Cost effectiveness of dabigatran etexila
Journal of Medical Economics
ISSN: 1369-6998 (Print) 1941-837X (Online) Journal homepage: http://www.tandfonline.com/loi/ijme20
Cost-effectiveness of dabigatran etexilate for
stroke prevention in non-valvular atrial fibrillation.
Applying RE-LY to clinical practice in Denmark
Lars K Langkilde, Mikael Bergholdt Asmussen & Mikkel Overgaard
To cite this article: Lars K Langkilde, Mikael Bergholdt Asmussen & Mikkel Overgaard
(2012) Cost-effectiveness of dabigatran etexilate for stroke prevention in non-valvular atrial
fibrillation. Applying RE-LY to clinical practice in Denmark, Journal of Medical Economics, 15:4,
695-703
To link to this article: http://dx.doi.org/10.3111/13696998.2012.673525
Accepted author version posted online: 07
Mar 2012.
Published online: 22 Mar 2012.
Submit your article to this journal
Article views: 389
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ijme20
Download by: [Boehringer Ingelheim]
Date: 10 February 2016, At: 01:22
Journal of Medical Economics
1369-6998
doi:10.3111/13696998.2012.673525
Vol. 15, No. 4, 2012, 695–703
Article 0001.R1/673525
All rights reserved: reproduction in whole or part not permitted
Lars K Langkilde
Abstract
se oa
d,
C
No op
Un t
yr
au fo
ig
th
r
ht
di or S
sp ize a
la d le ©
y, us
20
vi e o
ew p r
1
r
2I
C
o
an h
o
nf
d p ibi m
t
rin ed m
or
t a . Au e
m
si th rc
aU
ng or i
le is al
K
co ed
D
py us is Lim
fo ers tr
ite
rp c i
b
a
er n
d
so d ut
na ow io
l u nl n
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Original article
Cost-effectiveness of dabigatran etexilate
for stroke prevention in non-valvular atrial
fibrillation. Applying RE-LY to clinical
practice in Denmark
Wickstrøm & Langkilde ApS, Vejle, Denmark
Mikael Bergholdt Asmussen
Mikkel Overgaard
Boehringer Ingelheim Denmark, Copenhagen,
Denmark
Address for correspondence:
Lars K Langkilde, Wickstrøm & Langkilde ApS,
Hjulmagervej 4A Vejle, DK-7100, Denmark.
Tel.: þ4541111183; Fax: +45 3514 1110;
ll@wlpharma.com
Key words:
Anti-coagulation – Dabigatran etexilate – Warfarin –
Atrial fibrillation – Cost-effectiveness – Denmark
Accepted: 5 March 2012; published online: 22 March 2012
Citation: J Med Econ 2012; 15:695–703
Objective:
To estimate the economic implications of introducing dabigatran etexilate (‘dabigatran’) for anti-coagulation
therapy in Danish patients with non-valvular atrial fibrillation based on results of the RE-LY trial.
Methods:
The lifetime cost and outcomes of dabigatran and warfarin were estimated using a previously published
cost-effectiveness model. The model utilizes the data from the RE-LY study to estimate the costs and
outcomes of stroke prevention in atrial fibrillation. Cost estimates were based on official Danish tariffs and
prices, and published literature on the cost of stroke. In the base-case analysis a conservative approach was
adopted applying tariffs from the lowest range for the cost of International Normalized Ratio (INR) monitoring
associated with warfarin. The effectiveness measure of the analysis was quality-adjusted life-years (QALY).
Results:
The model estimated that the mean cost per patient for the remaining life-time is E16,886 treated with
warfarin and E18,752 treated with dabigatran. This was associated with mean QALYs per patient of 8.32
with warfarin and 8.59 with dabigatran. The resulting incremental cost-effectiveness ratio (ICER) of
E7000 per QALY gained is regarded as cost-effective by Danish standards. This conclusion was seen
to be robust to realistic variations in input parameters, including adjustment for the RE-LY centres achieving
the best INR monitoring quality. Threshold analysis revealed that dabigatran would be cost-saving in settings
where the cost of warfarin monitoring exceeds E744 per year.
Limitations:
The analysis does not include all aspects of Danish clinical practice anti-coagulation that will influence costeffectiveness of dabigatran, e.g., this study did not attempt to model quality of anticoagulation monitoring
and under-utilization in clinical practice.
Conclusions:
Based on the outcomes observed in the RE-LY trial, dabigatran represents a cost-effective alternative to
warfarin in Denmark for all patients with atrial fibrillation within the licensed indication of dabigatran.
Introduction
Atrial fibrillation (AF) is the most common cardiac arrhythmia, which increases
morbidity and mortality principally due to the resultant 5-fold increased risk of
stroke compared to patients without AF. Persons above 40 years of age have a
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
695
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
25% lifetime risk of developing AF1. Based on international literature1–3, Brandes4 estimated that more than
50,000 Danish patients suffer from AF, with prevalence
expected to treble by the year 2050.
Anticoagulation is recommended for AF patients with
risk factors for thromboembolic events5. The use of anticoagulation with vitamin K antagonists has evolved over
time. Two decades ago it was generally limited to AF
patients with rheumatic heart disease and prosthetic
heart valves6. With growing evidence from randomized
controlled trials an increasing number of AF patients are
recommended for anticoagulation treatment5.
The historical standard for anticoagulation treatment
in Denmark has been the vitamin K antagonist warfarin.
Warfarin reduces the risk of stroke by 60% compared to
placebo and 30–40% compared to treatment with aspirin7.
Warfarin can be cumbersome to use because of the multiple drug–drug and drug–food interactions, and the significant intra- and inter-patient variability in terms of
therapeutic response. Achieving a therapeutic dose in
clinical practice requires regular monitoring of the
International Normalized Ratio (INR) and dose adjustment. The quality of INR monitoring has been shown to
vary between countries8 and is related to how anticoagulation management is organized9, with implications for reallife effectiveness10. However, in Denmark, valid data on
warfarin management performance is not available. The
majority of Danish warfarin patients are managed by their
general practitioner.
Friberg et al.11 reported that only 36% of patients hospitalized with a diagnosis of AF in Denmark claimed a
prescription of vitamin-K antagonists (i.e., warfarin)
within 3 months of discharge. Furthermore, a recent
Danish study of secondary prophylaxis showed that
among patients older than 80 years only 25% of patients
with atrial fibrillation and no contraindications for anticoagulant therapy at the time of the initial hospital admission were treated with oral anticoagulation 1 year after the
initial ischemic stroke12.
Dabigatran etexilate (‘dabigatran’) is a new oral direct
thrombin inhibitor which provides a predictable anticoagulant effect with no need for anticoagulation monitoring.
In 2008, dabigatran was approved in Europe for primary
prevention of venous thromboembolic events in adult
patients who have undergone elective total hip or knee
replacement surgery. In August 2011 dabigatran was
approved in Denmark for stroke prevention for patients
with non-valvular AF with at least one risk factor.
In the Randomized Evaluation of Long-term
Anticoagulation Therapy (RE-LY) trial (n ¼ 18,113), the
efficacy and safety of dabigatran was compared with warfarin in patients with AF with risk factors for stroke13,14.
Two doses of dabigatran were tested and dabigatran
150 mg twice daily was associated with a significant
lower rate of stroke/systemic embolism and a comparable
696
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
rate of major haemorrhage compared to warfarin, whereas
dabigatran 110 mg twice daily was associated with a comparable rate of stroke and systemic embolism and a significant lower rate of major haemorrhage compared to
warfarin. Both doses significantly reduced the rate of intracranial and life-threatening bleeding compared to
warfarin14.
With an increasing prevalence of AF and a lower
threshold for initiating anticoagulation therapy it is relevant to consider which treatment option represents the
best value-for-money. The purpose of this analysis is to
estimate the long-term cost-effectiveness of stroke prevention in patients with AF using dabigatran as an alternative
to warfarin in a Danish setting.
Methods
Model of clinical events
We adapted an economic model of long-term (remaining
life-time) cost and outcomes based on the RE-LY trial
published by Sorensen et al.15,16.The model has Markov
properties, i.e., it is assumed that any patient’s health
can be represented as a finite number of health states.
Transitions between states can occur once every 3
months with transition probabilities conditional on the
present health state, baseline characteristics of the patient,
and progression of time. The health states represent ischemic stroke history, degree of disability following stroke
(independent, moderately disabled, or dependent), current
treatment (first line, primary prophylaxis using warfarin/
dabigatran treatment, second line prophylaxis using aspirin, or no treatment), and death. The model estimates the
distribution of patients on health states from entry into
the model and until all patients are dead, i.e., the time
perspective is remaining life-time. Calculations were performed in Microsoft Excel 2007.
Clinical events modelled include ischemic stroke,
transient ischemic attack, systemic embolism, acute myocardial infarction, intracranial haemorrhage and haemorrhagic stroke, other major haemorrhages, and minor
bleeds.
The publication by Sorensen et al.16 analyses the longterm cost and outcomes from a Canadian perspective and
was adapted further to the Danish healthcare setting. The
Canadian cost-effectiveness model is suitable for this adaptation because its main analysis was based on the RE-LY
data and applies the sequential dosing strategy of dabigatran that is common between the approved European and
Canadian labels. The posology reflected in the costeffectiveness model initiates patients under the age of 80
years at dabigatran 150 mg twice daily followed by 110 mg
twice daily at the age of 80 years and over.
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Journal of Medical Economics
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Risk of events
In our adaptation we used the baseline analysis of the
Canadian adaptation which is an extrapolation of outcomes in the RE-LY trial to long-term (lifetime) outcomes
and cost. Below we will only provide a short summary of
the modelling approach. For full details please consult
Sorensen et al.16.
Risk of clinical events for patients on warfarin and
dabigatran were based on observed event rates in the
RE-LY trial. Separate risk estimates for all modelled clinical events were estimated for the sub-group of patients
respectively younger than or at least 80 years of age at
baseline16.
