JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
VOL. 66, NO. 24, 2015
ª 2015 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
ISSN 0735-1097/$36.00
PUBLISHED BY ELSEVIER INC.
http://dx.doi.org/10.1016/j.jacc.2015.09.084
Time to Cost-Effectiveness Following
Stroke Reduction Strategies in AF
Warfarin Versus NOACs Versus LAA Closure
Vivek Y. Reddy, MD,* Ronald L. Akehurst, MFPHM,y Shannon O. Armstrong, BA,z Stacey L. Amorosi, MA,x
Stephen M. Beard, MSC,k David R. Holmes, JR, MD{
ABSTRACT
BACKGROUND Left atrial appendage closure (LAAC) and nonwarfarin oral anticoagulants (NOACs) have emerged as
safe and effective alternatives to warfarin for stroke prophylaxis in patients with nonvalvular atrial fibrillation (AF).
OBJECTIVES This analysis assessed the cost-effectiveness of warfarin, NOACs, and LAAC with the Watchman device
(Boston Scientific, Marlborough, Massachusetts) for stroke risk reduction in patients with nonvalvular AF at multiple time
points over a lifetime horizon.
METHODS A Markov model was developed to assess the cost-effectiveness of LAAC, NOACs, and warfarin from the
perspective of the Centers for Medicare & Medicaid Services over a lifetime (20-year) horizon. Patients were 70 years of
age and at moderate risk for stroke and bleeding. Clinical event rates, stroke outcomes, and quality of life information
were drawn predominantly from PROTECT AF (Watchman Left Atrial Appendage System for Embolic Protection in Patients with Atrial Fibrillation) 4-year data and meta-analyses of warfarin and NOACs. Costs for stroke risk reduction
therapies, treatment of associated acute events, and long-term care following disabling stroke were presented in 2015
U.S. dollars.
RESULTS Relative to warfarin, LAAC was cost-effective at 7 years ($42,994/quality-adjusted life-years [QALY]), and
NOACs were cost-effective at 16 years ($48,446/QALY). LAAC was dominant over NOACs by year 5 and warfarin by
year 10. At 10 years, LAAC provided more QALYs than warfarin and NOACs (5.855 vs. 5.601 vs. 5.751, respectively).
In sensitivity analyses, LAAC remained cost-effective relative to warfarin ($41,470/QALY at 11 years) and NOACs
($21,964/QALY at 10 years), even if procedure costs were doubled.
CONCLUSIONS Both NOACs and LAAC with the Watchman device were cost-effective relative to warfarin, but LAAC was
also found to be cost-effective and to offer better value relative to NOACs. The results of this analysis should be considered
when formulating policy and practice guidelines for stroke prevention in AF. (J Am Coll Cardiol 2015;66:2728–39)
© 2015 by the American College of Cardiology Foundation.
A
trial
fibrillation
6
of this expense (2,3). For decades, warfarin has
million Americans, a number expected to
(AF)
affects
nearly
been the standard of care for stroke prophylaxis in
double by 2050 (1). Stroke is the most
AF. Although inexpensive and effective, warfarin
debilitating and costly consequence of AF. Medicare
is associated with an increased risk of bleeding,
spends an estimated $16 billion annually on AF,
lower
with AF-related stroke accounting for nearly one-half
nonadherence.
quality
of
life
(QoL),
and
Listen to this manuscript’s
audio summary by
JACC Editor-in-Chief
From the *Mount Sinai Medical Center, New York, New York; yUniversity of Sheffield, Sheffield, United Kingdom; zGfK, Wayland,
Dr. Valentin Fuster.
Massachusetts; xBoston Scientific, Marlborough, Massachusetts; kBresMed, Sheffield, United Kingdom; and the {Mayo Clinic,
Rochester, Minnesota. Dr. Reddy, Prof. Akehurst, Ms. Armstrong, and Mr. Beard are paid consultants to Boston Scientific,
manufacturer of the Watchman device discussed in this paper. Ms. Amorosi is a full-time employee of Boston Scientific. Dr.
Holmes and the Mayo Clinic have a financial interest in the Watchman device. Dr. Reddy is a paid consultant for and has received
grant support from Coherex and St. Jude Medical. Bruce D. Lindsay, MD, served as Guest Editor for this paper.
Manuscript received September 4, 2015; revised manuscript received September 24, 2015, accepted September 25, 2015.
high
patient
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
2729
Time to Cost-Effectiveness Following AF Treatment
More recently, 4 nonwarfarin oral anticoagulants
patients could experience clinical events
ABBREVIATIONS
(NOACs) (dabigatran, rivaroxaban, apixaban, and
leading to death, disability, and/or therapy
AND ACRONYMS
edoxaban) were approved by the U.S. Food and Drug
discontinuation and could incur associated
Administration (FDA). As a group, they are noninferior
costs and QoL adjustments.
to warfarin for ischemic stroke reduction and superior
Cost-effectiveness was evaluated using
for hemorrhagic stroke and all-cause mortality, but
the U.S. accepted willingness-to-pay thres-
they have an increased risk for gastrointestinal
hold of $50,000 per quality-adjusted life-year
bleeding (4). As with warfarin, the effectiveness of
(QALY) gained, and it was reported as the
these drugs is contingent on patient adherence.
incremental cost-effectiveness ratio (ICER),
For patients who are poor candidates for long-term
oral
anticoagulation,
percutaneous
left
which provides a standardized approach
AF = atrial fibrillation
ICER = incremental costeffectiveness ratio
LAAC = left atrial appendage
closure
MRS = modified Rankin score
NOAC = nonwarfarin oral
anticoagulant
atrial
to measure cost per unit of health improve-
PSA = probabilistic sensitivity
appendage closure (LAAC) is a device-based alterna-
ment in and across health states. Cost-
analysis
tive for stroke prophylaxis in AF (5). The first
effectiveness
FDA-approved LAAC device, the Watchman device
determine whether there was an observable
life-year
(Boston Scientific, Marlborough, Massachusetts), was
time horizon over which treatment options
QoL = quality of life
demonstrated to be noninferior to warfarin for
reached accepted levels of cost-effectiveness.
TIA = transient ischemic attack
was
assessed
annually
to
QALY = quality-adjusted
ischemic stroke reduction and superior for hemor-
MODEL STRUCTURE AND CLINICAL PATHWAYS. The
rhagic stroke and all-cause mortality. However, it
model began with patients assigned to 1 of 3 treatments
is
complica-
(Figures 1A and 1B). Patients in the LAAC arm faced a
tions that predictably diminish in frequency with
associated
with
procedure-related
1-time procedure-related risk, including ischemic
operator experience (6). Importantly, most patients
stroke resulting from air embolism (1.10%), major
can discontinue lifelong anticoagulation following
bleeding (0.60%), and pericardial effusion (4.80%).
Watchman device implantation (6).
