ATC code: B01A) in DDD per 1000 inhabitants/day and the number of antithrombotic drug users per 1000 inhabitants. Specifically, we obtained data on, and illustrated graphically, use of antiplatelets (B01AC), including clopidogrel (B01AC04), low-dose aspirin (B01AC06), dipyridamole (B01AC07), prasugrel (B01AC22) and ticagrelor (B01AC24). In Medstat, the ATC code B01AC06 identifies low-dose aspirin specifically and therefore does not include higher doses (500 mg) used for pain relief.4 We also retrieved data on vitamin K antagonists (VKA) (B01AA), including warfarin (B01AA03) and phenprocoumon (B01AA04), as well as NOAC. NOAC included dabigatran (B01AE07), rivaroxaban (B01AF01) and apixaban (B01AF02). Finally, we assessed the use of heparins (B01AB), which included low-molecular weight heparins (LMWH): dalteparin (B01AB04), enoxaparin (B01AB05) and tinzaparin (B01AB10); unfractionated heparin (B01AB01) and fondaparinux (B01AX05). We stratified drug use by healthcare sector (hospital and primary healthcare), sex, age groups (20–39 years, 40–64 years, 64–80 years and above 80 years) and administrative regions (available since 2007). In addition, we stratified analyses by age groups, separate for men and women.
Results
During 1999–2014, the overall use of antithrombotic drugs per 1000 inhabitants increased from 64 to 96 DDD (table 1 and figure 1). In addition, the prevalence proportion of drug users increased from 5.1% to 9.6% of the entire Danish population (table 1). During 1999–2014, the number of Danish inhabitants remained relatively unchanged, and the age and sex distributions did not change substantially (see online supplementary table S1).
Antithrombotic drug use in the primary healthcare sector during 1999–2014 in Denmark
Timeline in use of antithrombotic drugs in Denmark and publication of landmark studies. AMPLIFY, oral apixaban for the treatment of acute venous thromboembolism study; ARISTOTLE, apixaban versus warfarin in patients with atrial fibrillation study; ARTEMIS, efficacy and safety of fondaparinux for the prevention of venous thromboembolism in older acute medical patients: randomised placebo-controlled trial; AVERROES, apixaban in patients with atrial fibrillation study; CLOT, low-molecular weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer study; CURE, effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation study; EINSTEIN, oral rivaroxaban for symptomatic venous thromboembolism study; MEDENOX, a comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients study; NOAC, non-vitamin K oral antagonists; OASIS 5, comparison of fondaparinux and enoxaparin in acute coronary syndromes study; PLATO, ticagrelor versus clopidogrel in patients with acute coronary syndromes study; POPADAD, the prevention of progression of arterial disease and diabetes trial; PREVENT, randomised, placebo-controlled trial of dalteparin for the prevention of venous thromboembolism in acutely ill medical patients study; PROFESS, aspirin and extended-release dipyridamole versus clopidogrel for recurrent stroke study; RCT, randomised controlled trial; RECOVER, dabigatran versus warfarin in the treatment of acute venous thromboembolism study; RE-LY, dabigatran versus warfarin in patients with atrial fibrillation study; ROCKET-AF, rivaroxaban versus warfarin in non-valvular atrial fibrillation study; TRITON-TIMI 38, prasugrel versus clopidogrel in patients with acute coronary syndromes study.
During the study period, the use of antiplatelet drugs increased from 60 to 79 DDD per 1000 inhabitants/day, and the use of VKA increased from 4 to 9 DDD per 1000 inhabitants/day. After its introduction in 2008, the use of NOAC rose abruptly to 8 DDD per 1000 inhabitants/day in 2014. The use of heparins also increased during the study period. Hospital use of antithrombotic drugs constituted a minor part of the total use (4 DDD per 1000 inhabitants/day in 2014), but the use increased during the 16-year study period (table 2). Although small regional differences were observed, the use of antithrombotic drugs in individual administrative regions of Denmark was consistent with the national trends (see online supplementary table 2).
Antithrombotic drug use at hospitals during 1999–2014 in Denmark*
Antiplatelet drugs
Low-dose aspirin was the most frequently used antiplatelet drug (figure 2). Its use increased over time before declining slightly after 2010. The use of clopidogrel increased steadily from 2002 to 2010, after which an additional increase in use was observed. The use of ticagrelor and prasugrel began in 2011 and reached a level of 1 DDD per 1000 inhabitants in 2014. Use of antiplatelet drugs overall showed an equal sex distribution in 1999, but over the study period, the total use increased relatively more for men than for women (figure 3), driven by similar trends in the use of aspirin.
Use of antiplatelet drugs (A), vitamin K antagonists (B), non-vitamin K oral antagonists (C) and heparins and fondaparinux (D) in defined daily doses per 1000 inhabitants/day in the primary healthcare sector, 1999–2014.
Age-stratified and sex-stratified primary care use of antiplatelet drugs and vitamin K antagonists, non-vitamin K antagonist oral anticoagulants (NOAC) and heparins in defined daily doses per 1000 inhabitants/day.