Risk of ischaemic stroke was estimated separately for
each sub-group of patients based on the CHADS2—
scoring of risk factors17 at baseline. In the model the
CHADS2 is reassessed every cycle to reflect changes in
the risk factors as a result of predicted clinical events
during the previous 3-month cycle.
Treatment and treatment effects
Patients entering the economic model are assumed to
receive either warfarin (target INR 2–3) or dabigatran.
Patients are assumed to discontinue from any of the treatments in the case of an intracranial haemorrhage (ICH).
Furthermore, we assume that 50% of other major haemorrhages lead to permanent discontinuation of antithrombotic treatment. These patients are assumed to be
at risk of clinical events at rates that reflect those of placebo-treated patients in published clinical trials, and
an increased mortality risk depending on the level of
disability16. Furthermore, treatment switches (from warfarin or dabigatran to aspirin) can occur in the model for
other reasons. In this case patients are assumed to experience risks of clinical events and costs as if treated with
aspirin. Rates of discontinuation for warfarin and dabigatran were estimated using data from RE-LY.
For warfarin and dabigatran the events rates were estimated from RE-LY. For aspirin (modelled as a second line
treatment option) and no treatment (used for modelling
the clinical events for patients who have discontinued all
anti-thrombotic treatment following a major haemorrhagic event) event rates are based on an adaptation of a
recently published meta-analysis of controlled clinical
trials involving warfarin, aspirin, and/or placebo18.
Estimation of health outcomes
Long-term health outcomes were estimated as number of
clinical events, life expectancy (life-years), and qualityadjusted life-years (QALYs).
QALYs were estimated by weighting 3-month survival
by a utility weight for the health state during that cycle. If a
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Volume 15, Number 4
August 2012
clinical event occurred during a cycle, a temporary utility
decrement was subtracted from the health state utility.
Utility weights relevant to AF in a Danish population
are not available; hence we used the same values as applied
in Sorensen et al.16.
Adaptation of clinical event model
Danish age- and gender-specific death rates from national
population statistics19 were applied. In order not to
double-count causes of deaths already accounted for in
the model as event-related mortality, only death from
other causes than stroke, bleeding, systemic emboli, and
acute myocardial infarction was included in the calculation of death rates.
Furthermore, since the efficacy and safety of warfarin
are related to the achieved quality of INR monitoring, a
sensitivity analysis was planned allowing the level of timein-therapeutic range (TTR) to deviate from that of the
warfarin treated patients in the RE-LY trial. The costs
and outcomes of dabigatran and warfarin were compared
using results from a post-hoc analysis of RE-LY based on
centre average TTR20. Relative risk ratios were reported by
quartile of patients according to the mean TTR achieved
by the trial centre that the patient attended during RE-LY.
The patients from the lowest quartile represent models of
care achieving a low level of TTR (centres with an average
TTR of less than 57.1%) and the highest quartile represent
models of care with a high level of TTR for INR monitoring (centres with average TTR at least 72.6% or above).
The relative risk ratios reported20 for ischaemic stroke and
systemic embolism, intra-cranial bleeding, and major
bleedings for these two extreme quartiles were applied as
sensitivity analyses.
Cost of medical products
Costs of medicinal products are based on the current pharmacist sales prices including dispensing fees but excluding
VAT21. For warfarin a daily cost of treatment based on a
5 mg daily dosage is assumed22.
Estimation of cost
The direct costs associated with anti-thrombotic treatment and clinical events were estimated using primarily
national tariffs where available23,24 and data from other
publicly-available sources otherwise. Transferrals such as
pensions, work disability compensation, and value
added tax (VAT) were excluded. Table 1 provides an overview of the included cost components in the model. In the
following all cost are reported in euro (E) based on the
2011 annual average currency rate (100 Danish
Kroner ¼ 13.42 E).
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
697
Journal of Medical Economics
Volume 15, Number 4
August 2012
Table 1. List of unit cost estimates applied in the model. In Euro (E).
Item
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Ischaemic stroke
3 month cost
Initial hospitalization
Other
Total
Lifelong cost
Heamorrhagic stroke
3 month cost
Initial hospitalization
Other
Total
Lifelong cost (dependent only)
Systemic embolism
Transient Ischemic Attack
Acute Myocardial Infarction
Major extracranial haemorrhage
Fatal
Baseline
18,771
1701
20,472
1287
16,665
1359
18,024
1028
3228
2253
8629
3223
Non-fatal/non-GI
1765
Non-fatal/GI
2623
Treatment switch/discontinuation
INR monitoring (per year)
17
513
Dabigatran initiation
176
Dabigatran maintenance
17
Dabigatran
Warfarin
2.63
0.26
Aspirin
0.09
Notes
(per 3 months)
(per 3 months)
DRG 114 (kidney
embolism)
DRG 125 (TIA)
See text
DRG 638 (GI
bleed age
417 with
complication)
DRG 323 (Blood
loss)
DRG 640
(Uncomplicated GI bleed
age 417)
GP visit
Per year. See
Table 2
First 3 months.
See Table 2
Per year. See
Table 2
Per day
Per day (assume
2 2.5 mg
dose of
warfarin)
Per day (assume
75 mg)
Cost of stroke
Cost of stroke is a key parameter in the costing of clinical
events. Tariffs are available for the initial hospitalization
in the Danish Disease Related Groups (DRG) classification of hospitalizations. However, it was not possible to
identify up-to-date Danish studies of resource utilization
associated with stroke (re-hospitalization, rehabilitation,
and social costs) following the initial hospitalization.
Porsdal and Boysen25 performed a comprehensive estimation of first year resource utilization following ischaemic and haemorrhagic stroke based on cost applicable to
1995. The hospital cost during the initial hospitalization
constituted 63.4% of the total first year direct medical and
social cost in patients admitted with ischaemic stroke and
59.4% of first year cost in haemorrhagic stroke. The initial
hospitalization cost for ischaemic stroke and ICH in the
model was based on Porsdal and Boysen25.
698
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
The first year follow-up cost from Porsdal and Boysen25
could not be used directly because it included cost related
to secondary strokes occurring within the first year. Since
secondary stroke is estimated and assigned a cost in our
clinical model this would imply double counting cost of
secondary strokes. From a Swedish study on cost of stroke
with 4 years follow-up26 it was possible to estimate that
11% of first year follow-up cost was related to secondary
stroke. The Danish estimate of first year follow-up cost was
reduced by this amount.
Furthermore, the long-term cost of stroke was estimated. The Swedish study reported that social cost and
follow-up cost are highest in the first year following the
initial stroke event and fairly stable the second through
fourth year. The annual cost years 2–4 was 24% lower
than first year follow-up cost. This relative reduction in
cost (based on the Swedish observation) was applied to our
estimate of the Danish first year follow-up cost. In order to
use the estimated long-term follow-up cost conservatively
we only applied it to patients with moderate or severe
disability after the initial stroke event. Estimates of
stroke cost were inflated to 2011 values using the consumer
price index27.
Cost of patient management
INR monitoring is organized in a number of different ways
in Denmark28. For the base-case analysis, the cost of INR
monitoring was estimated based on the most common and
least costly set-up where INR monitoring is managed by
general practitioners. Based on publicly available tariffs
and assumed annual number of contacts, the base-case
cost for INR monitoring was estimated to E513. In sensitivity analyses, the impact of different scenarios for INR
monitoring costs was analysed. In one scenario—reflecting
the split between primary care and hospital organized INR
monitoring in Denmark—the cost of INR monitoring was
estimated as a weighted average of cost for the three most
frequent models of care in Denmark. For each model of
care the annual number of contacts to the provider and the
proportion of the patient population being monitored were
assessed. Cost per contact was estimated using publicly
available tariffs. The resulting annual cost per patient is
E975 per year (weighted average based on estimated
patient distribution by models of care, see Table 2). The
publicly-available tariffs especially for hospital contacts
have been shown to deviate from variable cost of monitoring, i.e., cost estimated from the pure work input and materials used during one contact28. In another scenario only
variable cost per contact were taken into consideration.
Variable costs are—as also stated by Holm et al.28—only a
relevant measure in short-term analyses, since they do not
take into account the overhead cost associated with running the clinic or the alternative use of clinic space for
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Journal of Medical Economics
Volume 15, Number 4
August 2012
Table 2. Calculation of annual patient management cost on dabigatran and warfarin (Euro).
Frequency
Warfarin
GP (60%)
3rd party lab (50%)
Point of care instrument (50%)
Hospital (40%)
Paper-based (27%)
AC clinic (70%)
Patient self-monitoring (3%)
Weighted average (warfarin)
Dabigatran
1 year
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
2þ years
Contacts
per year
4 weeks
3.5 weeks
5 weeks
5 weeks
2 weeks
Specialist
GP
Total
GP
Description
Annual
cost (E)
Visit, test and shipment, follow-upa
Visit and follow-upa
Outpatient visitb
Outpatient visitb
Consultation by phone/internetb
513
516
1654
1654
2068
975
1
3
Outpatient visitb
Visita
1
Visita
159
52
211
17
AC, Anti-coagulation; GP, general practitioner.
a
Unit cost (E): visit 17.38, blood test and shipment 8.94, INR (PoC) 14.63, phone consultation 3.41, mail follow-up 5.47. Assumes 50% of follow-up is done faceto-face and 50% via phone/mail.
b
Unit cost (E): Outpatient visit 159.05, phone contact 79.52.
other purposes. Variable cost was estimated based on experience from a hospital-based anticoagulation clinic using
computer-assisted dosing (calculations not shown). The
resulting variable cost of monitoring was estimated to
E418 per year.
