Journal of Thrombosis and Haemostasis, 8: 14681474

DOI: 10.1111/j.1538-7836.2010.03856.x

ORIGINAL ARTICLE

Antiplatelet drugs and risk of subarachnoid hemorrhage:

a population-based case-control study
M. SCHMIDT,* M. B. JOHANSEN,* T. L. LASH,* C. F. CHRISTIANSEN,* S. CHRISTENSEN*
and H . T . S R E N S E N *
*Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; and Department of Epidemiology, Boston University
School of Public Health, Boston, MA, USA

To cite this article: Schmidt M, Johansen MB, Lash TL, Christiansen CF, Christensen S, Srensen HT. Antiplatelet drugs and risk of subarachnoid
hemorrhage: a population-based case-control study. J Thromb Haemost 2010; 8: 146874.

Summary. Background: Antiplatelet drug use increases bleeding risk, but its role in precipitating subarachnoid hemorrhage
remains unclear. Objectives: We examined whether the use of
low-dose acetylsalicylic acid (LDA), clopidogrel or dipyridamole increased the risk of subarachnoid hemorrhage. Patients/
Methods: This population-based casecontrol study was conducted in northern Denmark. We used the Danish National
Patient Registry to identify all persons admitted to neurosurgery or neurology departments with a rst diagnosis of
subarachnoid hemorrhage between 1997 and 2008
(n = 1186). Using risk-set sampling, we selected 10 population
controls (n = 11 840) for each case, matched by age and sex.
We obtained data on prescriptions for antiplatelet drugs, use of
other medications and comorbidity from medical databases.
We used conditional logistic regression to compute odds ratios
with 95% condence intervals (CIs), controlling for confounding factors. Results: One hundred and nine cases (9.2%) and
910 controls (7.7%) used antiplatelet drugs. Among cases, 104
(8.8%) used LDA and 11 (0.9%) used dipyridamole. Among
controls, 891 (7.5%) used LDA and 48 (0.4%) used dipyridamole. As compared with not using any antiplatelet drugs during
the study period, the adjusted odds ratios were 1.03
(95% CI 0.811.32) for long-term LDA use, 2.52 (95%
CI 1.374.62) for new LDA use, and 2.09 (95% CI 1.044.23)
for long-term dipyridamole use. Owing to the low number
of users, data were inconclusive for clopidogrel. Conclusions: Long-term dipyridamole use and new LDA use were
associated with an increased risk of subarachnoid hemorrhage.
Because of the limited precision of these risk estimates, however,
caution is advised in their interpretation. Long-term LDA use
was not associated with subarachnoid hemorrhage.

Keywords: acetylsalicylic acid, antiplatelet drugs, clopidogrel,

dipyridamole, subarachnoid hemorrhage.
Introduction
Antiplatelet drug therapy is a mainstay in the prevention and
treatment of occlusive vascular events in patients with coronary
artery disease or ischemic stroke [1]. As compared with
acetylsalicylic acid (ASA) monotherapy, dual antiplatelet drug
therapy with ASA plus clopidogrel following acute coronary
syndromes [24], or ASA plus dipyridamole following ischemic
stroke [5], reduces by 2025% the relative risk of thromboembolic cardiovascular events (non-fatal myocardial infarction,
non-fatal stroke, or cardiovascular death). The benets of
antiplatelet drug therapy (single or dual) accrue at the expense
of increased risk of hemorrhagic complications most notably
gastrointestinal bleeding, but also intracranial bleeding [6,7].
Subarachnoid hemorrhage (SAH) is caused by a ruptured
intracranial aneurysm in 85% of all cases [8]. Although familial
clustering suggests a genetic predisposition [9], aneurysms are
not congenital but, rather, develop over the life course [10].
Known risk factors for SAH include hypertension, smoking,
and alcohol abuse [11]. Hypercholesterolemia, obesity and
diabetes may protect against SAH occurrence [11]. Half of all
SAHs occur in persons younger than 55 years [8]. Given this
relatively young age and the 50% short-term case fatality rate
[8], any association of SAHs with commonly used antiplatelet
drugs is of clinical and etiologic importance. However, no data
exist on the effect of antiplatelet drug use on SAH occurrence.
We examined whether the use of low-dose ASA, clopidogrel or
dipyridamole increased SAH risk.
Materials and methods

Correspondence: Morten Schmidt, Department of Clinical

Epidemiology, Olof Palmes Alle 43-45, DK-8200 Aarhus N, Denmark.
Tel.: +45 89424800; fax: +45 89424801.
E-mail: morten.schmidt@dce.au.dk
Received 11 November 2009, accepted 8 March 2010

Setting

We conducted this population-based casecontrol study in

northern Denmark (North Jutland County and Aarhus
County) with a population of 1.15 million (approximately
20% of the Danish population). The study period was dened
2010 International Society on Thrombosis and Haemostasis