Oral anticoagulants
Throughout the study period, the use of warfarin increased consistently. In contrast, the use of phenprocoumon declined markedly (figure 2). The use of NOAC began with dabigatran, followed by rivaroxaban and apixaban (figure 2). In 2014, the use (per 1000 inhabitants/day) of dabigatran was 4 DDD compared with 3 DDD for rivaroxaban and 1 DDD for apixaban. More men than women used NOAC, and the use was most frequent in age categories above 65 years (figure 3).
Heparins
The use of heparins increased over the study period (figure 2); especially, the sale of dalteparin and tinzaparin increased. The sale of heparins was higher among women than men and highest in the age category above 65 years (figure 3).
For all the individual antithrombotic drugs, the sex-stratified analyses among men and women separately were also in accordance with the main findings (see online supplementary figure S1).
Discussion
This study is the first to provide nationwide data on long-term trends in antithrombotic drug use. Considerable changes have occurred in the use of antithrombotic drugs during the past 16 years. First, the volume of antithrombotic drug use has doubled, mainly due to increased use of antiplatelet drugs. Second, several new treatment modalities, such as NOAC, have been introduced, and these new drugs are increasingly gaining market share. Below, we discuss how the trends in use of individual drugs correlate with the publication of landmark studies.
Antiplatelet drugs
Aspirin: An explanation for the increased use of aspirin during the study period may be the increased prevalence of ischaemic heart disease during the study period8 and increased focus on primary prophylaxis supported by large randomised trials.9 ,10 The decline after 2010 may hold other explanations. Studies indicated that low-dose aspirin should be prescribed only to selected patients for primary prophylaxis due to an imbalance between protective and hazardous effects such as bleeding risk.11 Randomised studies on patients with diabetes receiving aspirin compared with placebo showed no benefit from aspirin in reducing the risk of cardiovascular events and mortality, whereas treatment with aspirin increased the risk of gastrointestinal bleeding.12 ,13 In addition, patients receiving aspirin, VKA and clopidogrel had a substantially increased risk of bleeding, compared with patients treated with only one or two antithrombotic drugs.9 It is therefore recommended that triple therapy should be used with caution and for the shortest possible time period.3 ,14 Aspirin has also been proven inferior to VKA in the thromboprophylaxis of patients with atrial fibrillation.15 Although aspirin has previously been recommended for reducing the risk of ischaemic stroke, it is now considered obsolete for patients with atrial fibrillation only.3
Clopidogrel: The increase in use of clopidogrel seen in our study correlates with the publication of the CURE study from 2001.16 In this study, patients with non-ST-segment elevation myocardial infarction or unstable angina pectoris were randomised to receive loading and maintenance doses with clopidogrel or placebo in addition to aspirin.16 Patients treated with clopidogrel had a reduction in 1-year mortality and risk of reinfarction and heart failure, while the risk of bleeding was increased compared with placebo.16 The additional rise in sale of clopidogrel seen in 2010 might be explained by the PROFESS study published in 2008.17 In this study, patients with stroke were randomised to receive secondary prevention with clopidogrel alone or aspirin in combination with dipyridamole.17 The study demonstrated similar risk of recurrent stroke and cardiovascular death in the two groups, but an increased risk of major haemorrhagic events was seen for patients treated with aspirin and dipyridamole compared with clopidogrel.17 Accordingly, this may explain the decline in use of dipyridamole after 2010.17
Ticagrelor and prasugrel: As an alternative to treatment with clopidogrel in patients with acute coronary syndrome (ACS), newer antiplatelet drugs such as ticagrelor and prasugrel have recently been implemented in clinical practice.18 The PLATO study in 2009 compared ticagrelor with clopidogrel and found a lower 1-year mortality in patients treated with ticagrelor.19 In the TRITON-TIMI 38 study, patients with ACS scheduled for percutaneous coronary intervention were randomised to either clopidogrel or prasugrel.20 The study demonstrated a reduction in 1-year cardiovascular mortality, risk of reinfarction and stent thrombosis in the prasugrel group at the expense of an increased risk of major and life-threatening bleedings.20 Although these studies were published more than 5 years ago, only a small rise in overall sale of these drugs was seen in Denmark. This may be attributed in part to a consistently higher price for ticagrelor than for clopidogrel during the study period21 and the convenience for patients to use clopidogrel, which is administered once a day compared with two times per day for ticagrelor.
Oral anticoagulants
VKA: The rise in use of VKA may have several explanations. The prevalence of patients diagnosed with atrial fibrillation (who receive anticoagulant treatment) has increased over the study period.22 Although the thromboembolic risk of atrial fibrillation has been known for several years, a meta-analysis in 2007 found a 67% reduction in the risk of ischaemic stroke among patients treated with VKA compared with patients without anticoagulant treatment.23 Due to new treatment opportunities and several clinical guidelines in this area,2 the importance of anticoagulation has also received increased attention. Recently, prediction tools such as the CHA2DS2VASc score have been recommended,3 which have likely lowered the threshold for prescribing anticoagulants. The use of warfarin has increased, while the use of phenprocoumon has declined, which may be attributed to stronger scientific evidence for treatment with warfarin than for phenprocoumon and, most importantly, the fact that warfarin has a shorter half-life than phenprocoumon, making warfarin more convenient and easier for clinicians to use.