There is no need for anticoagulation monitoring or dose
adjustment with dabigatran. However, a temporary higher
cost for initiation of dabigatran and a permanent cost of
dabigatran management were included. It was assumed
that all patients will be seen by a cardiologist (outpatient
visit) at initiation of dabigatran treatment and by a GP
three times during the first year and once yearly thereafter
specifically related to dabigatran. In reality, dabigatranrelated issues are likely to be part of the general patient
management and dealt with in the course of other AF
related visits. However, in the model these are considered
additional to any other visits the patients may have as a
consequence of their health state. Renal function testing
will also occur at those visits, according to the Danish
label.
Cost of bleeding events
Cost of ICH was based on published cost of stroke estimates25 as outlined above. Major bleeding events other
than intracranial haemorrhage were estimated using
DRG tariffs. Costs of fatal and non-fatal gastrointestinal
(GI) bleedings were estimated using the tariff for GI bleedings, respectively, with and without complications, while
costs for non-GI major bleedings were estimated using the
tariff for inpatient treatment of blood loss. Minor bleedings
were not applied any cost in the base-case analysis. In
sensitivity analyses a cost range of 20% of the baseline
estimates (both intracranial and other major bleedings)
! 2012 Informa UK Ltd www.informahealthcare.com/jme
were tested. We applied the higher bleeding cost to both
treatments and—in a separate sensitivity analysis—only to
dabigatran-treated patients and for this analysis we also
applied the cost of a GP visit for minor bleedings in dabigatran-treated patients. The asymmetric analysis of bleeding cost was applied to test whether an increase in resource
utilization associated with the new medicine would be critical for the conclusion.
Cost of acute myocardial infarction (AMI)
Cost of AMI is based on observed use of revascularization and imaging procedures in 24,952 Danish patients
with incident ACS29. Applying 2011 tariffs to the casemix reported leads to an estimated cost per event of
E8629.
Discounting
All future costs and outcomes (QALYs) were discounted
in accordance with standards set by the Danish Ministry of
Health30 at an annual rate of 2%, with 0% and 4% discounting tested in sensitivity analyses.
Patient cohort
We estimated outcomes and cost for a modelled cohort of
10,000 patients with a similar risk profile (described by
CHADS2) to the patients aged 80 or below included in
the RE-LY trial. The average age was 69 years and 65% of
patients were male.
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
699
Journal of Medical Economics
Volume 15, Number 4
August 2012
Results
Cost and outcomes
Table 3 reports the estimation of long-term cost and outcomes for the baseline analysis. The estimated total cost
per patient for the remaining life-time is E16,886 for
patients treated with warfarin and E18,752 for patients
Table 3. Baseline cost and quality-adjusted life-years per patient (life-long
projection) by treatment strategy. Incremental cost-effectiveness ratio
(warfarin compared to dabigatran).
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Sequentiala
Dabigatran use
Direct costs/Patient
Drug costs and Monitoring (E)
8232
Event costs (E)
8787
Follow-up costs (E)
1733
Total (E)
18,752
Incremental costs (E)
1866
Benefits/Patient (discounted at 2% per annum)
b
0.47
Expected number of events
Expected survival (years)
11.61
QALYs
8.59
Incremental QALY
0.27
ICER (Dabigatran vs warfarin)
Direct Costs (E) per QALY gained
6950
treated with dabigatran. The additional cost associated
with dabigatran (E1866) is mainly driven by the additional cost of medication. The estimated QALYs per
patient are 8.32 for warfarin and 8.59 for dabigatran. The
expected survival (discounted by 2% per annum) is 11.34
and 11.61 years, for warfarin and dabigatran, respectively.
The resultant incremental cost per QALY gained is
E6950. There is no officially accepted threshold for incremental cost per QALY in Denmark. However, an ICER
below 30,000 E/QALY is regarded as cost-effective, and,
hence, dabigatran is considerably below this threshold
compared to warfarin in the baseline analysis.
Warfarin
Sensitivity analysis
4935
9727
2224
16,886
0.55
11.34
8.32
QALY, Quality adjusted life years; ICER, incremental cost effectiveness ratio.
150 mg b.i.d. for age below 80 then 110 mg b.i.d.
Stroke, systemic embolism, and intracranial haemorrhage.
a
b
Table 4 reports the main findings from the sensitivity analysis. Two different scenarios for INR monitoring cost and/
or a high quality of INR monitoring were analysed.
If INR monitoring costs were based on a weighted average of the three most frequent models of care in Denmark,
the assigned cost of INR monitoring was E975. In this reallife scenario, dabigatran was shown to be dominant (i.e.,
lower cost and better outcome compared to warfarin). If
INR monitoring is assigned only the variable cost, the
incremental cost is E2633 per patient (compared to
E1866 in the base-case). This leads to an incremental
cost per QALY gained of E9806. Given these results dabigatran would be regarded as cost-effective compared to
Table 4. Univariate sensitivity analyses. Net present value of incremental cost and outcomes (dabigatran minus warfarin).
Incremental costs (E)
Incremental QALY
Baseline
1866
0.268
INR monitoring
Variable cost estimate
2633
0.268
Tariff based estimate
1704
0.268
Centres achieving TTR557.1%
1760
0.288
Centres achieving TTR 57.1–65.5%
1601
0.309
Centres achieving TTR 65.5–72.6%
2434
0.188
Centres achieving TTR472.6%
2952
0.102
Cost of stroke
20% lower
2188
0.268
20% higher
1544
0.268
Cost of bleeding
20% lower
1854
0.268
20% higher
1899
0.268
20% higher (dabigatran only)
2016
0.268
Discounting of future cost and outcomes (baseline 2% p.a.)
No discounting of future cost and outcomes
2114
0.263
Discounting at 4% p.a.
2183
0.272
Threshold analysis (annual cost of monitoring associated with equal lifetime cost of dabigatran and warfarin)
ICER E/QALY
6950
9806
–
6115
5175
12,923
29,019
8149
5752
6905
7075
7510
8037
8023
Threshold monitoring cost (E/year)
Baseline (all centres)
Centres achieving TTR557.1%
Centres achieving TTR 57.1–65.5%
Centres achieving TTR 65.5–72.6%
Centres achieving TTR472.6%
744
731
712
815
879
p.a., per annum; QALY, Quality-adjusted life years; TTR, Time in therapeutic interval.
*Incremental cost per QALY gained (dabigatran compared to warfarin).
700
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
Figure 1. Incremental cost and quality-adjusted life-years at varying level of thrombo-embologic and intra-cranial haemorrhage risk.
Note: Each cluster represents a level of ICH risk from 0.4 to 1.6 events per 100 patient years. Each point represents a level of CHADS2 from 0 (0.62
events/100 patient years) to 5–6 (2.77 event/100 patient years).
warfarin by any reasonable method of applying the cost of
INR monitoring under base-case assumptions.
Furthermore, it was shown in the sensitivity analysis that
if the annual cost of INR-monitoring exceeds E744, dabigatran would dominate warfarin.
Taken into consideration the achieved quality of INR
monitoring based on the centre average TTR, the incremental cost per QALY range from E6115 for the lowest
level of quality (centres with an average TTR of less than
57.1%) to E29,019 for the highest level of quality (centres
with average TTR at least 72.6% or above). Dabigatran
would be regarded as cost-effective to warfarin, irrespective
of average level of centre TTR.
The results were not sensitive to 20% changes in cost of
stroke or bleeding. In the case of bleeding cost we applied a
20% higher cost only to the dabigatran arm. This resulted
in an incremental cost per QALY of E7510, which is still
cost-effective compared to warfarin. The conclusion was
also found to be insensitive to the rate of discounting of
future cost and outcomes.
Figure 1 provides the results from sensitivity analysis
with respect to various combinations of risks of intracranial haemorrhage and stroke (based on CHADS2 score).
Each cluster represents a level of ICH risk (from 0.4 events
per 100 patient years to 1.6 events per 100 patient years)
and each point represents a CHADS2 score from 0–5/6. In
RE-LY the risk of ischaemic stroke on warfarin treatment
varied between 0.62 events per 100 patient years for
patients with a CHADS2 score of 0 at baseline to 2.77
events per 100 patient years for patients with a score of
5–6 at baseline16. The risk of ICH has a larger impact on
! 2012 Informa UK Ltd www.informahealthcare.com/jme
both incremental cost and incremental outcomes than the
risk of thromboembolic events. In general, higher levels of
risk are associated with lower incremental cost of dabigatran treatment and larger patient benefits (measured in
QALY gained) compared to warfarin. In all of the analysis
in Figure 1 the ICER is below E18,000 per QALY gained
and hence dabigatran remains cost-effective compared to
warfarin.
Discussion
This analysis evaluated the cost-effectiveness of dabigatran vs warfarin for patients with atrial fibrillation in a
Danish healthcare setting based on the results from the
RE-LY trial. In the base-case analysis dabigatran was
shown to be a cost-effective strategy for patients with
atrial fibrillation and the sensitivity analyses confirmed
the robustness of the base-case finding. The applied costeffectiveness model is—as with any extrapolation—
limited by the amount of evidence available. Hence, the
following limitations of this analysis should be mentioned.
The driving factor behind the costs and outcomes are
the risk levels for clinical events in the model. These are
primarily based on observed risk levels in the RE-LY trial
and the generalizability of the trial to the Danish setting in
terms of selection of patients, healthcare providers, and
patient management should be examined. This costeffectiveness does not address this real-life scenario with
patient exclusion however; Sorensen et al.16 showed that
the cost-effectiveness ratio will become even more
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
701
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
favourable for dabigatran in case a real-life scenario is
applied in addition to the applied RE-LY scenario.
Furthermore, we compare a new treatment option with
an established one. A number of assumptions were made to
estimate the annual dabigatran treatment cost, e.g., the
number of healthcare contacts, management of dabigatran-related bleedings, and treatment adherence.
Prospective data from a real-life setting will be needed to
provide more accurate estimates of dabigatran management and related cost.