Antiplatelet drugs and subarachnoid hemorrhage 1469

by the availability, starting in 1996, of complete computerized

prescription records. In order to provide at least 1 year of
prescription history for all members of the study population,
the study period was set as 1 January 1997 to 31 December
2008.
The Danish National Health Service provides universal taxsupported healthcare, guaranteeing unfettered access to general
practitioners and hospitals and partial reimbursement for
prescribed medications, including antiplatelet drugs. Other
than ASA, all antiplatelet drugs are available by prescription
only. In Denmark, all hospital care for patients with surgical
emergencies, including SAH, is provided by public hospitals.
We linked various registries using the unique Danish central
personal registry (CPR) number, which is assigned to each
Danish citizen at birth and to residents at immigration [12].
Cases with SAH

We used the Danish National Patient Registry (DNPR), which

covers all Danish hospitals [13], to identify all patients with a
rst in-hospital diagnosis of SAH. This registry contains data
on dates of admission and discharge, including all discharge
diagnoses from non-psychiatric hospitals after 1977 and
emergency room and outpatient clinic visits after 1995 [13].
Each discharge is associated with one primary diagnosis and
one or more secondary diagnoses classied according to the
International Classication of Diseases, 8th revision (ICD-8),
until the end of 1993, and the 10th revision (ICD-10) thereafter
[13]. More than 95% of stroke patients in Denmark receive
diagnostic computed tomography or magnetic resonance
imaging scans during hospitalization [14]. The proportion of
correctly diagnosed SAH patients, that is, the positive predictive value (PPV), recorded in the DNPR depends on the
admitting departments level of specialization [9,15]. To ensure
a high quality of SAH diagnoses, we included only patients
whose SAH diagnosis code originated from a neurosurgery
(PPV = 93%) or neurology (PPV = 75%) department [9].
We used ICD-8 code 430 and ICD-10 code I60 to identify SAH
cases. To distinguish aneurysmal rupture from traumatic SAH,
we did not include patients (n = 747) diagnosed with traumatic SAH (ICD-10 code S06.6) in the study period. Cases
coded with a traumatic SAH diagnosis at the same hospital
admission were excluded (n = 4). We considered the date of
the rst SAH diagnosis as the index date for cases.
Population controls

For each case, we selected from the Danish Civil Registration

System [16] 10 population controls with no previous SAH
hospitalization, matched on age and sex. This registry has
kept electronic records on all vital statistics including date of
birth, change of address, date of emigration, and exact date of
death for the Danish population since 2 April 1968 [16]. We
selected controls using risk-set sampling; that is, the controls
had to be alive and at risk of a rst SAH hospitalization at the
index date of the case to whom they were matched [17].
2010 International Society on Thrombosis and Haemostasis

Controls were assigned an index date identical to that of

corresponding cases.
Antiplatelet drug use

We used the population-based prescription database in the

region to identify prospectively recorded prescriptions for
antiplatelet drugs redeemed by cases and controls before their
index dates. Pharmacies in Denmark are equipped with
electronic accounting systems, which are primarily used to
secure reimbursement from the National Health Service. For
each redeemed prescription, the patients CPR number, the
type and amount of drug prescribed according to the
Anatomical Therapeutic Chemical (ATC) classication system
and the date of drug dispensation are transferred electronically
from the pharmacies to a prescription research database at
Aarhus university [18].
In Denmark, long-term antiplatelet drug therapy for
secondary prevention of cardiovascular disease is typically
recommended as follows: lifelong daily low doses of 75150 mg
of ASA (LDA), 75 mg of clopidogrel up to 12 months, and
400 mg of dipyridamole lifelong [19]. Prescriptions were
identied by ATC codes for LDA in tablet sizes of 75 and
150 mg, clopidogrel in a tablet size of 75 mg, and dipyridamole
in tablet sizes of 100 and 200 mg. In a secondary analysis, we
identied use of high-dose ASA (HDA) in a tablet size of
500 mg. In addition to ATC codes, we used drug identication
numbers to separate LDA and HDA. Furthermore, use of
LDA plus dipyridamole was identied by combining their
individual codes and using an ATC code for a combination pill
containing 25 mg of ASA and 200 mg of dipyridamole. ATC
and drug identication codes are provided in the Appendix. We
computed the number of days supplied by a prescription as the
prescribed package size divided by the recommended daily
dose.
Other characteristics of patients and controls

From the DNPR and the prescription database, we identied

risk factors for SAH [11] potentially associated with antiplatelet
drug use registered within 5 years of index dates. Specically,
we identied comorbidities as follows: (i) by using inpatient and
outpatient clinic diagnoses of hypertension, obesity and rare
causes of SAH as dened by van Gijn et al. [8] (coagulopathy,
arteritis, polycystic kidney disease, and brain tumor); and (ii) by
combining hospital diagnoses and prescription data for chronic
obstructive pulmonary disease (COPD), alcoholism-related
disease, diabetes mellitus, and ischemic vascular disease (ischemic heart disease, cerebral ischemia, or two or more prescriptions of nitroglycerin).
In addition, we identied current use of the following
medications from the prescription database: postmenopausal
hormone replacement therapy (which has been associated with
decreased risk of SAH [11]), angiotensin-converting enzyme
inhibitors, b-blockers, calcium channel blockers, glucocorticoids, and statins (because these drugs affect nitric oxide-