Patients undergoing LAAC could experience a successful or failed implantation procedure. In accor-
SEE PAGE 2740
dance with the PROTECT AF (Watchman Left Atrial
Both clinical and economic value are important in
evaluating new therapies. Published U.S. economic
analyses
have
effectiveness
of
largely
oral
focused
on
the
cost-
anticoagulants,
and
these
reports invariably use a lifetime analysis spanning
20 to 35 years (7–9). Herein, we estimate the costeffectiveness of all stroke prevention strategies
available in the United States at multiple points
over a lifetime horizon. Although both NOACs and
LAAC would initially be expected to be more costly
than warfarin, their clinical benefits may supervene
with time, thereby altering cost-effectiveness. This
comprehensive assessment may aid clinicians and
health care decision makers in making informed
choices regarding patient treatment.
Appendage System for Embolic Protection in Patients
With Atrial Fibrillation) trial, patients who had successful device implantation were assumed to receive
warfarin for 45 days, aspirin plus clopidogrel from 46
days to 6 months, and aspirin thereafter. Following a
failed procedure, patients were assumed to continue
warfarin therapy. Patients receiving warfarin or
NOACs could discontinue therapy following a bleeding
event or for nonclinical reasons. Patients who discontinued primary drug therapy were assumed to
switch to aspirin. Discontinuation of second-line
therapy was assumed to result in no treatment.
On entering the model, patients were assumed to
be “well,” or in normal, good health. Patients transitioned to new health states following a clinical event
or death. Only ischemic and hemorrhagic stroke
METHODS
affected disability outcomes. Patients who had a
second stroke could either remain in the same health
A Markov model was developed in Excel to evaluate
state or worsen to greater disability. Transient
the cost-effectiveness of 3 treatment strategies:
ischemic attack (TIA) and systemic embolism led to
1. LAAC
with
the
Watchman
device
(Boston
Scientific)
2. NOACs as a class
3. Adjusted-dose warfarin
The model was constructed from the perspective
patients being well with a history of stroke, which
increased their risk of subsequent stroke. All events,
except TIA, could lead to death.
CLINICAL INPUTS. Clinical inputs were taken from
several sources (Table 1). For LAAC, procedural
complications and event probabilities were drawn
of the Centers for Medicare & Medicaid Services
from PROTECT AF at 4-year follow-up (6). Relative
(CMS), with a lifetime horizon (defined as 20 years)
risks for post-procedural stroke and bleeding were
and a 3-month cycle length. Within each cycle,
used to apply a standard efficacy estimate to the
2730
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
F I G U R E 1 Model Schematics Depicting LAAC and OAC Patient Pathways
A
LAAC
Procedure
Procedural
Complications?
YES
NO
Pericardial
Effusion
Procedural
Stroke
NO
Implant
Successful?
Device
Embolization
Major Bleeding
Warfarin
Therapy
YES
Therapy
Discontinued?
No Therapy
YES
Ischemic Stroke
Transient
Ischemic Attack
Systemic
Embolism
Hemorrhagic
Stroke
Extracranial
Hemorrhage
Myocardial
Infarction
Unrelated
Mortality
Nondisabling
Well with Stroke
History
Well with Stroke
History
Nondisabling
Well
Well
Death
Death
Moderate
Disability
Death
Death
Moderate
Disability
Severe
Disability
Severe
Disability
Death
Death
B
OAC
Therapy
Therapy
Discontinued?
YES
YES
Therapy
Discontinued?
Aspirin
Therapy
No Therapy
NO
Ischemic
Stroke
Transient
Ischemic
Attack
Systemic
Embolism
Nondisabling
Well with Stroke
History
Well with Stroke
History
Moderate
Disability
Hemorrhagic
Stroke
Extracranial
Hemorrhage
Myocardial
Infarction
Unrelated
Mortality
Well
Well
Death
Nondisabling &
Therapy
Change
Death
Death
Moderate
Disability &
Therapy
Change
Severe
Disability
Well & Therapy
Change
Death
Well &
Temporary
Therapy
Change
Severe
Disability &
Therapy
Change
Death
Death
(A) Left atrial appendage closure (LAAC) decision nodes include procedural complications, implantation success, and subsequent outcomes. (B) Oral
anticoagulant (OAC) decision nodes include therapy discontinuation and subsequent outcomes.
derived baseline risks. For warfarin, the relative risk
a pooled analysis of multiple AF clinical trials
of stroke was derived from a meta-analysis of
involving warfarin (4,16–20). NOAC clinical event
warfarin
rates were derived from a NOAC meta-analysis and
trials
(10).
Risk
data
for
all
other
warfarin-related clinical events were estimated from
from individual NOAC trials (4,12,13,21,22).
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
T A B L E 1 Clinical Inputs Derived From Meta-Analyses and Pivotal Trials
Value
Range
Distribution
0.76–1.14*
Lognormal
Source (Ref. #)
LAAC
RR of post-procedure ischemic stroke
(relative to warfarin)
0.95
Holmes et al., 2009;
Reddy et al., 2014 (5,6)
RR of hemorrhagic stroke (relative to warfarin)
0.15
0.03–0.49†
Lognormal
Reddy et al., 2014 (6)
RR of post-procedure major bleeding (relative to warfarin)
0.55
0.44–0.66
Lognormal
Reddy et al., 2014 (6)
Annual risk of systemic embolism
0.20%
RR of myocardial infarction (relative to warfarin)
0.50
Risk of minor bleeding
0.0%–0.4%†
Beta
Reddy et al., 2014 (6)
0.40–0.60
Lognormal
Reddy et al., 2014 (6)
On the basis of concomitant drug therapy
Warfarin
RR of ischemic stroke (relative to no therapy)
0.33
0.23–0.46†
Lognormal
Hart et al., 2007 (10)
RR of major bleeding (relative to HAS-BLED)
1.00
0.80–1.20
Lognormal
Pisters et al., 2010 (11)
41.80%
33.40%–50.20%
Beta
Ruff et al., 2014 (4)
0.11%
0.90%–0.12%†
Beta
Connelly et al., 2009;
Granger et al., 2011 (12,13)
Percentage of major bleeding that is
hemorrhagic stroke
Annual risk of systemic embolism
Annual risk of myocardial infarction
1.47%
0.53%–1.47%†
Beta
Ruff et al., 2014 (4)
Annual risk of minor bleeding
7.70%
0.80%–16.40%
Beta
Connelly et al., 2009;
Granger et al., 2011 (12,13)
NOACs
RR of ischemic stroke (relative to warfarin)
0.92
0.83–1.02†
Lognormal
Ruff et al., 2014 (4)
RR of hemorrhagic stroke (relative to warfarin)
0.48
0.39–0.59†
Lognormal
Ruff et al., 2014 (4)
RR of extracranial hemorrhage (relative to warfarin)
1.25
1.01–1.55†
Lognormal
Ruff et al., 2014 (4)
RR of systemic embolism
0.92
0.83–1.02†
Lognormal
Ruff et al., 2014 (4)
RR of myocardial infarction (relative to warfarin)
0.97
0.78–1.20†
Lognormal
Ruff et al., 2014 (4)
Annual risk of minor bleeding
8.70%
7.00%–10.40%
Beta
Connelly et al., 2009;
Granger et al., 2011 (12,13)
Aspirin
RR of ischemic stroke (relative to no therapy)
0.78
0.61–0.98†
Lognormal
Hart et al., 2007 (10)
RR of hemorrhagic stroke (relative to warfarin)
0.50
0.25–1.00†
Lognormal
Hart et al., 2007 (10)
RR of extracranial hemorrhage (relative to warfarin)
0.30
0.30–7.10
Lognormal
Hart et al., 2007 (10)
Annual risk of systemic embolism
0.40%
0.30%–0.50%
Beta
RR of myocardial infarction (relative to warfarin)
2.00
1.60–2.40
Lognormal
Annual risk of minor bleeding
1.40%
1.00%–1.70%
Beta
1.00
0.80–1.20
Lognormal
RR of major bleeding (relative to warfarin)
0.50
0.20–0.69†
Lognormal
SPINAF, CAFA, BAAFT (16–18)
RR of minor bleeding (relative to warfarin)
0.62
0.36–1.05†
Lognormal
SPINAF, CAFA, BAAFT (16–18)
ACTIVE A, 2009 (14)
ACTIVE A, 2009;
ACTIVE W, 2006 (14,15)
ACTIVE A, 2009 (14)
No therapy
RR of ischemic stroke (relative to CHA2DS2-VASc score)
Assumption on the basis
of no treatment
RR of systemic embolism (relative to warfarin)
1.96
1.60–2.40
Lognormal
CAFA (18)
RR of myocardial infarction (relative to warfarin)
1.00
0.80–1.20
Lognormal
SPAF 1 (19)
Values are relative risks or risk percentages. *RR of post-procedural ischemic stroke (relative to warfarin) uses a range of 20% as the published 95% CI includes risk of
intraprocedure stroke. †Denotes 95% CI.