Another aspect which is not reflected in this cost-effectiveness analysis is the substantial lack of treatment for
AF-patients who according to risk profile are recommended to receive anticoagulation therapy. Friberg
et al.11 documented—based on a Danish database study
(n ¼ 68,546)—that only 36% of the patients hospitalized
with a diagnosis of AF claimed a prescription of vitamin-K
antagonists (i.e., warfarin) within 3 months from discharge. The reasons for this lack of treatment may be
related to the fact that treatment with warfarin is considered as cumbersome due to the interactions with other
drugs and food and the frequent need for INR monitoring
to ensure an acceptable level of treatment quality.
Treatment with dabigatran does not have any drug–food
interactions, has very limited drug–drug interactions, and
does not require anticoagulation monitoring. An analysis
of the use of dabigatran in AF patients eligible for anticoagulation treatment who are currently under-treated
indicated that this would be a cost-effective strategy16;
however, this should be investigated further in the
Danish setting.
An indicator for the quality of the warfarin treatment is
the time in therapeutic range (TTR). As a part of the
sensitivity analysis, a hypothetical situation with a TTR
similar to the highest quartile of the centre-based TTR
sub-group analysis of RE-LY for warfarin was compared
with dabigatran treatment. Also in this situation, dabigatran was demonstrated to be a cost-effective alternative to
warfarin. In Denmark, the perception is that the TTR for
warfarin treatment is relatively high compared to the TTR
for other countries, as also reported for the Danish centres
participating in RE-LY20. However, very little is known
about the actual achieved TTR for the majority of
Danish patients who are managed outside specialized
anticoagulation clinics. The introduction of dabigatran
holds the promise of a less diverse level of anticoagulation
treatment among different patient groups and most importantly a predictable high level of anticoagulation
treatment.
In the base-case analysis a conservative approach to
costing was applied. In regards to INR monitoring an estimate from the low range of tariffs that are paid to healthcare providers for managing patients was applied. This is
done in order to be conservative in evaluating dabigatran
compared to the established treatment option. For the
702
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
Danish healthcare budget holders, however, the actual—
higher—tariffs are more relevant. Sensitivity analysis
showed that when the actual tariffs are used, dabigatran
dominates warfarin, i.e., resulting in lower long-term cost
and better long-term outcomes. Also for the cost of stroke
a conservative estimate was applied. The Danish National
Board of Health refers to an estimate of the total first year
cost of stroke of E21,500 at 2001 cost levels31.
Furthermore, it has been shown that strokes related to
atrial fibrillation are more costly compared with strokes
not related to atrial fibrilation32,33.
Conclusions
Dabigatran is a new anticoagulant treatment for patients
with non-valvular atrial fibrillation. Based on findings
from an economic model based on the results from
RE-LY, dabigatran was found to be cost-effective from a
Danish healthcare perspective when compared to
warfarin.
Transparency
Declaration of funding
The project was funded by Boehringer Ingelheim, Denmark.
Declaration of financial/other relationships
Lars K. Langkilde was a paid consultant for Boehringer Ingelheim
for this project. Mikael Bergholdt Asmussen and Mikkel
Overgaard are employees of Boehringer Ingelheim Denmark.
Acknowledgements
Sonja V. Sorensen and Anuraag R. Kansal from United
BioSource Corporation developed the initial model15,16 and
kindly provided us access to the model and a detailed description
of methods and assumptions. Furthermore, we thank Steen
Husted, MD PhD, Regional Hospital West Jutland, Jens
Flensted Lassen, MD PhD, Aarhus University Hospital, and
Anna-Marie Mu¨nster, MD PhD, Aalborg Hospital, for acting as
advisors on clinical practice in anti-thrombotic management in
Denmark and giving valuable input on the analysis plan and
interpretation of results.
References
1. Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial
fibrillation: the Framingham Heart Study. Circulation 2004;110:1042–6
2. Heeringa J, van der Kuip DAM, Hofman A, et al. Prevalence, incidence, and
lifetime risk of atrial fibrillation: The Rotterdam Study. Eur Heart J 2006;
27:946-53
3. Miyasaka Y, Barnes ME, Gersh BJ, et al. Secular trends in incidence of atrial
fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on
the projections of future prevalence. Circulation 2006;114:119-25
4. Institute For Rational Pharmacotherapy. Behandling af atrieflimren. Rationel
Pharmakoterapi Juni 2008 No 6
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
5. The Task Force for the Management of Atrial Fibrillation of the European
Society of Cardiology. Guidelines for the management of atrial fibrillation.
Eur Heart J 2010;31:2369-429
6. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor
for stroke: the Framingham Study. Stroke 1991;22:983-8
7. Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to prevent
stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med
1999;131:492-501
8. Pengo V, Pegoraro C, Cucchini U, et al. Worldwide management of oral
anticoagulant therapy: the ISAM study. J Thromb Thrombolysis 2006;
21:73-7
9. Ansell J, Hollowell J, Pengo V, et al. Descriptive analysis of the process and
quality of oral anticoagulation management in real-life practice in patients
with chronic non-valvular atrial fibrillation: the international study of anticoagulation management (ISAM). J Thromb Thrombolysis 2007;23:83-91
10. Amouyel P, Mismetti P, Langkilde LK, et al. INR variability in atrial fibrillation: a
risk model for cerebrovascular events. Eur J Intern Med 2009;20:63-9
11. Friberg J, Gislason GH, Gadsbøll N, et al. Temporal trends in the prescription
of vitamin K antagonists in patients with atrial fibrillation. J Intern Med
2006;259:173-8
12. Palnum KH, Mehnert F, Andersen G, et al. Medical Prophylaxis following
hospitalization for Ischemic Stroke: age- and sex-related differences and
relation to mortality. Cerebrovasc Dis 2010;30:556-66
13. Ezekowitz MD, Connolly S, Parekh A, et al. Rationale and design of RE-LY:
randomized evaluation of long-term anticoagulant therapy, warfarin, compared with dabigatran. Am Heart J 2009;157:805-10
14. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in
patients with atrial fibrillation. N Engl J Med 2009;361:1139-51
15. Sorensen SV, Dewilde S, Singer DE, et al. Cost-effectiveness of warfarin: trial
versus ‘‘real-world’’ stroke prevention in atrial fibrillation. Am Heart J
2009;157:1064-73
16. Sorensen SV, Kansal AR, Connolly S, et al. Cost-effectiveness of
dabigatran etexilate for the prevention of stroke and systemic embolism in
atrial fibrillation: a Canadian payer perspective. Thromb Haemost 2011;
105:908-19
17. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification
schemes for predicting stroke: results from the National Registry of Atrial
Fibrillation. JAMA 2001; 285:2864-70
18. Roskell NS, Lip GY, Noack H, et al. Treatments for stroke prevention in atrial
fibrillation: a network meta-analysis and indirect comparisons versus dabigatran etexilate. Thromb Haemost 2010;104:1106-15
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Volume 15, Number 4
August 2012
19. Statistics Denmark Table HISB8. Life table (2 years tables) by age, sex and life
table (1981:1982-2009:2010). Copenhagen: Statistics Denmark. 2010
http://www.statistikbanken.dk. Accessed December 13, 2010
20. Wallentin L, Yusuf S, Ezekowitz MD, et al, on behalf of the RE-LY investigators.
Efficacy and safety of dabigatran compared with warfarin at different levels of
international normalised ratio control for stroke prevention in atrial fibrillation:
an analysis of the RE-LY trial. Lancet 2010; 6736:61194-4
21. Danish Medicines Agency. Medicinpriser. Copenhagen: Danish Medical
Agency, 2011 http://www.medicinpriser.dk. Accessed March 15, 2011
22. Danish Society of Cardiology. Antitrombotisk behandling ved kardiovaskulære
sygdomme. Trombokardiologi. DCS Guidelines. Copenhagen: Danish Society
of Cardiology, 2007
23. Ministry of Health. Takstsystem 2011 – vejledning. Copenhagen: Ministry of
Health, 2011
24. Danish Medical Association. Honorartabel 01-04-2011. Copenhagen: Danish
Medical Association, 2011
25. Porsdal V, Boysen G. Direct costs during the first year after intracerebral
hemorrhage. Eur J Neurol 1999;6:449-54
26. Ghatnekar O, Persson U, Glader EL, et al. Cost of stroke in Sweden: an
incidence estimate. Int J Technol Assess Health Care 2004;20:375-80
27. Statistics Denmark. Table PRIS8. Consumer price index, annual average
(1900=100) by type (1900-2010). Copenhagen: Statistics Denmark, 2010
http://www.statistikbanken.dk. Accessed December 10, 2010
28. Holm T, Lassen JF, Genefke J, et al. Cross-sectorial cooperation between
general practice and hospital – shared care elucidated using anticoagulant
therapy as an example. Medicinsk Teknologivurdering – puljeprojekter.
Copenhagen: National Board of Health, 2006
29. Hvelplund A, Galatius S, Madsen M, et al. Women with acute coronary syndrome are less invasively examined and subsequently less treated than men.
Eur Heart J 2010;31:684-90
30. Committee on Health Promotion and Disease. Vi kan leve længere og sundere.
Forebyggelseskommisionens anbefalinger til en styrket forbyggende indsats.
Report. Copenhagen: Ministry of Health, Committee on Health Promotion and
Disease Prevention, 2009
31. National Board on Health. Referenceprogram for behandling af patienter med
apopleksi. Copenhagen: National Board On Health, 2005
32. Ghatnekar O, Glader EL. The effect of atrial fibrillation on stroke-related
inpatient costs in Sweden: a 3-year analysis of registry incidence data from
2001. Value Health 2008;11:862-8
33. Bru¨ggenju¨rgen B, Rossnagel K, Roll S, et al. The impact of atrial fibrillation on
the cost of stroke: the Berlin acute stroke study. Value Health 2007;10:137-4
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
703
ISSN: 1369-6998 (Print) 1941-837X (Online) Journal homepage: http://www.tandfonline.com/loi/ijme20
Cost-effectiveness of dabigatran etexilate for
stroke prevention in non-valvular atrial fibrillation.