1470 M. Schmidt et al

mediated vascular endothelium function [2024], which may

play a role in SAH etiology [25]). We also identied current use
of non-steroidal anti-inammatory drugs (NSAIDs), selective
serotonin reuptake inhibitors (SSRIs), and vitamin K antagonists (warfarin and phenprocoumon), because these drugs
increase the risk of bleeding [2628]. ATC and ICD codes are
provided in the Appendix.

therefore compared the OR and 95% CI obtained from the

conventional conditional logistic regression model with the
median and 95% simulation interval yielded by a probabilistic
bias analysis [35]. Assuming a valid bias model, this comparison allows a quantitative assessment of the bias and
uncertainty arising from confounding by unmeasured smoking.
Analyses were performed using SAS version 9.2 (SAS Institute
Inc., Cary, NC, USA).

Statistical analysis

We dened current users of a medication as those whose

index date fell into the interval between the date on which a
medication was dispensed and this date plus the computed
number of days supplied plus 7 days. Seven days were added
to the number of days supplied to allow for a 7-day gap to
occur between prescription redemptions before a patient was
considered to have discontinued the medication. To examine
the sensitivity of the estimates to differences in exposure
denition, we repeated the analysis with allowance for a 14day gap between prescription redemptions. The ndings were
consistent with the primary analysis and are not further
reported. Because some adverse effects may arise shortly after
therapy initiation and because the inclusion of long-term
users may lead to underestimation of these complications
[29], we further categorized current users into two groups:
new users, for whom the number of days supplied plus
7 days from a rst-ever prescription covered the index date;
and long-term users, for whom the number of days supplied
plus 7 days from a prescription subsequent to their rst
prescription covered the index date. We dened former users
as those whose index date fell after the most recent
dispensation date plus the computed number of days
supplied plus 7 days. The reference category consisted of
those who had not used any antiplatelet drugs during the
study period (never users).
We assessed the association between antiplatelet drug use
and the risk of SAH using odds ratios (ORs) with 95%
condence intervals (CIs) as a measure of relative risk. Because
we used risk-set sampling of controls, the ORs are unbiased
estimates of the incidence rate ratio [30]. We used conditional
logistic regression to estimate ORs for SAH among current,
new, long-term and former users of LDA, HDA, clopidogrel,
dipyridamole, and LDA plus dipyridamole, as compared with
never users. In regression analyses, we controlled for confounders using the change-in-estimate method [31]. In a
forward stepwise fashion, the covariates that made the largest
change in estimate above 5% were included. Covariates eligible
for inclusion were hypertension, obesity, COPD, alcoholismrelated disease, diabetes mellitus, and use of NSAIDs, SSRIs,
or vitamin K antagonists. Only diabetes mellitus and hypertension fullled the criteria and were included in the model.
We lacked data on lifestyle factors such as smoking, which is
associated with a two-fold to three-fold increased risk of SAH
[11]. Given the association of smoking with cardiovascular
disease, it is also expected to be more common among
antiplatelet drug users than among never users [5,3234]. We

Results
We identied 1186 patients with a rst non-traumatic SAH
between 1997 and 2008 (incidence: 8.6 per 100 000 personyears) and 11 846 population controls (Table 1). The median
age was 57 years. Almost two-thirds (64.6%) of cases were
female. Comorbidity was more frequent among cases than
among controls: 1.4% of cases had alcoholism-related
disease, as compared with 0.8% of controls; 4.1% of cases
had COPD, as compared with 3.8% of controls; and 7.6%

Table 1 Characteristics of subarachnoid hemorrhage cases and their

population controls
Cases, no. (%)
n = 1186
Fenale sex*
Age* group, years
1550
5165
> 65
Antiplatelet drugs
Never use
Current use
Former use
Comorbidity
Alcoholism-related disease
COPD
Diabetes mellitus
Hypertension
Ischemic vascular disease
Obesity
Rare causes of SAH
Comedication
HRT
ACE inhibitors
b-Blockers
Calcium channel blockers
Oral glucocorticoids
Statins
NSAIDs
Vitamin K antagonists
SSRIs

Controls, no. (%)

n = 11 840

765 (64.6)

7650 (64.6)

397 (33.5)
504 (42.6)
283 (23.9)

4008 (33.9)
4981 (42.1)
2851 (24.1)

1005 (84.9)
109 (9.2)
70 (5.9)

10 308 (87.1)
910 (7.7)
622 (5.3)

16
48
17
90
67
5

(1.4)
(4.1)
(1.4)
(7.6)
(5.7)
(0.4)

97
452
322
587
591
127
37

(0.8)
(3.8)
(2.7)
(5.0)
(5.0)
(1.1)
(0.3)

