Compliance With Secondary Prevention of Ischemic Stroke

A Prospective Evaluation
Tanja Sappok, MD; Andreas Faulstich; Erika Stuckert; Holger Kruck;
Peter Marx, MD; Hans-Christian Koennecke, MD

Background and Purpose—Compliance with pharmacological therapy is essential for the efficiency of secondary
prevention of ischemic stroke. Few data exist regarding patient compliance with antithrombotic and risk factor treatment
outside of controlled clinical trials. The aim of the present study was to assess the rate of and predictors for compliance
with secondary stroke prevention 1 year after cerebral ischemia and to identify reasons for noncompliance.
Methods—Patients with a diagnosis of ischemic stroke or TIA and antithrombotic discharge medication were prospectively
recruited. At 1 year, the proportion of patients compliant with antithrombotic treatment and with medication for risk
factors (eg, hypertension, diabetes, hyperlipidemia) was evaluated through structured telephone interviews. In addition,
the reasons for nontreatment with antithrombotic and risk factor medication were determined. Independent predictors
for compliance were analyzed by logistic regression analyses.
Results—Of 588 consecutive patients admitted to our stroke unit, 470 had a discharge diagnosis of cerebral ischemia (TIA
26.2%, cerebral infarct 73.8%) and recommendations for antithrombotic therapy. At 1 year, 63 patients (13.4%) had died
and 21 (4.5%) were lost to follow-up, thus, 386 could finally be evaluated. Of the patients, 87.6% were still on
antithrombotic medication, and 70.2% were treated with the same agent prescribed on discharge. Of the patients with
hypertension, diabetes, and hyperlipidemia, 90.8%, 84.9%, and 70.2% were still treated for their respective risk factors.
Logistic regression analyses revealed age (OR 1.03, 95% CI 1.00 to 1.06), stroke severity on admission (OR 1.09, 95%
CI 1.00 to 1.20), and cardioembolic cause (OR 4.13, 95% CI 1.23 to 13.83) as independent predictors of compliance.
Conclusions—Compliance with secondary prevention in patients with ischemic stroke is rather good in the setting of our
study. Higher age, a more severe neurological deficit on admission, and cardioembolic stroke cause are associated with
better long-term compliance. Knowledge of these determinants may help to further improve the quality of stroke
Downloaded from http://ahajournals.org by on November 11, 2019

prevention. (Stroke. 2001;32:1884-1889.)

Key Words: antithrombotic therapy 䡲 medical management 䡲 prevention 䡲 stroke, ischemic

S troke is a leading cause of death and disability in

industrialized countries.1 The efficacy of antithrombotic
medication to prevent recurrence of stroke and death has been
proven in several trials.2,3 Furthermore, little doubt exists
regarding the efficiency of secondary stroke prevention
through the treatment of major stroke risk factors.4 –9 How-
ever, in particular, the stroke-preventive effect of antiplatelet
treatment is only moderate10 and is further reduced, if not
abolished, by poor compliance.
Patient compliance may be different under the monitoring
conditions of controlled trials compared with daily life
conditions. Information regarding compliance with anti-
thrombotic medication and with therapy of stroke risk factors
in routine general practice is scarce. In the present study, we
therefore sought to (1) determine the prevalence of compli-
ance with antithrombotic medication and pharmacological
therapy of stroke risk factors (hypertension, diabetes, and
hyperlipidemia) 1 year after an ischemic stroke or a TIA and
to (2) identify the predictors for compliance and reasons for
discontinuation or changes of a specific medication.
Subjects and Methods
The study was conducted at an academic medical center in the south
of Berlin, Germany, serving a population of ⬇500 000. Every year,
⬇400 patients with acute stroke are admitted to our stroke unit.
Consecutive patients with a confirmed diagnosis of TIA or nonfatal
ischemic stroke treated with an antithrombotic agent (eg, aspirin,
ticlopidine/clopidogrel, phenprocoumon) at discharge were prospec-
tively enrolled. Patients with a diagnosis of cerebral hemorrhage,
migrainous aura, seizure, or any other nonischemic pathology were
excluded. Patients were considered to have had an ischemic stroke or
a TIA if there was a focal neurological deficit of sudden onset with
no known alternative to an ischemic vascular cause. All patients
included in this study underwent CT or MRI. Deficits that lasted
⬎24 hours were diagnosed as infarctions,11 and those with a duration
of ⬍24 hours and a negative brain scan were diagnosed as TIAs.