BAAFT ¼ Boston Area Anticoagulation Trial for Atrial Fibrillation; CAFA ¼ Canadian Atrial Fibrillation Anticoagulation; CHA2DS2-VASc ¼ congestive heart failure, hypertension, age 75 years or older, diabetes mellitus, stroke–vascular disease; CI ¼ confidence interval; HAS-BLED ¼ hypertension, abnormal renal/liver function, stroke, bleeding
history or predisposition, labile international normalized ratio, elderly (>65 years), drugs/alcohol concomitantly; LAAC ¼ left atrial appendage closure; NOAC ¼ nonwarfarin
oral anticoagulant; RR ¼ relative risk; SPAF 1 ¼ Stroke Prevention in Atrial Fibrillation; SPINAF ¼ Stroke Prevention in Nonrheumatic Atrial Fibrillation.
Probabilities of death following systemic embo-
age 75 years or older, diabetes mellitus, stroke–
lism, extracranial hemorrhage, or myocardial infarc-
vascular disease) scores, and risk of bleeding was
tion were derived from Healthcare Utilization Project
derived from HAS-BLED (hypertension, abnormal
mortality rates for these diagnoses (23). Risk of death
renal/liver function, stroke, bleeding history or pre-
from unrelated causes was obtained from U.S. life
disposition,
tables, with disabled patients facing a 2.3-fold greater
elderly [>65 years], drugs/alcohol concomitantly)
risk of death (24,25).
scores (11,26). Because neither measure was prospec-
Baseline risk of stroke was assigned according to
CHA2DS2-VASc (congestive heart failure, hypertension,
labile
international
normalized
ratio,
tively collected in the PROTECT AF or NOAC trials,
both
were
estimated
using
available
baseline
2731
2732
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
HEALTH STATE UTILITIES AND STROKE OUTCOMES.
T A B L E 2 Stroke Outcomes and Health State Utilities
As is conventional in cost-effectiveness analyses, QoL
Utility
Value
LAAC
Warfarin (Ref. #)
NOAC
Nondisabling stroke
(MRS 0–2)
78.9
24.0 (16,17,19,20)
Moderately disabling
stroke (MRS 3)
5.3
29.0 (16,17,19,20)
21.4 (16,17,19,20)
0.390 (33)
Severely disabling stroke
(MRS 4–5)
5.3
35.0 (16,17,19,20)
25.8 (16,17,19,20)
0.110 (33)
10.5
12.0 (16,17,19,20)
8.8 (16,17,19,20)
0.000
44.0 (12,13)
0.760 (33)
was captured in the model as health utility. Health
utility values were on a scale of 0 to 1, with 1 representing perfect health and 0 representing death.
Utility values for well with warfarin (0.987), well with
NOAC (0.994), and well with aspirin (0.998) are
Fatal stroke (MRS 6)
consistent with values used in other AF stroke
reduction analyses (28–32). The utility value for well
with LAAC (0.999) was calculated by applying the
Values are percentages or utility values.
Nichol ordinary least squares (OLS) algorithm to Short
MRS ¼ modified Rankin score; other abbreviations as in Table 1.
Form-12 (SF-12) data collected during PROTECT AF
(33,34). The utility weights for all well-based health
characteristic data from the trials. Patients were
assumed to be 70 years of age, with a mean CHA2DS2VASc score of 3.2 (annual stroke risk 3.4%), and a
HAS-BLED score of 2 (annual bleeding risk 1.88%)
(11,26). To account for increasing risk with age, rates
of ischemic and hemorrhagic stroke were increased by
1.4- and 1.97-fold per decade, respectively (27,28).
Patients experiencing a TIA or systemic embolism
were assumed to have a 2.6-fold higher risk of experiencing a second ischemic event (28).
states were applied as a multiplying factor to an
underlying baseline utility of 0.82, representing QoL
at age 70 years (29). Baseline utility was decremented
by 2% per decade to account for general decline in
QoL with advancing age (35).
Because stroke is the most debilitating consequence of AF, QoL also was assessed by stroke
severity. Stroke outcomes were assigned using the
modified Rankin score (MRS) and were characterized as nondisabling (MRS 0 to 2), moderately
disabling (MRS 3), severely disabling (MRS 4 to 5),
and fatal (MRS 6) (Table 2). Stroke outcomes for
LAAC were gathered from additional analyses of
PROTECT AF at 4 years. Warfarin stroke outcomes
T A B L E 3 Cost Inputs
Acute Events
Costs
Code (Ref. #)
LAAC procedure (including 2 TEEs)*
$16,109
DRG 273/274 (43)
Fatal ischemic stroke
$8,854
DRG 063 (38)
were estimated using a weighted average of outcomes from 4 warfarin trials (16,17,19,20). NOAC
stroke
outcomes
were
derived
from
2
of
the
Severe ischemic stroke
$48,539
DRG 061/CMG 108–110 (38,39)
4 pivotal trials (12,13). Because only the rate of
Moderate ischemic stroke
$33,235
DRG 062/CMG 101–104 (38,39)
nondisabling strokes was reported, the inverse
Minor ischemic stroke
$23,236
DRG 063/CMG 105–107 (38,39)
represented disabling and fatal strokes, with the
Transient ischemic attack
$4,097
DRG 069 (38)
distribution of moderately disabling, severely dis-
Systemic embolism (nonfatal)
$4,924
DRG 068 (38)
abling, and fatal strokes assumed to be the same as
Systemic embolism (fatal)
$8,520
DRG 067 (38)
Fatal hemorrhagic stroke
$10,194
DRG 064 (38)
with warfarin.