79
113
85
100
23
72
163
8
59

(6.7)
(9.5)
(7.2)
(8.5)
(1.9)
(6.1)
(13.8)
(0.7)
(5.0)

826
1302
936
826
187
788
905
156
579

(7.0)
(11.0)
(7.9)
(7.0)
(1.6)
(6.7)
(7.6)
(1.3)
(4.9)

ACE, angiotensin-converting enzyme; COPD, chronic obstructive

pulmonary disease; HRT, hormone replacement therapy; NSAID,
non-steroidal anti-inammatory drug; SAH, subarachnoid hemorrhage; SSRI, selective serotonin reuptake inhibitor. *Controls were
matched to cases on this factor. Acetylsalicylic acid (low or high
dose), clopidogrel, or dipyridamole. Ischemic heart disease, cerebral
ischemia, or two or more prescriptions of nitroglycerin. Coagulopathy, arteritis, polycystic kidney disease, and brain tumors. Current
use.
2010 International Society on Thrombosis and Haemostasis

Antiplatelet drugs and subarachnoid hemorrhage 1471

of cases had diagnosed hypertension, as compared with 5.0%

of controls.
One hundred and nine cases (9.2%) and 910 controls (7.7%)
were current antiplatelet drug users. Among cases, 104 (8.8%)
were current LDA users, one (0.1%) used HDA, three (0.3%)
used clopidogrel, 11 (0.9%) used dipyridamole, and 12 (1.0%)
used LDA plus dipyridamole. Among controls, 891 (7.5%)
used LDA, two (0.02%) used HDA, 24 (0.2%) used clopidogrel, 48 (0.4%) used dipyridamole, and 58 (0.5%) used LDA
plus dipyridamole. The low numbers of clopidogrel and HDA
users hindered further analysis of the association between these
drugs and SAH occurrence. As compared with never use, the
adjusted ORs associating SAH occurrence with LDA use were
1.13 (95% CI 0.891.42) among current users, 2.52 (95% CI
1.374.62) among new users, and 1.03 (95% CI 0.811.32)
among long-term users (Table 2). There were no new dipyridamole users. Therefore, all current users were long-term users
with an adjusted OR for SAH of 2.09 (95% CI 1.044.23). A
similar increased risk was seen for long-term users of LDA plus
dipyridamole (OR 1.80, 95% CI 0.903.60). Former use of
LDA or dipyridamole was not associated with SAH occurrence. The bias analysis revealed that lack of adjustment for
smoking may have led us to slightly overestimate the ORs, by
2% for LDA users and by 10% for dipyridamole users.
Assuming a valid bias model, the ORs were thus not changed
substantially.
Discussion
In this population-based casecontrol study, long-term use of
dipyridamole, but not of LDA, was associated with an

Table 2 Relative risks for subarachnoid hemorrhage among current users

of low-dose acetylsalicylic acid (LDA) and/or dipyridamole
No. of
cases/
controls
Never use
1005/10 308
LDA
Current use
104/891
New use
14/52
Long-term use
90/839
Former use
69/608
Dipyridamole
Current use
11/48
New use

Long-term use
11/47
Former use
2/18
LDA plus dipyridamole
Current use
12/58
New use
1/3
Long-term use
11/55

Relative risks* with 95%

condence intervals
Unadjusted
1 (reference)
1.22
2.67
1.12
1.15

(0.971.53)
(1.464.90)
(0.881.42)
(0.881.49)

Adjusted
1 (reference)
1.13
2.52
1.03
1.07

(0.891.42)
(1.374.62)
(0.811.32)
(0.821.40)

2.38 (1.214.68)

2.44 (1.244.82)
1.26 (0.295.56)

2.04 (1.014.12)

2.09 (1.044.23)
1.18 (0.275.20)

2.19 (1.154.17)
2.86 (1.595.17)
2.10 (1.074.10)

1.88 (0.973.67)
2.68 (1.484.84)
1.80 (0.903.60)

*Odds ratios were computed with conditional logistic regression.

Adjusted for diabetes mellitus and hypertension, which were selected

using a 5% change-in-estimate method. No redeemed prescription for
acetylsalicylic acid (low or high dose), clopidogrel, or dipyridamole.
2010 International Society on Thrombosis and Haemostasis