Received January 30, 2001; final revision received March 21, 2001; accepted May 9, 2001.
From the Department of Neurology, Stroke Unit, Universitätsklinikum Benjamin Franklin, Freie Universität Berlin, Germany.
Correspondence to Hans-Christian Koennecke, MD, Department of Neurology, Ev Krankenhaus Königin Elisabeth Herzberge, Herzbergstr 79, 10362
Berlin, Germany. E-mail h.koennecke@keh-berlin.de
© 2001 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org

1884
 Sappok et al Compliance With Secondary Prevention of Ischemic Stroke 1885

TABLE 1. Baseline Characteristics of the 470 Patients

Primarily Enrolled
Characteristic
Age, mean⫾SD, y 67.2⫾12.6
Sex, n (%)
Men 265 56.4
Women 205 43.6
Risk factors, n (%)
Hypertension 318 67.7
Diabetes 118 25.1
Hyperlipidemia 193 41.1
Figure 1. Compliance with antithrombotic therapy as recom-
History of smoking 162 34.5 mended on discharge. Numbers in bold and percentages refer
Subtype of cerebral ischemia, n (%) to the 470 patients who were initially enrolled. Numbers in
parentheses indicate percentages of the 386 patients who were
TIA 123 26.2 finally evaluated at 1 year.
Cerebral infarct 347 73.8
secondary analysis assessed changes among different subgroups of
antithrombotic agents during follow-up.
Informed consent was obtained from all patients or their next of kin. Data were recorded in a personal computer using a special mask
The study was approved by the ethics committee of our hospital. for data entry. In a first step, univariate analyses (␹2 test for
In addition to the respective discharge medication, different categorical factors, Mann-Whitney U test for ordinal factors) were
sociodemographic and medical variables were recorded, including conducted to reveal associations of different variables with compli-
age, sex, care situation, occupation, major stroke risk factors, stroke ance with antithrombotic treatment. In a second step, the independent
severity (as measured by the National Institutes of Health Stroke effects of the variables with P⬍0.1 in the univariate analyses were
Scale [NIHSS] on admission and the Barthel Index and Rankin Scale assessed by stepwise backward logistic regression analyses. For each
at discharge), and subtype of cerebral ischemia (according to the association of interest, an adjusted OR and the 95% CI were
TOAST criteria12). calculated. OR, 95% CI, and P value were determined just before
Compliance with discharge medication, smoking habits, and elimination. A difference was regarded as statistically significant at
complications were evaluated at 1 year after admission. Follow-up P⬍0.05. Data were analyzed with SPSS (release 9.0.1) statistical
information was obtained through a structured telephone interview software.
Downloaded from http://ahajournals.org by on November 11, 2019

with the patient. If the patient could not be contacted, information

was obtained from relatives, caregivers, or the patient’s general
practitioner. The interview consisted of questions regarding compli-
cations (death, recurrent cerebral ischemia or hemorrhage, myocar-
dial infarction) and the patient’s current medication. If the medica-
tion differed from the recommendations at discharge, the
questionnaire was extended to reveal (1) the person who initiated the
discontinuation or change (general practitioner, physician at rehabil-
itation facility, patient, unknown) and (2) the reasons for the change
or discontinuation (side effects, contraindication, inefficacy, other
[eg, costs, drug interaction], unknown). For antithrombotic therapy,
inefficacy was assumed in patients with recurrent cerebrovascular
ischemic events, whereas insufficient control of the respective risk
factor was the criterion for inefficacy of risk factor medication. The
primary analyses were focused on the proportion of patients treated
with any antithrombotic agent or stroke risk factor medication at 1
year and predictors for compliance with antithrombotic treatment. A
Results
Between March 1998 and June 1999, a total of 588 consec-
utive patients were admitted to our stroke unit. One hundred
eight patients were excluded from the study due to intracra-
nial hemorrhage or other discharge diagnoses (eg, migrainous
aura, seizures), and 10 were excluded due to fatal ischemic
stroke. Of the remaining 470 patients with cerebral ischemia
(for baseline characteristics, see Table 1), 63 (13.4%) died
during the year after the event and an additional 21 (4.5%)
were lost to follow-up. Thus, 386 patients were finally
evaluated at 1 year (Figure 1). The distributions of stroke
severity on admission and at discharge for patients with
cerebral infarction are given in Figure 2. During the

Figure 2. Frequency distribution of stroke severity on admission and at discharge of patients with confirmed cerebral infarction
(n⫽347).
 1886 Stroke August 2001