NOAC: The use of NOAC has been rapidly adopted after 2008, owing to at least non-inferiority compared with warfarin in the prevention of thromboembolism among patients with atrial fibrillation and a lower risk of bleeding.24–26 Although the prices still are considerably higher for NOAC than for VKA,21 NOAC are increasingly gaining market share. In accordance with the inclusion criteria of the RE-LY,24 ROCKET-AF25 and ARISTOTLE26 studies, the use of these drugs occurred frequently in ages between 65 and 79 years, but several patients outside this age group were also prescribed the medication. NOAC have also been approved for treatment of venous thromboembolism which also has contributed to the rise in sale, particularly for rivaroxaban.27
Heparins
LMWH: The increased use of dalteparin and tinzaparin since 2003 may be attributed to an increased use of medical thromboprophylaxis. Several studies including MEDENOX, PREVENT and ARTEMIS for acutely ill medical patients found that 14 days treatment with LMWH or fondaparinux compared with placebo reduced the risk of deep vein thrombosis within 3 months by approximately 50%.28 Furthermore, the use of LMWH has increased among patients with cancer after the CLOT trial in 2003 compared with warfarin in patients with venous thromboembolism, and it has been found that among patients with cancer the risk of recurrent venous thromboembolism was reduced among patients treated with LMWH.29 In contrast to other LMWHs, the consumption of enoxaparin declined after 2006. This may partly be explained by the release of the OASIS 5 study comparing enoxaparin with fondaparinux in patients with ACS.30 The study showed similar effects on mortality and risk of reinfarction, but fondaparinux halved the risk of bleeding compared with enoxaparin.30 Since enoxaparin was primarily used for ACS in Denmark, the consumption fell, and the use of fondaparinux increased correspondingly.
Strengths and limitations
Several issues should be considered when interpreting our results. The nationwide coverage eliminated selection bias. In Denmark, it is mandatory to report drug sales to the Register of Medical Product Statistics; hence, Medstat is complete. The data were prospectively recorded and have previously proven a valid source for long-term trends in drug use.4 Data were based on redeemed prescriptions as a proxy for use and not just written prescriptions. Still, prescription redemption may not be identical to actual drug use, but such potential discrepancies cannot explain the overall trends observed in drug use. As data were not linkable to other registries, it falls for future studies to examine to what extent the changes in drug use are influenced by temporal changes in patients' comorbidity burden. Finally, it should be noted that landmark studies may explain some, but not all of the observed patterns of use. Our results are likely generalisable to most Western countries, where clinical guidelines have followed the published evidence on antithrombotic drugs.
Conclusion
Several changes have occurred in the use of antithrombotic drugs over the last 16 years, including the introduction of new and increasingly used drugs such as NOAC. The overall increase in use of antithrombotic drugs was driven by an increase in the use of antiplatelet drugs. The trends in use of individual drugs correlate with landmark publications.
Key messages
What is already known on this subject?
Antithrombotic drugs are prescribed for cardiovascular prevention and for patients with venous thromboembolism, ischaemic heart disease, ischaemic stroke, peripheral arterial disease, prosthetic heart valves and atrial fibrillation/flutter.
Studies on long-term nationwide trends in their use in real-world practice are needed.
What might this study add?
During the 16-year period, the use of antithrombotic drugs per 1000 inhabitants/day increased from 64 to 96 defined daily doses, and the prevalence proportion of users doubled from 5.1% to 9.6% of the Danish population.
Several new treatment modalities, such as non-vitamin K oral antagonists, have been introduced, and are increasingly gaining market share.
The study shows how these trends in use of antithrombotic drugs correlate well with the publication of landmark studies throughout the period.
Supplemental material
How might this impact on clinical practice?
Information about clinical practice patterns is necessary to improve treatment quality by evaluating whether guidelines, which translate the results of clinical trials, are being implemented in clinical practice.
References
Footnotes
Contributors KA and MS conceived the idea and designed the study. KA ascertained the statistical results. KA, ELG, JS, ML and MS interpreted the data and reviewed the literature. KA drafted the first manuscript. ELG, JS, ML and MS critically reviewed the manuscript and approved the final version for submission. MS has the overall responsibility for the accuracy of the data and the manuscript.
Funding Funding was provided through the Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark.
Competing interests ELG has received speaker honoraria from AstraZeneca, Baxter, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb and Pfizer and has participated in advisory board meetings for AstraZeneca, Bayer, Boehringer Ingelheim and Bristol-Myers Squibb.
Ethics approval Approval from an ethical committee was not required for this study according to Danish law.
Provenance and peer review Not commissioned; externally peer reviewed.