increased risk of SAH. Supporting these ndings, a similar

increased risk was found among patients receiving long-term
treatment with LDA plus dipyridamole. New LDA use was
also associated with an increased SAH risk. Because of the
limited precision of the risk estimates for dipyridamole and
new LDA use, however, caution is advised in their interpretation.
Our study of 1186 SAH patients is the rst to examine
specically the SAH risk associated with antiplatelet drug use.
The CAPRIE study, which compared the efcacy of antiplatelet treatment with clopidogrel and ASA in patients with
ischemic vascular disease, reported a trend towards a lower
incidence of all types of intracranial hemorrhage (ICH) among
clopidogrel users (0.31%) than among ASA users (0.42%)
[36,37]. The relative ICH risk between these drugs and placebo
could not be examined. Dual antiplatelet drug therapy with
ASA plus clopidogrel in patients with acute coronary syndromes increases (although non-signicantly) the risk of lifethreatening bleeds as compared with ASA monotherapy [24].
In patients with ischemic stroke or transient ischemic attack,
adding ASA to clopidogrel increases the risk of ICH [38],
whereas adding dipyridamole to ASA does not [5]. However,
earlier reports of ICH risks associated with use of antiplatelet
drugs may not be generalizable to the risks of SAH, because of
substantial differences in pathogenesis between different types
of ICH.
Ruptured aneurysms form part of the pathogenesis in 85%
of SAHs. Dipyridamole potentiates endogenous adenosine
actions by blocking its cellular uptake, and thereby inhibits
platelet aggregation. Dipyridamole also acts as a potent
vasodilator [39]. Thus, stretching the aneurysm wall may also
increase the risk of rupture, and explain why dipyridamole use
increased SAH risk.
The incidence of SAH and the median age of SAH patients
in our study were consistent with those reported in previous
studies [8,40]. Although SAH is a rare event, comprising only
5% of all strokes [8], it occurs at a young age and has a 50%
short-term case fatality rate [8], making the loss of productive
life years in the general population as large as that from
cerebral infarction, the most common type of stroke [41]. Our
nding of long-term dipyridamole use and new LDA use as
risk factors for SAH is thus of clinical importance, although the
absolute risk increase is too small to warrant a change in
practice regarding their use in preventing thromboembolic
cardiovascular events.
Several issues should be considered when interpreting our
results. The population-based design within a tax-supported
universal healthcare system, with complete, prospectively
recorded hospital and prescription histories and with access
to appropriate population controls, reduced the risk of referral,
diagnostic and information biases [42].
In Denmark, most SAH patients are admitted to the
nearest hospital and are then subsequently transferred to
specialized departments at university hospitals for more
advanced diagnostic testing and treatment. However, patients
for whom no curative treatment exists such as very old and

1472 M. Schmidt et al

frail patients, and patients with very extensive SAHs are

occasionally not transferred to the specialized centers, and
instead are given palliative therapy at the site of primary
admission. If antiplatelet drug users had more severe SAHs
leading to high immediate mortality, then they would have a
low likelihood of being transferred to neurosurgery or
neurology departments. This scenario would lead to underestimation of the association between antiplatelet drug use and
SAH risk, and thus could not explain our ndings of an
increased SAH risk.
We restricted our study to patients who were admitted to
neurosurgery or neurology departments at some point during
their hospitalization with SAH, because the PPV of SAH
diagnosis has been reported to be adequate among patients
admitted to these departments [9,14,15]. Although this
requirement implies that some true SAH cases were omitted
from the case set, the misclassication should not affect the
prescription of antiplatelet drugs, resulting in little bias of the
estimates of association [43]. The vast majority (95%) of our
cases were admitted to neurosurgery departments, where the
proportion of correctly diagnosed SAH patients is very high.
It is therefore unlikely that the lower quality of SAH diagnosis
at neurology departments has inuenced our results [9,14,15].
If current antiplatelet drug users were more likely than never
users to be diagnosed with SAH because of more thorough
clinical work-up, it would have increased our relative risk
estimates.
We relied on redeemed prescriptions as a proxy for actual
drug use, but this may not always be correct [44]. However, the
known benecial effect of antiplatelet drugs in patients at high
risk of severe cardiovascular disease, the fact that the drug
exposure information was based on actual dispensing at
pharmacies and the copayment requirements all increase the
likelihood of adherence to prescribed medication. Also, any
misclassication due to non-adherence would attenuate the
estimates of association.
All ASA preparations are available over the counter in
Denmark. However, LDA is used almost exclusively for
secondary prevention of cardiovascular disease, and is generally prescribed by physicians, because the prescriptions are then
50% reimbursable through the national health insurance
program [45,46]. Clopidogrel and dipyridamole are not sold
over the counter in Denmark. We therefore probably succeeded in identifying most patients who were using antiplatelet
drugs on a regular basis.
Extensive efforts were made to adjust the risk estimates for
possible measured and unmeasured confounding factors,
particularly known risk factors for SAH. We found that the
increased SAH risk associated with long-term use of dipyridamole or new use of LDA was unlikely to be explained by
unmeasured confounding by smoking. Our non-randomized
design, however, may still be vulnerable to residual confounding by undiagnosed hypertension, lifestyle factors such as body
mass index, or unmeasured variables, particularly the underlying condition leading to antiplatelet drug use. However,
former dipyridamole use was not associated with SAH