TABLE 2. Compliance With Medical Stroke Prevention and Initiation of

Discontinuance/Change in 386 Patients Evaluated at 1 Year
Antiaggregant Antihypertensive Insulin or Oral Antilipidemic
or Anticoagulant Drug Antidiabetic Drug Drug

n % n % n % n %
Compliance rate
On any treatment 338 87.6 198 90.8 62 84.9 40 70.2
On discharge medication 271 70.2 122 56.0 48 65.8 21 36.8
No medication 48 12.4 20 9.2 11 15.1 17 29.8
Total (surveyed at 1 year) 386 100 218 100 73 100 57 100
Discontinuance/change initiated by
Hospital physician 27 23.5 11 11.5 5 20 4 11.1
General practitioner 62 53.9 74 77.1 10 40.0 24 66.7
Patient’s will 14 12.2 2 2.1 0 0 2 5.6
Unknown 12 10.4 9 9.4 10 40.0 6 16.7
Total (not on discharge 115 100 96 100 25 100 36 100
medication)

follow-up period, 25 (6.5%) patients had a recurrent cerebral
ischemia, 1 patient had a cerebral hemorrhage, and 3 patients
had a myocardial infarction. Treatment rates at 1 year are
given in Table 2. Of 386 patients, 87.6% were still on
antithrombotic treatment, and 70.2% received the same anti-
thrombotic agent as prescribed on discharge.
Table 3 demonstrates the different compliance rates ac-
cording to subclasses of antithrombotic medication. In par-
ticular, poor compliance (⬍60%) was determined in patients
Downloaded from http://ahajournals.org by on November 11, 2019
sons for discontinuation or changes of antithrombotic ther-
apy, side effects (47.5%) and inefficacy (23.7%) were most
common, whereas contraindications (13.6%) and other rea-
sons (15.2%) accounted for the remainder. For risk factor
medication, inefficacy (40.7%) was stated most frequently,
followed by other reasons (33.9%), side effects (20.3%), and
contraindications (5.1%). However, due to the large propor-
tion of unknown reasons in either group (48.7% and 62.4%
for antithrombotic and risk factor medication, respectively),

discharged on thienopyridines (ticlopidine n⫽35, compliance
40%; clopidogrel n⫽55, compliance 54.5%), whereas ⬎75%
of the patients discharged on aspirin or phenprocoumon were
on the identical medication at follow-up. Only 4.5% of the
patients with a recommendation for anticoagulation were
without any antithrombotic medication at 1 year, whereas
14% of the patients discharged on aspirin or thienopyridines
at discharge received no antithrombotic agent at follow-up.
Changes in therapy among the different antithrombotic agents
during the first year after cerebral ischemia are depicted in
Figure 3. About one third of patients discharged on thienopy-
ridines had been switched to aspirin at 1 year.
Compliance rates for risk factor treatment were highest in
patients with hypertension and lowest in patients with hyper-
lipidemia (Table 2). At 1 year, 47 (35.9%) of 131 smokers
had quit smoking. In the majority of patients, changes in the
antithrombotic and risk factor medication were initiated by
the general practitioner (Table 2). Among the reported rea-
these numbers have to be regarded with caution.
Univariate analyses revealed no association of compliance
with the patient’s care situation or occupation, whereas higher
age, nonsmoking status, more severe clinical deficit (NIHSS
on admission and Rankin Scale at discharge), cardioembolic
cause, and anticoagulation at discharge were associated with
better compliance (P⬍0.1) (Tables 4 and 5). In logistic
regression analyses, higher age, stroke severity, and cardio-
embolic cause were independently associated with better
compliance, but only higher age and cardioembolic cause
remained as significantly associated predictors (Table 6).
Discussion
Detailed knowledge about long-term compliance with sec-
ondary stroke prevention is necessary to focus future strate-
gies for improvement in stroke prevention.13 No long-term
in-depth analyses exist regarding the use of antithrombotic
agents after hospital discharge. The observed compliance

TABLE 3. Compliance According to Subclasses of Antithrombotic Medication

Ticlopidine/
Aspirin Clopidogrel Phenprocoumon

n % n % n %
On any treatment 197 86.0 77 85.6 64 95.5
On discharge medication 176 76.9 44 48.9 51 76.1
No medication 32 14.0 13 14.4 3 4.5
Total (surveyed at 1 year) 229 100 90 100 67 100
n⫽386.
 Sappok et al Compliance With Secondary Prevention of Ischemic Stroke 1887