Severe hemorrhagic stroke
$42,562
DRG 064/CMG 108–110 (38,39)
Additionally, a series of disutilities was applied in
the model for acute clinical events, representing a
Moderate hemorrhagic stroke
$28,595
DRG 065/CMG 101–104 (38,39)
Minor hemorrhagic stroke
$18,797
DRG 066/CMG 105–107 (38,39)
1-time decrement to QoL experienced for a finite
Major extracranial hemorrhage (nonfatal)
$5,877
DRG 377 (38)
length of time. Utility decrements were assessed for
DRG 378 (38)
stroke (0.139), extracranial hemorrhage (0.181),
CPT 42970 (44)
TIA (0.103), systemic embolism (0.120), and
Major extracranial hemorrhage (fatal)
Minor bleeding
$10,425
$427
Myocardial infarction (nonfatal)
$6,862
DRG 280, 281, 282 (38)
myocardial
Myocardial infarction (fatal)
$5,771
DRG 283, 284, 285 (38)
of 0.0315 was used for the LAAC procedure itself,
Quarterly costs
Warfarin þ INR monitoring
NOAC
$91
$945
CPT 85610, 99211 (44,45)
(45)
infarction
(0.125)
(36).
A
value
which assumed a 2-week disruption to healthy life.
QALYs were calculated by multiplying the health
CPT 99214 (44)
state utility value of each health state by the mean
Moderately disabled post-stroke
$9,293
(40–42)
time spent in the health state. Future QALYs were
Severely disabled post-stroke
$15,131
(40–42)
discounted at an annual rate of 3% (37).
Independent post-stroke
$108
*Costs for the LAAC procedure reflect 2016 Centers for Medicare & Medicaid Services reimbursement rates.
Weighting was obtained from 2012 U.S. hospital data (Healthcare Utilization Project [HCUP]) and reflects an
18%/82% split across DRG 250 and 251 (24).
COSTS. The model incorporated all direct health
CPT ¼ Current Procedural Terminology; DRG ¼ diagnosis-related group; INR ¼ international normalized ratio;
TEE ¼ transesophageal echocardiogram; other abbreviations as in Table 1.
ciated acute events, as well as costs for long-term
care costs for the therapies and treatment of assocare following a disabling stroke (Table 3). Costs
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
for
acute
events
2015
gained more QALYs than did warfarin-treated patients
average
(0.015). LAAC became cost-effective relative to
values, and costs for post-stroke inpatient rehabili-
warfarin by year 7, with a cost per QALY gained of
tation were from 2015 case-mix group (CMG) reim-
$42,994. By year 10, LAAC was dominant (more effec-
bursement rates (38,39). Long-term stroke disability
tive and less costly) over warfarin and remained so
costs were obtained from previous publications
throughout the remainder of the 20-year time horizon.
(30–32,40–42).
Over
diagnosis-related
LAAC
were
group
procedure
costs
taken
(DRG)
from
U.S.
national
were
calculated
as
a
lifetime,
LAAC
provided
an
additional
a
0.506 life-years and 0.638 QALYs relative to warfarin.
weighted average of the 2 newly created DRGs for
Online Figure 1 shows the total QALYs for all 3 treat-
percutaneous intracardiac procedures (273 and 274),
ment groups over 20 years.
plus the cost of 2 follow-up transesophageal echo-
NOACs
cardiograms (43,44). The annual cost of warfarin
effective than warfarin in year 1 and achieved
therapy was also applied for LAAC-treated patients
cost-effectiveness relative to warfarin at year 16
who were unable to discontinue warfarin. Warfarin
($48,446/QALY). NOACs were not cost saving relative
costs were from U.S. pharmaceutical wholesale
to warfarin over 20 years, although the ICER
acquisition data, in combination with reimbursement
continued to decrease over time, such that the cost
rates for Current Procedural Terminology (CPT) codes
per QALY at 20 years was $40,602.
related to international normalized ratio monitoring
(44,45). NOAC costs were calculated as an average of
U.S. wholesale acquisition costs for the first 3
approved drugs: dabigatran, rivaroxaban, and apixaban (45). All costs, with the exception of LAAC procedure costs, are in 2015 U.S. dollars and are
discounted at an annual rate of 3%.
VERSUS
WARFARIN. NOACs
were
more
LAAC VERSUS NOACs. When comparing LAAC with
NOACs, LAAC was expectedly more expensive than
NOACs in the first post-procedure year. However, by
year 5, LAAC was less expensive ($20,892 vs. $20,924)
and more effective (3.455 vs. 3.448 QALYs) than
NOACs.
LAAC remained
dominant over
NOACs
throughout the modeled time horizon. Over a pa-
SENSITIVITY ANALYSIS. To assess the impact of
tient’s lifetime, LAAC was estimated to provide an
parameter uncertainty on model results, 1-way
additional 0.298 life-years and 0.349 QALYs relative
sensitivity
sensitivity
to NOACs.
analysis (PSA) were undertaken. One-way sensi-
ONE-WAY
tivity analysis reveals which inputs have the great-
grams illustrating the 10 most impactful variables in
est impact on results, and PSA generates thousands
descending order of influence at 20 years are depicted
analysis
and
2733
Time to Cost-Effectiveness Following AF Treatment
probabilistic
SENSITIVITY
ANALYSIS. Tornado
dia-
of outcomes by using varied inputs to allow estimation of the effect of individual parameter uncertainty on uncertainty around model results (46).
T A B L E 4 QALY and Cost Results at 5, 10, 15, and 20 Years
Inputs were varied within 95% confidence intervals
Total
QALYs
Incremental
QALYs
(Relative to
Warfarin)
LAAC
3.455
Warfarin
3.387
NOAC
Total
Costs
Incremental
Costs
(Relative to
Warfarin)
Incremental
Cost per QALY
Versus Warfarin
Incremental
Cost per QALY
Versus NOAC
0.068
$20,892
$10,128
$149,468
Dominant
—
$10,764
—
—
—
3.448
0.061
$20,924
$10,160
$167,446
—
(CIs), where available, and by 20% where CIs were
not published. Additionally, the impact of LAAC
procedure cost was explored independently, with
100% variation. Table 1 shows ranges and distributions used for clinical inputs. Stroke outcomes
assumed a Dirichlet distribution, and health state
utilities assumed a beta distribution. All costs were
5 yrs
10 yrs
varied by 20% and assumed a gamma distribution.