occurrence, which supports a true drug effect. Even when

there is a close temporal relationship between drug intake and
the development of SAH, we cannot rule out at least some
inuence from disease severity, because new users may have
more severe underlying disease. Both dipyridamole and LDA
are indicated in ischemic stroke patients, who are at increased
risk of subsequent intracerebral hemorrhage [37]. New users of
LDA had an increased risk of SAH, which may be attributable
either to early adverse effects of LDA or to an increased risk of
SAH developing from an intracerebral bleed expanding into
the subarachnoid space. These early-onset mechanisms may
also apply to dipyridamole use. We could not examine this
supposition, however, because all dipyridamole users were
long-term users.
Conclusion
Although risk estimates were imprecise, long-term dipyridamole use and new LDA use were associated with an increased
risk of SAH. Long-term LDA use was not associated with
SAH.
Acknowledgements
The Danish Medical Research Council (grant 271-05-0511)
and the Clinical Epidemiological Research Foundation, Denmark, funded this study. Neither had a role in its design,
conduct, analysis, or reporting. None of the authors received
any fees, honoraria, grants or consultancies that would
constitute a conict of interest with the current study. The
Department of Clinical Epidemiology, Aarhus University
Hospital, receives funding for other studies from companies
in the form of research grants to (and administered by) Aarhus
University. None of these studies have any relation to the
present study.
Disclosure of Conflict of Interests
The authors state that they have no conict of interest.
Appendix: ICD and ATC codes
Antiplatelet drugs

LDA: ATC: B01AC06 or N02BA01 (restricted to drug

identication codes: 17048, 19019, 24503, 24514, 24618,
39178, 39187, 44642, 50526, 65052, 95828, 152231, 432682,
459065, 459479, 459883, 506485, 519371)
HDA: ATC: N02BA01 (restricted to drug identication codes:
32375, 111708, 137232, 196766, 198754, 453985, 453993,
501403, 504209, 512905, 512947, 594093)
Clopidogrel: ATC: B01AC04
Dipyridamole: ATC: B01AC07
LDA plus dipyridamole: ATC: B01AC30 or combinations of
LDA and dipyridamole codes

2010 International Society on Thrombosis and Haemostasis

Antiplatelet drugs and subarachnoid hemorrhage 1473

Comorbidity

Hypertension: ICD-8: 401.99; ICD-10: DI10DI15

Obesity: ICD-10: DE65DE68
Rare causes of SAH
Coagulopathy: ICD-8: 286 (except 286.59), 287; ICD-10:
DD66DD69 (except DD683)
Arteritis: ICD-8: 446.29, 136.02, 446.09; ICD-10: DM313,
DM352, DA692E, DM301, DM300, DI675
Polycystic kidney disease: ICD-8: 753.19; ICD-10: Q61.2
Brain tumor: ICD-8: 191, 192.19, 238.19, 238.39; ICD-10:
DC71, DC700, DC793, DC793A, DC793B, DD420,
DD430DD432
Alcoholism-related disease: ICD-8: 291, 303, 577.10, 571.09,
571.10; ICD-10: F10 (except F10.0), G31.2, G62.1, G72.1, I
42.6, K29.2, K86.0, Z72.1; ATC: N07BB01
COPD: ICD-10: DJ44; ATC: R03
Diabetes mellitus: ICD-8: 249250; ICD-10: E10E14, O24
(except O24.4), H36.0; ATC: A10B, A10A
Ischemic vascular disease
Ischemic heart disease: ICD-8: 410414; ICD-10: DI20DI25
Cerebral ischemia: ICD-8: 433434, 436; ICD-10: DI63
Nitroglycerin (if two or more prescriptions): ATC: C01DA02,
C01DA08, C01DA014

6
7

8
9

10

Comedication

Hormone replacement therapy (prescription redeemed within

120 days of index date): ATC: G03C
Angiotensin-converting enzyme inhibitors (prescription redeemed within 120 days of index date): ATC: C09 (C02
before 1 January 1996)
b-Blockers (prescription redeemed within 120 days of index
date): ATC: C07
Calcium channel blockers (prescription redeemed within
120 days of index date): ATC: C08
Oral glucocorticoids (prescription redeemed within 60 days of
index date): ATC: H02AB
Statins (prescription redeemed within 120 days of index date):
ATC: C10AA, B04AB0
NSAIDs (prescription redeemed within 60 days of index date):
ATC: M01A
Vitamin K antagonists (prescription redeemed within 90 days
of index date): ATC: B01AA03, B01AA04
SSRIs (prescription redeemed within 120 days of index date):
ATC: N06AB
References
1 Antithrombotic Trialists Collaboration. Collaborative meta-analysis
of randomised trials of antiplatelet therapy for prevention of death,
myocardial infarction, and stroke in high risk patients. BMJ 2002; 324:
7186.
2 Sabatine MS, Cannon CP, Gibson CM, Lopez-Sendon JL, Montalescot G, Theroux P, Claeys MJ, Cools F, Hill KA, Skene AM, McCabe
CH, Braunwald E, for the CLARITY-TIMI 28 Investigators; Addition of clopidogrel to aspirin and brinolytic therapy for myocardial