Applying RE-LY to clinical practice in Denmark
Lars K Langkilde, Mikael Bergholdt Asmussen & Mikkel Overgaard
To cite this article: Lars K Langkilde, Mikael Bergholdt Asmussen & Mikkel Overgaard
(2012) Cost-effectiveness of dabigatran etexilate for stroke prevention in non-valvular atrial
fibrillation. Applying RE-LY to clinical practice in Denmark, Journal of Medical Economics, 15:4,
695-703
To link to this article: http://dx.doi.org/10.3111/13696998.2012.673525
Accepted author version posted online: 07
Mar 2012.
Published online: 22 Mar 2012.
Submit your article to this journal
Article views: 389
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ijme20
Download by: [Boehringer Ingelheim]
Date: 10 February 2016, At: 01:22
Journal of Medical Economics
1369-6998
doi:10.3111/13696998.2012.673525
Vol. 15, No. 4, 2012, 695–703
Article 0001.R1/673525
All rights reserved: reproduction in whole or part not permitted
Lars K Langkilde
Abstract
se oa
d,
C
No op
Un t
yr
au fo
ig
th
r
ht
di or S
sp ize a
la d le ©
y, us
20
vi e o
ew p r
1
r
2I
C
o
an h
o
nf
d p ibi m
t
rin ed m
or
t a . Au e
m
si th rc
aU
ng or i
le is al
K
co ed
D
py us is Lim
fo ers tr
ite
rp c i
b
a
er n
d
so d ut
na ow io
l u nl n
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Original article
Cost-effectiveness of dabigatran etexilate
for stroke prevention in non-valvular atrial
fibrillation. Applying RE-LY to clinical
practice in Denmark
Wickstrøm & Langkilde ApS, Vejle, Denmark
Mikael Bergholdt Asmussen
Mikkel Overgaard
Boehringer Ingelheim Denmark, Copenhagen,
Denmark
Address for correspondence:
Lars K Langkilde, Wickstrøm & Langkilde ApS,
Hjulmagervej 4A Vejle, DK-7100, Denmark.
Tel.: þ4541111183; Fax: +45 3514 1110;
ll@wlpharma.com
Key words:
Anti-coagulation – Dabigatran etexilate – Warfarin –
Atrial fibrillation – Cost-effectiveness – Denmark
Accepted: 5 March 2012; published online: 22 March 2012
Citation: J Med Econ 2012; 15:695–703
Objective:
To estimate the economic implications of introducing dabigatran etexilate (‘dabigatran’) for anti-coagulation
therapy in Danish patients with non-valvular atrial fibrillation based on results of the RE-LY trial.
Methods:
The lifetime cost and outcomes of dabigatran and warfarin were estimated using a previously published
cost-effectiveness model. The model utilizes the data from the RE-LY study to estimate the costs and
outcomes of stroke prevention in atrial fibrillation. Cost estimates were based on official Danish tariffs and
prices, and published literature on the cost of stroke. In the base-case analysis a conservative approach was
adopted applying tariffs from the lowest range for the cost of International Normalized Ratio (INR) monitoring
associated with warfarin. The effectiveness measure of the analysis was quality-adjusted life-years (QALY).
Results:
The model estimated that the mean cost per patient for the remaining life-time is E16,886 treated with
warfarin and E18,752 treated with dabigatran. This was associated with mean QALYs per patient of 8.32
with warfarin and 8.59 with dabigatran. The resulting incremental cost-effectiveness ratio (ICER) of
E7000 per QALY gained is regarded as cost-effective by Danish standards. This conclusion was seen
to be robust to realistic variations in input parameters, including adjustment for the RE-LY centres achieving
the best INR monitoring quality. Threshold analysis revealed that dabigatran would be cost-saving in settings
where the cost of warfarin monitoring exceeds E744 per year.
Limitations:
The analysis does not include all aspects of Danish clinical practice anti-coagulation that will influence costeffectiveness of dabigatran, e.g., this study did not attempt to model quality of anticoagulation monitoring
and under-utilization in clinical practice.
Conclusions:
Based on the outcomes observed in the RE-LY trial, dabigatran represents a cost-effective alternative to
warfarin in Denmark for all patients with atrial fibrillation within the licensed indication of dabigatran.
Introduction
Atrial fibrillation (AF) is the most common cardiac arrhythmia, which increases
morbidity and mortality principally due to the resultant 5-fold increased risk of
stroke compared to patients without AF. Persons above 40 years of age have a
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
695
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
25% lifetime risk of developing AF1. Based on international literature1–3, Brandes4 estimated that more than
50,000 Danish patients suffer from AF, with prevalence
expected to treble by the year 2050.
Anticoagulation is recommended for AF patients with
risk factors for thromboembolic events5. The use of anticoagulation with vitamin K antagonists has evolved over
time. Two decades ago it was generally limited to AF
patients with rheumatic heart disease and prosthetic
heart valves6. With growing evidence from randomized
controlled trials an increasing number of AF patients are
recommended for anticoagulation treatment5.
The historical standard for anticoagulation treatment
in Denmark has been the vitamin K antagonist warfarin.
Warfarin reduces the risk of stroke by 60% compared to
placebo and 30–40% compared to treatment with aspirin7.
Warfarin can be cumbersome to use because of the multiple drug–drug and drug–food interactions, and the significant intra- and inter-patient variability in terms of
therapeutic response. Achieving a therapeutic dose in
clinical practice requires regular monitoring of the
International Normalized Ratio (INR) and dose adjustment. The quality of INR monitoring has been shown to
vary between countries8 and is related to how anticoagulation management is organized9, with implications for reallife effectiveness10. However, in Denmark, valid data on
warfarin management performance is not available. The
majority of Danish warfarin patients are managed by their
general practitioner.
Friberg et al.11 reported that only 36% of patients hospitalized with a diagnosis of AF in Denmark claimed a
prescription of vitamin-K antagonists (i.e., warfarin)
within 3 months of discharge. Furthermore, a recent
Danish study of secondary prophylaxis showed that
among patients older than 80 years only 25% of patients
with atrial fibrillation and no contraindications for anticoagulant therapy at the time of the initial hospital admission were treated with oral anticoagulation 1 year after the
initial ischemic stroke12.
Dabigatran etexilate (‘dabigatran’) is a new oral direct
thrombin inhibitor which provides a predictable anticoagulant effect with no need for anticoagulation monitoring.
In 2008, dabigatran was approved in Europe for primary
prevention of venous thromboembolic events in adult
patients who have undergone elective total hip or knee
replacement surgery. In August 2011 dabigatran was
approved in Denmark for stroke prevention for patients
with non-valvular AF with at least one risk factor.
In the Randomized Evaluation of Long-term
Anticoagulation Therapy (RE-LY) trial (n ¼ 18,113), the
efficacy and safety of dabigatran was compared with warfarin in patients with AF with risk factors for stroke13,14.
Two doses of dabigatran were tested and dabigatran
150 mg twice daily was associated with a significant
lower rate of stroke/systemic embolism and a comparable
696
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
rate of major haemorrhage compared to warfarin, whereas
dabigatran 110 mg twice daily was associated with a comparable rate of stroke and systemic embolism and a significant lower rate of major haemorrhage compared to
warfarin. Both doses significantly reduced the rate of intracranial and life-threatening bleeding compared to
warfarin14.
With an increasing prevalence of AF and a lower
threshold for initiating anticoagulation therapy it is relevant to consider which treatment option represents the
best value-for-money. The purpose of this analysis is to
estimate the long-term cost-effectiveness of stroke prevention in patients with AF using dabigatran as an alternative
to warfarin in a Danish setting.
Methods
Model of clinical events
We adapted an economic model of long-term (remaining
life-time) cost and outcomes based on the RE-LY trial
published by Sorensen et al.15,16.The model has Markov
properties, i.e., it is assumed that any patient’s health
can be represented as a finite number of health states.
Transitions between states can occur once every 3
months with transition probabilities conditional on the
present health state, baseline characteristics of the patient,
and progression of time. The health states represent ischemic stroke history, degree of disability following stroke
(independent, moderately disabled, or dependent), current
treatment (first line, primary prophylaxis using warfarin/
dabigatran treatment, second line prophylaxis using aspirin, or no treatment), and death. The model estimates the
distribution of patients on health states from entry into
the model and until all patients are dead, i.e., the time
perspective is remaining life-time. Calculations were performed in Microsoft Excel 2007.
Clinical events modelled include ischemic stroke,
transient ischemic attack, systemic embolism, acute myocardial infarction, intracranial haemorrhage and haemorrhagic stroke, other major haemorrhages, and minor
bleeds.
The publication by Sorensen et al.16 analyses the longterm cost and outcomes from a Canadian perspective and
was adapted further to the Danish healthcare setting. The
Canadian cost-effectiveness model is suitable for this adaptation because its main analysis was based on the RE-LY
data and applies the sequential dosing strategy of dabigatran that is common between the approved European and
Canadian labels. The posology reflected in the costeffectiveness model initiates patients under the age of 80
years at dabigatran 150 mg twice daily followed by 110 mg
twice daily at the age of 80 years and over.
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Journal of Medical Economics
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Risk of events
In our adaptation we used the baseline analysis of the
Canadian adaptation which is an extrapolation of outcomes in the RE-LY trial to long-term (lifetime) outcomes
and cost. Below we will only provide a short summary of
the modelling approach. For full details please consult
Sorensen et al.16.
Risk of clinical events for patients on warfarin and
dabigatran were based on observed event rates in the
RE-LY trial. Separate risk estimates for all modelled clinical events were estimated for the sub-group of patients
respectively younger than or at least 80 years of age at
baseline16.
Risk of ischaemic stroke was estimated separately for
each sub-group of patients based on the CHADS2—
scoring of risk factors17 at baseline. In the model the
CHADS2 is reassessed every cycle to reflect changes in
the risk factors as a result of predicted clinical events
during the previous 3-month cycle.