LAAC
5.855
0.254
$25,425
-$1,409
Dominant
Dominant
The PSA followed a standard Monte Carlo approach
Warfarin
5.601
—
$26,834
—
—
—
on the basis of 5,000 randomly drawn simulations
NOAC
5.751
0.150
$39,260
$12,426
$82,684
—
LAAC
7.309
0.466
$29,075
-$12,251
Dominant
Dominant
Warfarin
6.843
—
$41,326
—
—
—
NOAC
7.077
0.234
$53,431
$12,105
$51,755
of parameter values.
RESULTS
15 yrs
—
20 yrs
LAAC VERSUS WARFARIN. As expected, LAAC was
LAAC
8.031
0.638
$31,198
-$18,748
Dominant
Dominant
more expensive than warfarin in the first post-
Warfarin
7.392
—
$49,946
—
—
—
procedure year, and patients had fewer QALYs because
NOAC
7.682
0.290
$61,701
$40,602
—
of the disutility applied for the procedure (Table 4,
Central Illustration). By year 3, LAAC-treated patients
$11,755
QALY ¼ quality-adjusted life-year; other abbreviations as in Table 1.
2734
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
CE NTR AL ILL UST RAT ION
Warfarin Versus NOACs Versus LAAC: Cumulative Cost and Time to Cost-Effectiveness Following
Treatment Initiation
$70,000
LAAC
$60,000
Warfarin
NOAC
$50,000
$40,000
$30,000
$20,000
$10,000
$1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Years
Time to Clinical Effectiveness
Time to Cost-Effectiveness
(Incremental QALYs)
QALYS)
(Cost
(Cost per
per QALY)
QALY)
Time to Dominance
(MoreEffective,
Effective,Less
LessCostly)
Costly)
(More
LAAC vs. warfarin
Year 3
(0.015)
Year 7
($42,994/QALY)
NOACs vs. warfarin
Year 1
(0.008)
Year 16
($48,446/QALY)
N/A
LAAC vs. NOACs
Year 5
(0.007)
Year 5
(Dominant)
Year 5
Year 10
Reddy, V.Y. et al. J Am Coll Cardiol. 2015; 66(24):2728–39.
(Upper panel) Average total cumulative costs per patient over 20 years. Left atrial appendage closure (LAAC) becomes less costly than nonwarfarin oral
anticoagulants (NOACs) at 5 years and less costly than warfarin at 10 years. (Lower panel) Summary of time to clinical and cost-effectiveness for each
treatment arm. LAAC becomes cost-effective at 7 years relative to warfarin and at 5 years relative to NOACs. NOACs achieve cost-effectiveness relative to
warfarin at 16 years. QALY ¼ quality-adjusted life-year.
in Figure 2 (Online Figures 2A to 2C for 10-year re-
threshold. In the NOAC versus warfarin analysis
sults). One-way sensitivity analyses of LAAC versus
(Figure 2C), the baseline risk of bleeding had the
warfarin demonstrated that the health utility values
greatest impact on modeled results, followed by the
used for well with LAAC had a significant impact on
utility value for well with NOAC, baseline risk of
model results, with this being the only variable for
stroke, and the proportion of NOAC ischemic strokes
which the upper threshold increased the ICER at
that were severely disabling.
20 years beyond $50,000/QALY (Figure 2A). The
Additional analyses demonstrated that, even with
baseline risk of ischemic stroke also influenced the
a doubling of LAAC procedure cost, LAAC remained
ICER, with higher risks resulting in more favorable
cost-effective. As expected, time to cost-effectiveness
cost-effectiveness for LAAC. All other variables had
increased, with LAAC becoming cost-effective at 11
a <1% impact on model outcomes. When comparing
years relative to warfarin ($41,470/QALY) and at 10
LAAC with NOACs (Figure 2B), the 20-year results
years relative to NOACs ($21,964/QALY); however,
were most sensitive to the utility values for well with
LAAC remained dominant over both drug compara-
LAAC and well with NOACs along with the baseline
tors over the lifetime analysis.
risk of bleeding. However, only varying well with
PROBABILISTIC SENSITIVITY ANALYSIS. PSA simu-
LAAC increased the ICER beyond the $50,000
lations (Figure 3) demonstrated that LAAC had
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
F I G U R E 2 Tornado Plots of 1-Way Sensitivity Analyses at 20 Years of LAAC Versus Warfarin, LAAC Versus NOACs, and NOACs
Versus Warfarin
A
Well with LAAC
Baseline risk of stroke
0.9
18.00%
Baseline risk of bleeding
2.00%
1.00%
8.70%
Well with Warfarin
1
0.9
Quarterly Cost of Severe Post Stroke
$18,157
$12,105
Cost of LAAC Procedure
$12,887
$19,331
Warfarin Relative risk of ischemic stroke
0.46
LAAC % of IS Independent
63.10%
Warfarin % HS Severe
42%
28%
Warfarin % HS Severe
42%
28%
$(80,000)
0.23
90.00%
$(30,000)
$20,000
$70,000
$120,000
$170,000
$220,000
LAAC vs. WARFARIN (20 years)
B
0.9
Well with LAAC
1
Well with NOAC
LAAC % of IS Independent
NOAC % of IS Severe
0.9
1.00%
Baseline risk of bleeding
8.70%
90.00%
63.10%
52.80%
LAAC Relative risk of hemorrhagic stroke
25.80%
0.49
0.03
Well with aspirin
1
Cost of LAAC Procedure
$12,887
LAAC Relative risk of ischemic stroke
1.14
0.9
$19.331
0.76
17%
NOAC % of IS Independent
$(130,000)
$(80,000)
$(30,000)
$20,000
$70,000
LAAC v. NOAC (20 years)
C
Baseline risk of bleeding
Well with NOAC
1.00%
8.70%
0.9
Baseline risk of stroke
2.00%
18.00%
NOAC % of IS Severe
52.80%
Well with Warfarin
0.9
1
Warfarin % of hemorrhage ICH
50.20%
Warfarin Relative risk of hemorrhage
1.20
Warfarin % HS Severe
42%
Quarterly Cost of Severe Post Stroke
$18,157
Warfarin % of IS Severe
42%
$(50,000)
$(30,000)
$(10,000)
$10,000
33.40%
0.80
28%
$12,105
28%
$30,000
$50,000
$70,000
NOAC vs. WARFARIN (20 years)
(A) LAAC versus warfarin. (B) LAAC versus nonwarfarin oral anticoagulants (NOACs). (C) NOACs versus warfarin. The 10 most impactful variables are presented in descending order of influence. All were varied, in standard fashion, as published 95% confidence intervals or 20%.
HS ¼ hemorrhagic stroke; ICH ¼ intracranial hemorrhage; IS ¼ ischemic stroke.