2010 International Society on Thrombosis and Haemostasis

11

12
13

14

15

16

17

18
19
20
21

22

infarction with ST-segment elevation. N Engl J Med 2005; 352: 1179

89.
Steinhubl SR, Berger PB, Mann JT, Fry ET, DeLago A, Wilmer C,
Topol EJ; Credo Investigators. Clopidogrel for the Reduction of
Events During Observation. Early and sustained dual oral antiplatelet
therapy following percutaneous coronary intervention: a randomized
controlled trial. JAMA 2002; 288: 241120.
Yusuf S, Zhao F, Mehta SR, Chrolavicius S, Tognoni G, Fox KK;
Clopidogrel in Unstable Angina to Prevent Recurrent Events Trial
Investigators. Eects of clopidogrel in addition to aspirin in patients
with acute coronary syndromes without ST-segment elevation. N Engl
J Med 2001; 345: 494502.
ESPRIT Study Group, Halkes PH, van Gijn J, Kappelle LJ,
Koudstaal PJ, Algra A. Aspirin plus dipyridamole versus aspirin alone
after cerebral ischaemia of arterial origin (ESPRIT): randomised
controlled trial. Lancet 2006; 367: 166573.
Vallurupalli NG, Goldhaber SZ. Gastrointestinal complications of
dual antiplatelet therapy. Circulation 2006; 113: e6558.
Serebruany VL, Malinin A, Eisert R, Sane D. Risk of bleeding complications with antiplatelet agents: meta-analysis of 338,191 patients
enrolled in 50 randomized controlled trials. Am J Hematol 2004; 75:
407.
van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet
2007; 369: 30618.
Gaist D, Vaeth M, Tsiropoulos I, Christensen K, Corder E, Olsen J,
Sorensen HT. Risk of subarachnoid haemorrhage in rst degree
relatives of patients with subarachnoid haemorrhage: follow up
study based on national registries in Denmark. BMJ 2000; 320: 141
5.
Rinkel GJ, Djibuti M, Algra A, van Gijn J. Prevalence and risk of
rupture of intracranial aneurysms: a systematic review. Stroke 1998;
29: 2516.
Feigin VL, Rinkel GJ, Lawes CM, Algra A, Bennett DA, van Gijn J,
Anderson CS. Risk factors for subarachnoid hemorrhage: an updated
systematic review of epidemiological studies. Stroke 2005; 36: 277380.
Frank L. Epidemiology. When an entire country is a cohort. Science
2000; 287: 23989.
Andersen TF, Madsen M, Jorgensen J, Mellemkjaer L, Olsen JH. The
Danish National Hospital Register. A valuable source of data for
modern health sciences. Dan Med Bull 1999; 46: 2638.
Ingeman A, Pedersen L, Hundborg HH, Petersen P, Zielke S, Mainz J,
Bartels P, Johnsen SP. Quality of care and mortality among
patients with stroke: a nationwide follow-up study. Med Care 2008; 46:
639.
Johnsen SP, Pedersen L, Friis S, Blot WJ, McLaughlin JK, Olsen JH,
Srensen HT. Nonaspirin nonsteroidal anti-inammatory drugs and
risk of hospitalization for intracerebral hemorrhage: a populationbased casecontrol study. Stroke 2003; 34: 38791.
Pedersen CB, Gtzsche H, Mller JO, Mortensen PB. The Danish
Civil Registration System. A cohort of eight million persons. Dan Med
Bull 2006; 53: 4419.
Wacholder S, McLaughlin JK, Silverman DT, Mandel JS. Selection of
controls in casecontrol studies. I. Principles. Am J Epidemiol 1992;
135: 101928.
Gaist D, Sorensen HT, Hallas J. The Danish prescription registries.
Dan Med Bull 1997; 44: 4458.
Available drugs and package sizes on the Danish drug market. http://
www.lmk.dk; Accessed 4 June 2010.
Bowman AJ, Chen CP, Ford GA. Nitric oxide mediated venodilator
eects of nebivolol. Br J Clin Pharmacol 1994; 38: 199204.
Ghiadoni L, Magagna A, Versari D, Kardasz I, Huang Y, Taddei S,
Salvetti A. Dierent eect of antihypertensive drugs on conduit artery
endothelial function. Hypertension 2003; 41: 12816.
Whitworth JA, Schyvens CG, Zhang Y, Andrews MC, Mangos GJ,
Kelly JJ. The nitric oxide system in glucocorticoid-induced hypertension. J Hypertens 2002; 20: 103543.