Treatment and treatment effects
Patients entering the economic model are assumed to
receive either warfarin (target INR 2–3) or dabigatran.
Patients are assumed to discontinue from any of the treatments in the case of an intracranial haemorrhage (ICH).
Furthermore, we assume that 50% of other major haemorrhages lead to permanent discontinuation of antithrombotic treatment. These patients are assumed to be
at risk of clinical events at rates that reflect those of placebo-treated patients in published clinical trials, and
an increased mortality risk depending on the level of
disability16. Furthermore, treatment switches (from warfarin or dabigatran to aspirin) can occur in the model for
other reasons. In this case patients are assumed to experience risks of clinical events and costs as if treated with
aspirin. Rates of discontinuation for warfarin and dabigatran were estimated using data from RE-LY.
For warfarin and dabigatran the events rates were estimated from RE-LY. For aspirin (modelled as a second line
treatment option) and no treatment (used for modelling
the clinical events for patients who have discontinued all
anti-thrombotic treatment following a major haemorrhagic event) event rates are based on an adaptation of a
recently published meta-analysis of controlled clinical
trials involving warfarin, aspirin, and/or placebo18.
Estimation of health outcomes
Long-term health outcomes were estimated as number of
clinical events, life expectancy (life-years), and qualityadjusted life-years (QALYs).
QALYs were estimated by weighting 3-month survival
by a utility weight for the health state during that cycle. If a
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Volume 15, Number 4
August 2012
clinical event occurred during a cycle, a temporary utility
decrement was subtracted from the health state utility.
Utility weights relevant to AF in a Danish population
are not available; hence we used the same values as applied
in Sorensen et al.16.
Adaptation of clinical event model
Danish age- and gender-specific death rates from national
population statistics19 were applied. In order not to
double-count causes of deaths already accounted for in
the model as event-related mortality, only death from
other causes than stroke, bleeding, systemic emboli, and
acute myocardial infarction was included in the calculation of death rates.
Furthermore, since the efficacy and safety of warfarin
are related to the achieved quality of INR monitoring, a
sensitivity analysis was planned allowing the level of timein-therapeutic range (TTR) to deviate from that of the
warfarin treated patients in the RE-LY trial. The costs
and outcomes of dabigatran and warfarin were compared
using results from a post-hoc analysis of RE-LY based on
centre average TTR20. Relative risk ratios were reported by
quartile of patients according to the mean TTR achieved
by the trial centre that the patient attended during RE-LY.
The patients from the lowest quartile represent models of
care achieving a low level of TTR (centres with an average
TTR of less than 57.1%) and the highest quartile represent
models of care with a high level of TTR for INR monitoring (centres with average TTR at least 72.6% or above).
The relative risk ratios reported20 for ischaemic stroke and
systemic embolism, intra-cranial bleeding, and major
bleedings for these two extreme quartiles were applied as
sensitivity analyses.
Cost of medical products
Costs of medicinal products are based on the current pharmacist sales prices including dispensing fees but excluding
VAT21. For warfarin a daily cost of treatment based on a
5 mg daily dosage is assumed22.
Estimation of cost
The direct costs associated with anti-thrombotic treatment and clinical events were estimated using primarily
national tariffs where available23,24 and data from other
publicly-available sources otherwise. Transferrals such as
pensions, work disability compensation, and value
added tax (VAT) were excluded. Table 1 provides an overview of the included cost components in the model. In the
following all cost are reported in euro (E) based on the
2011 annual average currency rate (100 Danish
Kroner ¼ 13.42 E).
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
697
Journal of Medical Economics
Volume 15, Number 4
August 2012
Table 1. List of unit cost estimates applied in the model. In Euro (E).
Item
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Ischaemic stroke
3 month cost
Initial hospitalization
Other
Total
Lifelong cost
Heamorrhagic stroke
3 month cost
Initial hospitalization
Other
Total
Lifelong cost (dependent only)
Systemic embolism
Transient Ischemic Attack
Acute Myocardial Infarction
Major extracranial haemorrhage
Fatal
Baseline
18,771
1701
20,472
1287
16,665
1359
18,024
1028
3228
2253
8629
3223
Non-fatal/non-GI
1765
Non-fatal/GI
2623
Treatment switch/discontinuation
INR monitoring (per year)
17
513
Dabigatran initiation
176
Dabigatran maintenance
17
Dabigatran
Warfarin
2.63
0.26
Aspirin
0.09
Notes
(per 3 months)
(per 3 months)
DRG 114 (kidney
embolism)
DRG 125 (TIA)
See text
DRG 638 (GI
bleed age
417 with
complication)
DRG 323 (Blood
loss)
DRG 640
(Uncomplicated GI bleed
age 417)
GP visit
Per year. See
Table 2
First 3 months.
See Table 2
Per year. See
Table 2
Per day
Per day (assume
2 2.5 mg
dose of
warfarin)
Per day (assume
75 mg)
Cost of stroke
Cost of stroke is a key parameter in the costing of clinical
events. Tariffs are available for the initial hospitalization
in the Danish Disease Related Groups (DRG) classification of hospitalizations. However, it was not possible to
identify up-to-date Danish studies of resource utilization
associated with stroke (re-hospitalization, rehabilitation,
and social costs) following the initial hospitalization.
Porsdal and Boysen25 performed a comprehensive estimation of first year resource utilization following ischaemic and haemorrhagic stroke based on cost applicable to
1995. The hospital cost during the initial hospitalization
constituted 63.4% of the total first year direct medical and
social cost in patients admitted with ischaemic stroke and
59.4% of first year cost in haemorrhagic stroke. The initial
hospitalization cost for ischaemic stroke and ICH in the
model was based on Porsdal and Boysen25.
698
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
The first year follow-up cost from Porsdal and Boysen25
could not be used directly because it included cost related
to secondary strokes occurring within the first year. Since
secondary stroke is estimated and assigned a cost in our
clinical model this would imply double counting cost of
secondary strokes. From a Swedish study on cost of stroke
with 4 years follow-up26 it was possible to estimate that
11% of first year follow-up cost was related to secondary
stroke. The Danish estimate of first year follow-up cost was
reduced by this amount.
Furthermore, the long-term cost of stroke was estimated. The Swedish study reported that social cost and
follow-up cost are highest in the first year following the
initial stroke event and fairly stable the second through
fourth year. The annual cost years 2–4 was 24% lower
than first year follow-up cost. This relative reduction in
cost (based on the Swedish observation) was applied to our
estimate of the Danish first year follow-up cost. In order to
use the estimated long-term follow-up cost conservatively
we only applied it to patients with moderate or severe
disability after the initial stroke event. Estimates of
stroke cost were inflated to 2011 values using the consumer
price index27.
Cost of patient management
INR monitoring is organized in a number of different ways
in Denmark28. For the base-case analysis, the cost of INR
monitoring was estimated based on the most common and
least costly set-up where INR monitoring is managed by
general practitioners. Based on publicly available tariffs
and assumed annual number of contacts, the base-case
cost for INR monitoring was estimated to E513. In sensitivity analyses, the impact of different scenarios for INR
monitoring costs was analysed. In one scenario—reflecting
the split between primary care and hospital organized INR
monitoring in Denmark—the cost of INR monitoring was
estimated as a weighted average of cost for the three most
frequent models of care in Denmark. For each model of
care the annual number of contacts to the provider and the
proportion of the patient population being monitored were
assessed. Cost per contact was estimated using publicly
available tariffs. The resulting annual cost per patient is
E975 per year (weighted average based on estimated
patient distribution by models of care, see Table 2). The
publicly-available tariffs especially for hospital contacts
have been shown to deviate from variable cost of monitoring, i.e., cost estimated from the pure work input and materials used during one contact28. In another scenario only
variable cost per contact were taken into consideration.
Variable costs are—as also stated by Holm et al.28—only a
relevant measure in short-term analyses, since they do not
take into account the overhead cost associated with running the clinic or the alternative use of clinic space for
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Journal of Medical Economics
Volume 15, Number 4
August 2012
Table 2. Calculation of annual patient management cost on dabigatran and warfarin (Euro).
Frequency
Warfarin
GP (60%)
3rd party lab (50%)
Point of care instrument (50%)
Hospital (40%)
Paper-based (27%)
AC clinic (70%)
Patient self-monitoring (3%)
Weighted average (warfarin)
Dabigatran
1 year
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
2þ years
Contacts
per year
4 weeks
3.5 weeks
5 weeks
5 weeks
2 weeks
Specialist
GP
Total
GP
Description
Annual
cost (E)
Visit, test and shipment, follow-upa
Visit and follow-upa
Outpatient visitb
Outpatient visitb
Consultation by phone/internetb
513
516
1654
1654
2068
975
1
3
Outpatient visitb
Visita
1
Visita
159
52
211
17
AC, Anti-coagulation; GP, general practitioner.
a
Unit cost (E): visit 17.38, blood test and shipment 8.94, INR (PoC) 14.63, phone consultation 3.41, mail follow-up 5.47. Assumes 50% of follow-up is done faceto-face and 50% via phone/mail.
b
Unit cost (E): Outpatient visit 159.05, phone contact 79.52.
other purposes. Variable cost was estimated based on experience from a hospital-based anticoagulation clinic using
computer-assisted dosing (calculations not shown). The
resulting variable cost of monitoring was estimated to
E418 per year.
There is no need for anticoagulation monitoring or dose
adjustment with dabigatran. However, a temporary higher
cost for initiation of dabigatran and a permanent cost of
dabigatran management were included. It was assumed
that all patients will be seen by a cardiologist (outpatient
visit) at initiation of dabigatran treatment and by a GP
three times during the first year and once yearly thereafter
specifically related to dabigatran. In reality, dabigatranrelated issues are likely to be part of the general patient
management and dealt with in the course of other AF
related visits. However, in the model these are considered
additional to any other visits the patients may have as a
consequence of their health state. Renal function testing
will also occur at those visits, according to the Danish
label.