2735
2736
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
F I G U R E 3 Scatter Plots of Incremental Costs and Incremental QALYs at 20 Years for LAAC Versus Warfarin, LAAC Versus NOACs, and NOACs
Versus Warfarin
A
Less Effective & More Expensive
B
More Effective & More Expensive
Less Effective & More Expensive
$70,000
-2.0
-1.5
-1.0
$50,000
$50,000
$30,000
$30,000
-0.5
0.0
-$10,000
Less Effective & Less Expensive
$10,000
0.5
1.0
1.5
2.0
2.5
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
-$10,000
-$30,000
-$30,000
-$50,000
-$50,000
-$70,000
More Effective & More Expensive
$70,000
$10,000
-2.5
$90,000
More Effective & Less Expensive
Less Effective & Less Expensive -$70,000
0.5
1.0
1.5
2.0
2.5
More Effective & Less Expensive
-$90,000
-$90,000
Incrermental QALYs - LAAC vs warfarin (20 years)
Incrermental QALYs - LAAC vs NOACs (20 years)
C
$90,000
More Effective & More Expensive
Less Effective & More Expensive
$70,000
$50,000
$30,000
$10,000
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
-$10,000
0.5
1.0
1.5
2.0
2.5
-$30,000
-$50,000
More Effective & Less Expensive
Less Effective & Less Expensive -$70,000
-$90,000
Incrermental QALYs - NOACs vs warfarin (20 years)
Probabilistic sensitivity analysis results reflect 5,000 model simulations to estimate the effect of uncertainty on model results. QALYs ¼ quality-adjusted
life-years; other abbreviations as in Figures 1 and 2.
lower average total costs than warfarin and NOACs
warfarin. Similar outcomes, albeit with greater scat-
at 20 years: $31,127 (95% CI: $17,350 to $51,067)
ter, are seen when the PSAs were conducted at 10
versus $48,817 (95% CI: $28,719 to $77,484) versus
years (Online Figures 3A to 3C).
$58,254 (95% CI: $44,865 to $77,763). Over 20 years,
there was a 94% probability that LAAC was cost
DISCUSSION
saving relative to warfarin, a 97% probability that
LAAC provided more QALYs, and a 100% probability
This analysis demonstrates that both novel thera-
that it provided more life-years. The overall proba-
pies, NOACs and LAAC with the Watchman device,
bility
a
are cost-effective relative to warfarin for stroke risk
willingness-to-pay threshold of $50,000/QALY. At
reduction in patients with nonvalvular AF. When
20 years, there was a 95% probability that LAAC
compared over a lifetime, LAAC proved to be the
was cost-effective relative to NOACs and a 75%
most cost-effective treatment. LAAC becomes cost-
probability that NOACs were cost-effective relative to
effective relative to warfarin at 7 years, with an
of
cost-effectiveness
was
98%,
using
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
ICER of $42,994. In contrast, NOACs require twice
anticoagulation therapy, and 46% of these patients
as long to achieve cost-effectiveness relative to
had a gap in anticoagulation therapy during the 2
warfarin, with an ICER of $48,446. LAAC was less
years analyzed (48). Additionally, 63% of all stroke-
costly than warfarin at 10 years and NOACs at 5
related hospitalizations occurred in patients who
years, with cost savings generated annually there-
were not receiving oral anticoagulation. With nearly
after. NOACs are more effective than warfarin, but
$8 billion spent annually on AF-associated strokes (3),
they remain more expensive throughout the life-
improving adherence is crucial to reducing this cost
time horizon.
burden.
Sensitivity
analyses
suggest
that
these
cost-
By its very nature, LAAC is not subject to patient
effectiveness results are robust and not overly
adherence issues. Once implanted, the device pro-
sensitive to variation in individual parameters, with
vides lifelong stroke prophylaxis without the risk of
the exception of well with LAAC, well with NOACs,
complications associated with oral anticoagulants.
and baseline risk of bleeding. Specifically, 1-way
Although the procedure has risks, they are reduced
sensitivity analyses found that LAAC is more cost-
with operator experience. In recognition of the clin-
effective in patients at higher risk of stroke and
ical risk/benefit relationship, the 2014 American
bleeding. Given the importance of baseline utility
Heart
values in this analysis, future prospective studies of
updated guidelines include consideration of LAAC in
these novel therapies should evaluate QoL over
patients at high risk for stroke who are “unsuitable”
time. PSA indicates a high probability that LAAC is
for long-term anticoagulation, if they can tolerate the
cost saving relative to warfarin and NOACs over
risk of at least 45 days of post-procedural anti-
20 years.
coagulation (49).
Association/American
Stroke
Association
Although costs of any new health intervention
To the best of our knowledge, the only other study
must be considered, patient outcomes remain of
assessing the cost-effectiveness of LAAC in the
paramount importance. Cost-effectiveness modeling
United States was published in abstract form. In this
uses 1 metric to account for cost, clinical effective-
study, LAAC dominated all NOACs, except for the
ness, and patient outcomes. For stroke prevention,
110-mg dose of dabigatran, and was cost-effective,
the risk and cost of stroke are highly important, but so
but not cost saving relative to warfarin (50). This
too are QoL and functional outcomes following
difference from our analysis is likely explained by our
stroke. Indeed, a preference study of stroke outcomes
detailed patient-level accounting of stroke severity
revealed that most patients rate severe disability as
and QoL. A cost-effectiveness analysis for the Cana-
worse than death (47).
dian health system compared LAAC with Watchman,
LAAC-treated
patients
fewer
dabigatran, and warfarin (51). Despite overall con-
disabling strokes than did patients taking warfarin
clusions directionally consistent with our analysis
or NOACs: 79% of strokes in patients who had un-
(particularly
dergone LAAC resulted in an MRS score of 0 to 2
effectiveness relative to warfarin and LAAC domi-
versus 24% of strokes in patients taking warfarin
nance over dabigatran), the earlier analysis had
and 44% of strokes in patients taking NOACs. This
several limitations. First, only 1.5-year follow-up
finding suggests that most LAAC-treated patients
PROTECT
can return to daily life without assistance following
follow-up data in our present analysis. Second,
stroke, whereas more than one-half the patients
whereas only dabigatran outcomes were available at
who have a stroke while taking oral anticoagulants
that time, the present analysis additionally in-
require lifetime assistance. Furthermore, previously
corporates outcomes with rivaroxaban, apixaban, and
published SF-12 data collected during PROTECT AF
edoxaban.
found an overall improvement in QoL and physical
examined only a lifetime horizon, the present anal-
functioning in LAAC-treated patients relative to
ysis considers cost-effectiveness at multiple time
warfarin-treated
points. Finally, the present analysis uses a U.S. payer
patients
experienced
at
12
months
(34).
Disability following a stroke has long-term implica-
both
AF
LAAC
data
Third,
were
whereas
and
used,
that
dabigatran
versus
earlier
cost-
3.8-year
analysis
perspective.
tions for both patients and health care costs, thus
warranting further study.