1474 M. Schmidt et al
23 Wolfrum S, Jensen KS, Liao JK. Endothelium-dependent eects of
statins. Arterioscler Thromb Vasc Biol 2003; 23: 72936.
24 Zhang X, Hintze TH. Amlodipine releases nitric oxide from canine
coronary microvessels: an unexpected mechanism of action of a
calcium channel-blocking agent. Circulation 1998; 97: 57680.
25 Khurana VG, Meissner I, Sohni YR, Bamlet WR, McClelland RL,
Cunningham JM, Meyer FB. The presence of tandem endothelial nitric oxide synthase gene polymorphisms identifying brain aneurysms
more prone to rupture. J Neurosurg 2005; 102: 52631.
26 Johnsen SP, Srensen HT, Mellemkjoer LF, Blot WJ, Nielsen GL,
McLaughlin JK, Olsen JH. Hospitalisation for upper gastrointestinal
bleeding associated with use of oral anticoagulants. Thromb Haemost
2001; 86: 5638.
27 Mellemkjaer L, Blot WJ, Sorensen HT, Thomassen L, McLaughlin
JK, Nielsen GL, Olsen JH. Upper gastrointestinal bleeding among
users of NSAIDs: a population-based cohort study in Denmark. Br J
Clin Pharmacol 2002; 53: 17381.
28 Serebruany VL. Selective serotonin reuptake inhibitors and increased
bleeding risk: are we missing something? Am J Med 2006; 119: 11316.
29 Ray WA. Evaluating medication eects outside of clinical trials: newuser designs. Am J Epidemiol 2003; 158: 91520.
30 Rothman KJ, Greenland S, Lash TL. Casecontrol studies. In:
Rothman KJ, Greenland S, Lash TL, eds. Modern Epidemiology, 3th
edn. Philadelphia, PA: Lippincott Williams & Wilkins, 2008: 1245.
31 Rothman KJ, Greenland S, Lash TL. Selecting confounders for control. In: Rothman KJ, Greenland S, Lash TL, eds. Modern Epidemiology, 3th edn. Philadelphia, PA: Lippincott Williams & Wilkins, 2008:
2614.
32 Gender-specic smoking prevalence in Denmark 2003. World Development Indicators Database. http://www.nationmaster.com/graph/
hea_smo_pre_mal_of_adu-health-smoking-prevalence-males-adults&
date=2003.
33 Centers for Disease Control. State-specic prevalence of cigarette
smoking among adults in Maryland, US, 2002. http://www.cdc.gov/
mmwr/preview/mmwrhtml/mm5253a1.htm.
34 Blasco-Colmenares E, Perl TM, Guallar E, Baumgartner WA, Conte
JV, Alejo D, Pastor-Barriuso R, Sharrett AR, Faraday N. Aspirin plus
clopidogrel and risk of infection after coronary artery bypass surgery.
Arch Intern Med 2009; 169: 78896.
35 Lash TL, Fox MP, Fink AK. Applying Quantitative Bias Analysis to
Epidemiologic Data. New York: Springer, 2009: 13842.

36 CAPRIE Steering Committee. A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events (CAPRIE).
Lancet 1996; 348: 132939.
37 Hart RG, Tonarelli SB, Pearce LA. Avoiding central nervous system
bleeding during antithrombotic therapy: recent data and ideas. Stroke
2005; 36: 158893.
38 Diener HC, Bogousslavsky J, Brass LM, Cimminiello C, Csiba L,
Kaste M, Leys D, Matias-Guiu J, Rupprecht HJ, MATCH investigators. Aspirin and clopidogrel compared with clopidogrel alone after
recent ischaemic stroke or transient ischaemic attack in high-risk
patients (MATCH): randomised, double-blind, placebo-controlled
trial. Lancet 2004; 364: 3317.
39 Gamboa A, Abraham R, Diedrich A, Shibao C, Paranjape SY, Farley
G, Biaggioni I. Role of adenosine and nitric oxide on the mechanisms
of action of dipyridamole. Stroke 2005; 36: 21705.
40 de Rooij NK, Linn FH, van der Plas JA, Algra A, Rinkel GJ. Incidence of subarachnoid haemorrhage: a systematic review with
emphasis on region, age, gender and time trends. J Neurol Neurosurg
Psychiatry 2007; 78: 136572.
41 Johnston SC, Selvin S, Gress DR. The burden, trends, and demographics of mortality from subarachnoid hemorrhage. Neurology 1998;
50: 141318.
42 Srensen HT. Regional administrative health registers as a resource in
clinical epidemiology. A study of options, strengths, limitations and
data quality provided with examples of use. Int J Risk Safety Med
1997; 10: 122.
43 Brenner H, Savitz DA. The eects of sensitivity and specicity of case
selection on validity, sample size, precision, and power in hospitalbased casecontrol studies. Am J Epidemiol 1990; 132: 18192.
44 Ho PM, Bryson CL, Rumsfeld JS. Medication adherence: its
importance in cardiovascular outcomes. Circulation 2009; 119: 3028
35.
45 Mose H, Larsen M, Riis A, Johnsen SP, Thomsen RW, Srensen HT.
Thirty-day mortality after peptic ulcer bleeding in hospitalized patients
receiving low-dose aspirin at time of admission. Am J Geriatr Pharmacother 2006; 4: 24450.
46 Srensen HT, Mellemkjaer LF, Blot WJ, Nielsen GL, Steensen FH,
McLaughlin JK, Olsen JH. Risk of upper gastrointestinal bleeding
associated with use of low-dose aspirin. Am J Gastroenterol 2000; 95:
221824.

2010 International Society on Thrombosis and Haemostasis

Copyright of Journal of Thrombosis & Haemostasis is the property of Wiley-Blackwell and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.