Cost of bleeding events
Cost of ICH was based on published cost of stroke estimates25 as outlined above. Major bleeding events other
than intracranial haemorrhage were estimated using
DRG tariffs. Costs of fatal and non-fatal gastrointestinal
(GI) bleedings were estimated using the tariff for GI bleedings, respectively, with and without complications, while
costs for non-GI major bleedings were estimated using the
tariff for inpatient treatment of blood loss. Minor bleedings
were not applied any cost in the base-case analysis. In
sensitivity analyses a cost range of 20% of the baseline
estimates (both intracranial and other major bleedings)
! 2012 Informa UK Ltd www.informahealthcare.com/jme
were tested. We applied the higher bleeding cost to both
treatments and—in a separate sensitivity analysis—only to
dabigatran-treated patients and for this analysis we also
applied the cost of a GP visit for minor bleedings in dabigatran-treated patients. The asymmetric analysis of bleeding cost was applied to test whether an increase in resource
utilization associated with the new medicine would be critical for the conclusion.
Cost of acute myocardial infarction (AMI)
Cost of AMI is based on observed use of revascularization and imaging procedures in 24,952 Danish patients
with incident ACS29. Applying 2011 tariffs to the casemix reported leads to an estimated cost per event of
E8629.
Discounting
All future costs and outcomes (QALYs) were discounted
in accordance with standards set by the Danish Ministry of
Health30 at an annual rate of 2%, with 0% and 4% discounting tested in sensitivity analyses.
Patient cohort
We estimated outcomes and cost for a modelled cohort of
10,000 patients with a similar risk profile (described by
CHADS2) to the patients aged 80 or below included in
the RE-LY trial. The average age was 69 years and 65% of
patients were male.
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
699
Journal of Medical Economics
Volume 15, Number 4
August 2012
Results
Cost and outcomes
Table 3 reports the estimation of long-term cost and outcomes for the baseline analysis. The estimated total cost
per patient for the remaining life-time is E16,886 for
patients treated with warfarin and E18,752 for patients
Table 3. Baseline cost and quality-adjusted life-years per patient (life-long
projection) by treatment strategy. Incremental cost-effectiveness ratio
(warfarin compared to dabigatran).
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Sequentiala
Dabigatran use
Direct costs/Patient
Drug costs and Monitoring (E)
8232
Event costs (E)
8787
Follow-up costs (E)
1733
Total (E)
18,752
Incremental costs (E)
1866
Benefits/Patient (discounted at 2% per annum)
b
0.47
Expected number of events
Expected survival (years)
11.61
QALYs
8.59
Incremental QALY
0.27
ICER (Dabigatran vs warfarin)
Direct Costs (E) per QALY gained
6950
treated with dabigatran. The additional cost associated
with dabigatran (E1866) is mainly driven by the additional cost of medication. The estimated QALYs per
patient are 8.32 for warfarin and 8.59 for dabigatran. The
expected survival (discounted by 2% per annum) is 11.34
and 11.61 years, for warfarin and dabigatran, respectively.
The resultant incremental cost per QALY gained is
E6950. There is no officially accepted threshold for incremental cost per QALY in Denmark. However, an ICER
below 30,000 E/QALY is regarded as cost-effective, and,
hence, dabigatran is considerably below this threshold
compared to warfarin in the baseline analysis.
Warfarin
Sensitivity analysis
4935
9727
2224
16,886
0.55
11.34
8.32
QALY, Quality adjusted life years; ICER, incremental cost effectiveness ratio.
150 mg b.i.d. for age below 80 then 110 mg b.i.d.
Stroke, systemic embolism, and intracranial haemorrhage.
a
b
Table 4 reports the main findings from the sensitivity analysis. Two different scenarios for INR monitoring cost and/
or a high quality of INR monitoring were analysed.
If INR monitoring costs were based on a weighted average of the three most frequent models of care in Denmark,
the assigned cost of INR monitoring was E975. In this reallife scenario, dabigatran was shown to be dominant (i.e.,
lower cost and better outcome compared to warfarin). If
INR monitoring is assigned only the variable cost, the
incremental cost is E2633 per patient (compared to
E1866 in the base-case). This leads to an incremental
cost per QALY gained of E9806. Given these results dabigatran would be regarded as cost-effective compared to
Table 4. Univariate sensitivity analyses. Net present value of incremental cost and outcomes (dabigatran minus warfarin).
Incremental costs (E)
Incremental QALY
Baseline
1866
0.268
INR monitoring
Variable cost estimate
2633
0.268
Tariff based estimate
1704
0.268
Centres achieving TTR557.1%
1760
0.288
Centres achieving TTR 57.1–65.5%
1601
0.309
Centres achieving TTR 65.5–72.6%
2434
0.188
Centres achieving TTR472.6%
2952
0.102
Cost of stroke
20% lower
2188
0.268
20% higher
1544
0.268
Cost of bleeding
20% lower
1854
0.268
20% higher
1899
0.268
20% higher (dabigatran only)
2016
0.268
Discounting of future cost and outcomes (baseline 2% p.a.)
No discounting of future cost and outcomes
2114
0.263
Discounting at 4% p.a.
2183
0.272
Threshold analysis (annual cost of monitoring associated with equal lifetime cost of dabigatran and warfarin)
ICER E/QALY
6950
9806
–
6115
5175
12,923
29,019
8149
5752
6905
7075
7510
8037
8023
Threshold monitoring cost (E/year)
Baseline (all centres)
Centres achieving TTR557.1%
Centres achieving TTR 57.1–65.5%
Centres achieving TTR 65.5–72.6%
Centres achieving TTR472.6%
744
731
712
815
879
p.a., per annum; QALY, Quality-adjusted life years; TTR, Time in therapeutic interval.
*Incremental cost per QALY gained (dabigatran compared to warfarin).
700
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
Figure 1. Incremental cost and quality-adjusted life-years at varying level of thrombo-embologic and intra-cranial haemorrhage risk.
Note: Each cluster represents a level of ICH risk from 0.4 to 1.6 events per 100 patient years. Each point represents a level of CHADS2 from 0 (0.62
events/100 patient years) to 5–6 (2.77 event/100 patient years).
warfarin by any reasonable method of applying the cost of
INR monitoring under base-case assumptions.
Furthermore, it was shown in the sensitivity analysis that
if the annual cost of INR-monitoring exceeds E744, dabigatran would dominate warfarin.
Taken into consideration the achieved quality of INR
monitoring based on the centre average TTR, the incremental cost per QALY range from E6115 for the lowest
level of quality (centres with an average TTR of less than
57.1%) to E29,019 for the highest level of quality (centres
with average TTR at least 72.6% or above). Dabigatran
would be regarded as cost-effective to warfarin, irrespective
of average level of centre TTR.
The results were not sensitive to 20% changes in cost of
stroke or bleeding. In the case of bleeding cost we applied a
20% higher cost only to the dabigatran arm. This resulted
in an incremental cost per QALY of E7510, which is still
cost-effective compared to warfarin. The conclusion was
also found to be insensitive to the rate of discounting of
future cost and outcomes.
Figure 1 provides the results from sensitivity analysis
with respect to various combinations of risks of intracranial haemorrhage and stroke (based on CHADS2 score).
Each cluster represents a level of ICH risk (from 0.4 events
per 100 patient years to 1.6 events per 100 patient years)
and each point represents a CHADS2 score from 0–5/6. In
RE-LY the risk of ischaemic stroke on warfarin treatment
varied between 0.62 events per 100 patient years for
patients with a CHADS2 score of 0 at baseline to 2.77
events per 100 patient years for patients with a score of
5–6 at baseline16. The risk of ICH has a larger impact on
! 2012 Informa UK Ltd www.informahealthcare.com/jme
both incremental cost and incremental outcomes than the
risk of thromboembolic events. In general, higher levels of
risk are associated with lower incremental cost of dabigatran treatment and larger patient benefits (measured in
QALY gained) compared to warfarin. In all of the analysis
in Figure 1 the ICER is below E18,000 per QALY gained
and hence dabigatran remains cost-effective compared to
warfarin.
Discussion
This analysis evaluated the cost-effectiveness of dabigatran vs warfarin for patients with atrial fibrillation in a
Danish healthcare setting based on the results from the
RE-LY trial. In the base-case analysis dabigatran was
shown to be a cost-effective strategy for patients with
atrial fibrillation and the sensitivity analyses confirmed
the robustness of the base-case finding. The applied costeffectiveness model is—as with any extrapolation—
limited by the amount of evidence available. Hence, the
following limitations of this analysis should be mentioned.
The driving factor behind the costs and outcomes are
the risk levels for clinical events in the model. These are
primarily based on observed risk levels in the RE-LY trial
and the generalizability of the trial to the Danish setting in
terms of selection of patients, healthcare providers, and
patient management should be examined. This costeffectiveness does not address this real-life scenario with
patient exclusion however; Sorensen et al.16 showed that
the cost-effectiveness ratio will become even more
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
701
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
Volume 15, Number 4
August 2012
favourable for dabigatran in case a real-life scenario is
applied in addition to the applied RE-LY scenario.
Furthermore, we compare a new treatment option with
an established one. A number of assumptions were made to
estimate the annual dabigatran treatment cost, e.g., the
number of healthcare contacts, management of dabigatran-related bleedings, and treatment adherence.
Prospective data from a real-life setting will be needed to
provide more accurate estimates of dabigatran management and related cost.