STUDY LIMITATIONS. Our model allowed for only 1
NOACs offer significant advantages over warfarin,
clinical event per 3-month cycle, whereas multiple
but the problem of therapy adherence persists.
events may occur. Data were taken from multiple
Analysis of Medicare MarketScan data found that only
sources, including meta-analyses and randomized
38% to 40% of patients with AF who had a moderate
controlled trials. Individual trials had different time
to high risk of stroke received a prescription for
horizons, and all data were extrapolated to 20 years.
2737
2738
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
Time to Cost-Effectiveness Following AF Treatment
By necessity, an indirect comparison method was
used, with warfarin as the common control value.
REPRINT REQUESTS AND CORRESPONDENCE: Dr.
Although this methodology is well established in
Vivek Y. Reddy, Helmsley Trust, Electrophysiology
health economic analyses (52), these data are not
Center, Icahn School of Medicine at Mount Sinai,
derived from a direct, randomized comparison. How-
One Gustave L. Levy Place, Box 1030, New York, New
ever, it will likely be many years before an LAAC
York 10029. E-mail: [email protected].
versus NOAC randomized trial is performed. Additionally, the datasets from which probabilities were
PERSPECTIVES
drawn were all on the basis of intention-to-treat analyses, whereas the model allowed for the possibility
of therapy change. Finally, treatments administered
in
clinical
practice
may
vary
in
effectiveness
COMPETENCY IN SYSTEMS-BASED PRACTICE:
As an alternative to long-term warfarin therapy for
stroke prophylaxis in patients with AF who are poor
compared with that observed in randomized trials,
candidates for long-term anticoagulation, percuta-
which generally enroll selected patients, achieve
neous LAAC become cost-effective at 7 years. On the
higher levels of adherence, and monitor patients more
basis of current cost estimates, LAAC would be asso-
intensively.
ciated with more QALYs gained by 10 years than
warfarin or NOACs.
CONCLUSIONS
Both
NOACs
and
LAAC
with
the
Watchman
device have the potential to change the treatment
paradigm for patients with nonvalvular AF who
are at risk of stroke. Both novel therapies demonstrated
cost-effectiveness
relative
to
warfarin,
but only LAAC demonstrated cost savings by year 10
TRANSLATIONAL OUTLOOK: Longer-term clinical
studies are needed to define more clearly the return
on investment in nonpharmacologic strategies to
prevent stroke in patients with AF and incorporate the
findings into clinical practice guidelines and reimbursement policies.
relative to warfarin and by year 5 relative to NOACs.
REFERENCES
1. Go AS, Mozaffarian D, Roger VL, et al. Heart
disease and stroke statistics—2013 update: a
8. Harrington AR, Armstrong EP, Nolan PE Jr.,
et al. Cost-effectiveness of apixaban, dabigatran,
anticoagulation for atrial fibrillation in the Atrial
fibrillation Clopidogrel Trial with Irbesartan for
report from the American Heart Association.
Circulation 2013;127:e6–245.
rivaroxaban, and warfarin for stroke prevention in
atrial fibrillation. Stroke 2013;44:1676–81.
2. Lee WC, Lamas GA, Balu S, et al. Direct treatment cost of atrial fibrillation in the elderly
American population: a Medicare perspective.
J Med Econ 2008;11:281–98.
9. You JHS. Novel oral anticoagulants versus
warfarin therapy at various levels of anticoagulation control in atrial fibrillation: a costeffectiveness analysis. J Gen Intern Med 2014;
29:438–46.
prevention of Vascular Events (ACTIVE W): a
randomised controlled trial. Lancet 2006;367:
1903–12.
3. Coyne KS, Paramore C, Grandy S, et al. Assessing the direct costs of treating nonvalvular atrial
fibrillation in the United States. Value Health
2006;9:348–56.
4. Ruff CT, Giugliano RP, Braunwald E, et al.
Comparison of the efficacy and safety of new oral
anticoagulants with warfarin in patients with atrial
fibrillation: a meta-analysis of randomized trials.
Lancet 2014;383:955–62.
5. Holmes DR, Reddy VY, Turi ZG, et al. Percutaneous closure of the left atrial appendage versus
warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomized noninferiority trial. Lancet 2009;374:534–42.
6. Reddy VY, Sievert H, Halperin J, et al. Percutaneous left atrial appendage closure vs warfarin
for atrial fibrillation: a randomized clinical trial.
JAMA 2014;312:1988–98.
7. Coyle D, Coyle K, Cameron C, et al. Costeffectiveness of new oral anticoagulants
compared with warfarin in preventing stroke and
other cardiovascular events in patients with atrial
fibrillation. Value Health 2013;16:498–506.
10. Hart RG, Pearce LA, Aguilar MI. Meta-analysis:
antithrombotic therapy to prevent stroke in
patients who have nonvalvular atrial fibrillation.
Ann Intern Med 2007;146:857–67.
11. Pisters R, Lane DA, Nieuwlaat R, et al. A novel
user-friendly score (HAS-BLED) to assess 1-year
risk of major bleeding in patients with atrial
fibrillation: the Euro Heart Survey. Chest 2010;
138:1093–100.
16. Ezekowitz MD, Bridgers SL, James KE, et al.
Warfarin in the prevention of stroke associated
with nonrheumatic atrial fibrillation. N Engl J Med
1992;327:1406–12.
17. Boston Area Anticoagulation Trial for Atrial
Fibrillation Investigators. The effect of low-dose
warfarin on the risk of stroke in patients with
nonrheumatic atrial fibrillation. N Engl J Med
1990;323:1505–11.
18. Connelly SJ, Laupaucis A, Gent M, et al. Canadian Atrial Fibrillation Anticoagulation (CAFA)
study. J Am Coll Cardiol 1991;18:349–55.
19. Stroke Prevention in Atrial Fibrillation in-
12. Connelly SJ, Ezekowitz MD, Yusuf S, et al.
Dabigatran versus warfarin in patients with atrial
fibrillation. N Engl J Med 2009;361:1139–51.
13. Granger CB, Alexander JH, McMurray JJV,
et al. Apixaban versus warfarin in patients
with atrial fibrillation. N Engl J Med 2011;365:
981–92.
14. ACTIVE Investigators. Effects of clopidogrel
added to aspirin in patients with atrial fibrillation.
N Engl J Med 2009;360:2066–78.
15. ACTIVE Writing Group of the ACTIVE Investigators. Clopidogrel plus aspirin versus oral
vestigators. Stroke Prevention in Atrial Fibrillation study: final results. Circulation 1991;84:
527–39.
20. Petersen P, Boysen G, Godtfredsen J, et al.
Placebo-controlled, randomized trial of warfarin
and aspirin for prevention of thromboembolic
complications in chronic atrial fibrillation: the
Copenhagen AFASAK study. Lancet 1989;8631:
175–8.
21. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial
fibrillation. N Engl J Med 2011;365:883–91.
Reddy et al.
JACC VOL. 66, NO. 24, 2015
DECEMBER 22, 2015:2728–39
22. Giugliano RP, Ruff CT, Braunwald E, et al.
Edoxaban versus warfarin in patients with atrial
fibrillation. N Engl J Med 2013;369:2093–104.
23. HCUP Databases. Healthcare Cost and Utilization Project (HCUP). Overview of the National
(Nationwide) Inpatient Sample. Rockville, MD:
Time to Cost-Effectiveness Following AF Treatment
(Percutaneous Closure of the Left Atrial
Appendage Versus Warfarin Therapy for Prevention of Stroke in Patients with Atrial Fibrillation)
trial of patients at risk for stroke with nonvalvular
atrial fibrillation. J Am Coll Cardiol 2013;61:
1790–8.
Agency for Heathcare Research and Quality,
May 2015. Available at: www.hcup-us.ahrq.gov/
nisoverview.jsp. Accessed October 4, 2015.
35. Fryback DG, Dunham NC, Palta M, et al. U.S.
norms for six generic health-related quality of life
indexes from the National Health Measurement
24. Social Security Administration. Actuarial Life
Table 2011. Available at: http://www.ssa.gov/
oact/STATS/table4c6.html. Accessed October 4,
2015.
study. Med Care 2007;45:1162–70.
25. Dennis MS, Burn JPS, Sandercock PAG, et al.
Long-term survival after first-ever stroke: the
Oxfordshire Community Stroke Project. Stroke
high risk of stroke in the US. Pharmacoeconomics
2006;24:1021–33.
1993;24:796–800.
26. Lip GYH, Nieuwlaat R, Pisters R, et al. Refining
clinical risk stratification for predicting stroke and
thromboembolism in atrial fibrillation using a
novel risk factor–based approach: the Euro Heart
Survey on Atrial Fibrillation. Chest 2010;137:
263–72.
27. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: analysis of
pooled data from five randomized controlled trials. Arch Intern Med 1994;154:1449–57.
28. Ariesen MJ, Claus SP, Rinkel GJE, et al. Risk
factors for intracerebral hemorrhage in the general population. Stroke 2003;34:2060–5.
36. Sullivan PW, Arant TW, Ellis SL, et al. The cost
effectiveness of anticoagulation management
services for patients with atrial fibrillation and at
37. Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the Panel on Cost-Effectiveness
in Health and Medicine. JAMA 1996;276:1253–8.
38. Centers for Medicare & Medicaid Services.
Acute Inpatient PPS FY 2015 Final Rule Tables.
2014. Available at: https://www.cms.gov/Medicare/
Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/
FY2015-IPPS-Final-Rule-Home-Page-Items/FY2015Final-Rule-Tables.html. Accessed October 4, 2015.
39. Centers for Medicare & Medicaid Services.
Inpatient Rehabilitation Facility PPS Data Files.
2014. Available at: https://www.cms.gov/Medicare/
Medicare-Fee-for-Service-Payment/InpatientRehab
FacPPS/Data-Files.html. Accessed October 4, 2015.
40. Cipriano LE, Steinberg ML, Gazelle GS, et al.
29. O’Brien CL, Gage BF. Costs and effectiveness
of ximelagatran for stroke prophylaxis in chronic
atrial fibrillation. JAMA 2005;293:699–706.
30. Shah SV, Gage BF. Cost-effectiveness of
dabigatran for stroke prophylaxis in atrial fibrillation. Circulation 2011;123:2562–70.
31. Gage BF, Cardinalli AB, Albers GW, et al. Costeffectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial
fibrillation. JAMA 1995;274:1839–45.
32. Lee S, Anglade MW, Pham D, et al. Costeffectiveness of rivaroxaban compared to warfarin
for stroke prevention in atrial fibrillation. Am J
Cardiol 2012;110:845–51.
33. Nichol MB, Sengupta N, Globe DR. Evaluating
quality-adjusted life years: estimation of the
health utility index (HU12) from the SF-36. Med
Decis Making 2001;21:105–12.
34. Alli O, Doshi S, Kar S, et al. Quality of life
assessment in the randomized PROTECT AF
Comparing and predicting the costs and outcomes
of patients with major and minor stroke using the
Boston Acute Stroke Imaging Scale neuroimaging
classification system. AJNR Am J Neuroradiol
2009;30:703–9.
41. Mercaldi CJ, Sui K, Sander SD, et al. Long-term
costs of ischemic stroke and major bleeding events
among Medicare patients with nonvalvular atrial
fibrillation. Cardiol Res Pract 2012;2012:645469.
42. Caro JJ, Huybrechts KF. Stroke treatment
economic model (STEM): predicting long-term
costs from functional status. Stroke 1999;30:
2574–9.
43. Centers for Medicare & Medicaid Services.
Acute IPPS FY 2016 Final Rule and Correction Notice
Tables. 2015. Available at: https://www.cms.
gov/Medicare/Medicare-Fee-for-Service-Payment/
AcuteInpatientPPS/FY2016-IPPS-Final-Rule-HomePage-Items/FY2016-IPPS-Final-Rule-Tables.html.
Accessed October 4, 2015.
44. The Coding Institute. Available at: http://
www.codinginstitute.com/supercoder. Accessed
October 5, 2015.
45. DMD
America.
AnalysourceBiweekly.com.
Available at: https://www.analysource.com/pricing.
html. Accessed October 5, 2015.
46. Briggs AH, Weinstein MC, Fenwick EAL, et al.
Model parameter estimation and uncertainty: a
report of the ISPOR-SMDM Modeling Good
Research Practices Task Force-6. Value Health
2012;15:835–42.
47. Solomon NA, Glick HA, Russo CJ, et al. Patient
preferences for stroke outcomes. Stroke 1994;25:
1721–5.
48. Lang K, Bozkaya D, Patel AA, et al. Anticoagulant use for the prevention of stroke in
patients with atrial fibrillation: findings from a
multi-payer analysis. BMC Health Serv Res 2014;
14:329.
49. Meschia JF, Bushnell C, Boden-Albala B, et al.
Guidelines for the primary prevention of stroke: a
statement for healthcare professionals from the
American Heart Association/American Stroke Association. Stroke 2014;45:3754–832.
50. Lee VW, Yan B, Chow I, et al. Cost-effectiveness of left atrial appendage occlusion device
compared with 7 oral anticoagulants for stroke
prevention in atrial fibrillation (abstr). J Am Coll
Cardiol 2014;64.
51. Singh SM, Micieli A, Wijeysundera HC. An
economic evaluation of percutaneous left atrial
appendage occlusion: dabigatran and warfarin
for stroke prevention in patients with nonvalvular atrial fibrillation. Circulation 2013;127:
2414–23.
52. Jansen JP, Fleurence R, Devine B, et al.
Interpreting indirect treatment comparisons
and network meta-analysis for health-care decision making: report of the ISPOR Task Force
on Indirect Treatment Comparisons Good Research Practices: part 1. Value Health 2011;14:
417–28.
KEY WORDS apixaban, dabigatran,
edoxaban, quality of life, rivaroxaban,
Watchman
A PP END IX For supplemental figures, please
see the online version of this article.
2739