Another aspect which is not reflected in this cost-effectiveness analysis is the substantial lack of treatment for
AF-patients who according to risk profile are recommended to receive anticoagulation therapy. Friberg
et al.11 documented—based on a Danish database study
(n ¼ 68,546)—that only 36% of the patients hospitalized
with a diagnosis of AF claimed a prescription of vitamin-K
antagonists (i.e., warfarin) within 3 months from discharge. The reasons for this lack of treatment may be
related to the fact that treatment with warfarin is considered as cumbersome due to the interactions with other
drugs and food and the frequent need for INR monitoring
to ensure an acceptable level of treatment quality.
Treatment with dabigatran does not have any drug–food
interactions, has very limited drug–drug interactions, and
does not require anticoagulation monitoring. An analysis
of the use of dabigatran in AF patients eligible for anticoagulation treatment who are currently under-treated
indicated that this would be a cost-effective strategy16;
however, this should be investigated further in the
Danish setting.
An indicator for the quality of the warfarin treatment is
the time in therapeutic range (TTR). As a part of the
sensitivity analysis, a hypothetical situation with a TTR
similar to the highest quartile of the centre-based TTR
sub-group analysis of RE-LY for warfarin was compared
with dabigatran treatment. Also in this situation, dabigatran was demonstrated to be a cost-effective alternative to
warfarin. In Denmark, the perception is that the TTR for
warfarin treatment is relatively high compared to the TTR
for other countries, as also reported for the Danish centres
participating in RE-LY20. However, very little is known
about the actual achieved TTR for the majority of
Danish patients who are managed outside specialized
anticoagulation clinics. The introduction of dabigatran
holds the promise of a less diverse level of anticoagulation
treatment among different patient groups and most importantly a predictable high level of anticoagulation
treatment.
In the base-case analysis a conservative approach to
costing was applied. In regards to INR monitoring an estimate from the low range of tariffs that are paid to healthcare providers for managing patients was applied. This is
done in order to be conservative in evaluating dabigatran
compared to the established treatment option. For the
702
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
Danish healthcare budget holders, however, the actual—
higher—tariffs are more relevant. Sensitivity analysis
showed that when the actual tariffs are used, dabigatran
dominates warfarin, i.e., resulting in lower long-term cost
and better long-term outcomes. Also for the cost of stroke
a conservative estimate was applied. The Danish National
Board of Health refers to an estimate of the total first year
cost of stroke of E21,500 at 2001 cost levels31.
Furthermore, it has been shown that strokes related to
atrial fibrillation are more costly compared with strokes
not related to atrial fibrilation32,33.
Conclusions
Dabigatran is a new anticoagulant treatment for patients
with non-valvular atrial fibrillation. Based on findings
from an economic model based on the results from
RE-LY, dabigatran was found to be cost-effective from a
Danish healthcare perspective when compared to
warfarin.
Transparency
Declaration of funding
The project was funded by Boehringer Ingelheim, Denmark.
Declaration of financial/other relationships
Lars K. Langkilde was a paid consultant for Boehringer Ingelheim
for this project. Mikael Bergholdt Asmussen and Mikkel
Overgaard are employees of Boehringer Ingelheim Denmark.
Acknowledgements
Sonja V. Sorensen and Anuraag R. Kansal from United
BioSource Corporation developed the initial model15,16 and
kindly provided us access to the model and a detailed description
of methods and assumptions. Furthermore, we thank Steen
Husted, MD PhD, Regional Hospital West Jutland, Jens
Flensted Lassen, MD PhD, Aarhus University Hospital, and
Anna-Marie Mu¨nster, MD PhD, Aalborg Hospital, for acting as
advisors on clinical practice in anti-thrombotic management in
Denmark and giving valuable input on the analysis plan and
interpretation of results.
References
1. Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial
fibrillation: the Framingham Heart Study. Circulation 2004;110:1042–6
2. Heeringa J, van der Kuip DAM, Hofman A, et al. Prevalence, incidence, and
lifetime risk of atrial fibrillation: The Rotterdam Study. Eur Heart J 2006;
27:946-53
3. Miyasaka Y, Barnes ME, Gersh BJ, et al. Secular trends in incidence of atrial
fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on
the projections of future prevalence. Circulation 2006;114:119-25
4. Institute For Rational Pharmacotherapy. Behandling af atrieflimren. Rationel
Pharmakoterapi Juni 2008 No 6
www.informahealthcare.com/jme ! 2012 Informa UK Ltd
Downloaded by [Boehringer Ingelheim] at 01:22 10 February 2016
Journal of Medical Economics
5. The Task Force for the Management of Atrial Fibrillation of the European
Society of Cardiology. Guidelines for the management of atrial fibrillation.
Eur Heart J 2010;31:2369-429
6. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor
for stroke: the Framingham Study. Stroke 1991;22:983-8
7. Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to prevent
stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med
1999;131:492-501
8. Pengo V, Pegoraro C, Cucchini U, et al. Worldwide management of oral
anticoagulant therapy: the ISAM study. J Thromb Thrombolysis 2006;
21:73-7
9. Ansell J, Hollowell J, Pengo V, et al. Descriptive analysis of the process and
quality of oral anticoagulation management in real-life practice in patients
with chronic non-valvular atrial fibrillation: the international study of anticoagulation management (ISAM). J Thromb Thrombolysis 2007;23:83-91
10. Amouyel P, Mismetti P, Langkilde LK, et al. INR variability in atrial fibrillation: a
risk model for cerebrovascular events. Eur J Intern Med 2009;20:63-9
11. Friberg J, Gislason GH, Gadsbøll N, et al. Temporal trends in the prescription
of vitamin K antagonists in patients with atrial fibrillation. J Intern Med
2006;259:173-8
12. Palnum KH, Mehnert F, Andersen G, et al. Medical Prophylaxis following
hospitalization for Ischemic Stroke: age- and sex-related differences and
relation to mortality. Cerebrovasc Dis 2010;30:556-66
13. Ezekowitz MD, Connolly S, Parekh A, et al. Rationale and design of RE-LY:
randomized evaluation of long-term anticoagulant therapy, warfarin, compared with dabigatran. Am Heart J 2009;157:805-10
14. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in
patients with atrial fibrillation. N Engl J Med 2009;361:1139-51
15. Sorensen SV, Dewilde S, Singer DE, et al. Cost-effectiveness of warfarin: trial
versus ‘‘real-world’’ stroke prevention in atrial fibrillation. Am Heart J
2009;157:1064-73
16. Sorensen SV, Kansal AR, Connolly S, et al. Cost-effectiveness of
dabigatran etexilate for the prevention of stroke and systemic embolism in
atrial fibrillation: a Canadian payer perspective. Thromb Haemost 2011;
105:908-19
17. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification
schemes for predicting stroke: results from the National Registry of Atrial
Fibrillation. JAMA 2001; 285:2864-70
18. Roskell NS, Lip GY, Noack H, et al. Treatments for stroke prevention in atrial
fibrillation: a network meta-analysis and indirect comparisons versus dabigatran etexilate. Thromb Haemost 2010;104:1106-15
! 2012 Informa UK Ltd www.informahealthcare.com/jme
Volume 15, Number 4
August 2012
19. Statistics Denmark Table HISB8. Life table (2 years tables) by age, sex and life
table (1981:1982-2009:2010). Copenhagen: Statistics Denmark. 2010
http://www.statistikbanken.dk. Accessed December 13, 2010
20. Wallentin L, Yusuf S, Ezekowitz MD, et al, on behalf of the RE-LY investigators.
Efficacy and safety of dabigatran compared with warfarin at different levels of
international normalised ratio control for stroke prevention in atrial fibrillation:
an analysis of the RE-LY trial. Lancet 2010; 6736:61194-4
21. Danish Medicines Agency. Medicinpriser. Copenhagen: Danish Medical
Agency, 2011 http://www.medicinpriser.dk. Accessed March 15, 2011
22. Danish Society of Cardiology. Antitrombotisk behandling ved kardiovaskulære
sygdomme. Trombokardiologi. DCS Guidelines. Copenhagen: Danish Society
of Cardiology, 2007
23. Ministry of Health. Takstsystem 2011 – vejledning. Copenhagen: Ministry of
Health, 2011
24. Danish Medical Association. Honorartabel 01-04-2011. Copenhagen: Danish
Medical Association, 2011
25. Porsdal V, Boysen G. Direct costs during the first year after intracerebral
hemorrhage. Eur J Neurol 1999;6:449-54
26. Ghatnekar O, Persson U, Glader EL, et al. Cost of stroke in Sweden: an
incidence estimate. Int J Technol Assess Health Care 2004;20:375-80
27. Statistics Denmark. Table PRIS8. Consumer price index, annual average
(1900=100) by type (1900-2010). Copenhagen: Statistics Denmark, 2010
http://www.statistikbanken.dk. Accessed December 10, 2010
28. Holm T, Lassen JF, Genefke J, et al. Cross-sectorial cooperation between
general practice and hospital – shared care elucidated using anticoagulant
therapy as an example. Medicinsk Teknologivurdering – puljeprojekter.
Copenhagen: National Board of Health, 2006
29. Hvelplund A, Galatius S, Madsen M, et al. Women with acute coronary syndrome are less invasively examined and subsequently less treated than men.
Eur Heart J 2010;31:684-90
30. Committee on Health Promotion and Disease. Vi kan leve længere og sundere.
Forebyggelseskommisionens anbefalinger til en styrket forbyggende indsats.
Report. Copenhagen: Ministry of Health, Committee on Health Promotion and
Disease Prevention, 2009
31. National Board on Health. Referenceprogram for behandling af patienter med
apopleksi. Copenhagen: National Board On Health, 2005
32. Ghatnekar O, Glader EL. The effect of atrial fibrillation on stroke-related
inpatient costs in Sweden: a 3-year analysis of registry incidence data from
2001. Value Health 2008;11:862-8
33. Bru¨ggenju¨rgen B, Rossnagel K, Roll S, et al. The impact of atrial fibrillation on
the cost of stroke: the Berlin acute stroke study. Value Health 2007;10:137-4
Cost-effectiveness of dabigatran etexilate for stroke prevention Langkilde et al.
703