TABLE 4. Association of Patient Characteristics and TABLE 5. Association of Patient Characteristics and
Compliance With Antithrombotic Therapy in 386 Patients Compliance With Antithrombotic Therapy
Evaluated at 1 Year
Mann-Whitney
Frequency of Ordinal Factor Compliant Noncompliant Test P
Nontreatment Age, y 66.7⫾12 62.2⫾14.8 0.062
Categorical Factor n n % ␹ Test P
2 Barthel Index at 88.0⫾24.2 90.8⫾24.8 0.265
discharge
Sex 0.162
Rankin Scale score at 1.5⫾1.4 1.1⫾1.4 0.034
Men 221 23 10.4
discharge
Women 165 25 15.2
NIHSS on admission 4.4⫾4.9 2.6⫾4.12 0.002
History of cigarette smoking 0.063
Values are mean⫾SD. n⫽386.
Yes 131 22 16.8
No 255 26 10.2
their relatives. Moreover, compliance rates may be lower
Cigarette smoking at follow-up 0.043
with extended follow-up periods. Finally, different treatment
Yes 85 16 18.8 guidelines used in various countries could account for the
No 301 32 10.6 different results.16 In 2 other studies that reported lower
Stroke subtype 0.029 treatment rates, either compliance was evaluated retrospec-
Atherothrombotic 50 7 14.0 tively17 or compliance was defined as the hospital physician’s
Cardioembolic 88 3 3.4 adherence to published guidelines for treatment.18 Conse-
Lacunar 81 14 17.3 quently, these results are hardly comparable to our study.
Even worse treatment rates, ranging from 39% to 62%, have
Unknown/other 167 24 14.4
been reported for primary stroke prevention in patients with
Syndrome 0.481
atrial fibrillation.19 –21 On the other hand, compliance with
TIA 112 16 14.3 antithrombotic medication for secondary prevention of myo-
Stroke 274 32 11.7 cardial infarction is much higher and similar to our results
Discharge medication 0.094 (92%).22
Aspirin 229 32 14.0 To the best of our knowledge, the present study is the first
Ticlopidine/clopidogrel 90 13 14.4 to compare compliance with different antithrombotic agents.
Downloaded from http://ahajournals.org by on November 11, 2019

Phenprocoumon 67 3 4.5 Overall, changes to a different antithrombotic medication

occurred in only 17% of the patients. However, in consider-
ation of the subclasses in detail, only one third of patients
with antithrombotic treatment in the present study is higher discharged on thienopyridines were on this medication at 1
than that in 2 recent studies, which reported rates of 64% and year. Aspirin replaced ticlopidine/clopidogrel in the majority
76%, respectively.14,15 However, 1 study was focused on of changes (88%), although the discharge recommendations
young adults (age 15 to 44 years) only,14 and the duration of for thienopyridines were either aspirin failure (ie, pretreat-
the follow-up period was either markedly shorter (3 ment with aspirin) or contraindications for aspirin. Drug
months)15 or significantly longer (mean 8 years)14 in these prescriptions for outpatients with nonprivate health insurance
studies than in the present study. The present surprisingly are limited by a yearly budget in our country. Thus, the much
high rate of adherence to therapy might further be due to the higher costs for thienopyridines may account for some
specific setting of our study, because all our patients were changes in treatment. However, because the cost factor was
recruited from a stroke unit. In other studies, recruitment of not a specific part of our questionnaire and because the
patients relied on a population-based register15 or multiple
centers of a study group.14 Stroke unit staff might be more TABLE 6. Predictors for Compliance With
committed to provide detailed explanations about the disease Antithrombotic Agents
and the necessity of long-term treatment to the patients and
Rank* OR 95% CI P
Rankin Scale score at 1 0.88 0.62 –1.24 0.458
discharge
Anticoagulant at discharge 2 1.91 0.50 –7.26 0.344
History of cigarette smoking 3 0.64 0.33 –1.25 0.193
NIHSS on admission 䡠䡠䡠 1.09 1.00 –1.20 0.058
Cardioembolic cause 䡠䡠䡠 4.13 1.23–13.83 0.022
Age 䡠䡠䡠 1.03 1.00 –1.06 0.022
*Rank of elimination in stepwise backward logistic regression.
ORs were derived from a backward logistic regression model. An OR ⬍1
indicated a lower likelihood of being on treatment for those with the indicated
Figure 3. Changes among subgroups of antithrombotic agents characteristic compared with those without the characteristic.
compared with discharge treatment (386 patients). n⫽386.
 1888 Stroke August 2001
reasons for treatment changes remained unknown in ⬎60%
of patients receiving thienopyridines, this assumption remains
speculative.
Another noteworthy result is the low nontreatment rate in
the anticoagulant subgroup (4.5% versus 14% in the anti-
platelet group). A higher awareness of risk for stroke recur-
rence by these patients and their general practitioners may
account for this result. Moreover, in contrast to treatment
with antiplatelet agents, anticoagulant therapy requires regu-
lar physician-patient contact. Compliance with antihyperten-
sive treatment was excellent in the present study, which
concurs with the results of other studies.23,24 Again, the
possibility of directly assessing treatment effects (ie, measur-
ing blood pressure) might account for patient adherence to the
recommended medication. In contrast, the high costs for
statins and the remaining uncertainty regarding their effi-
ciency in stroke prevention may have contributed to the
relatively low compliance rate (70%) in this subgroup.
In the present study, patients of older age, those with a
more severe neurological deficit, and a cardioembolic stroke
cause were more likely to adhere to the recommended
medication, whereas another study from Great Britain re-
ported that patients with severe strokes were less compliant at
3 months after stroke.15 This phenomenon was explained by
“possible difficulties experienced in attending hospital out-
patient clinics and general practitioner surgeries for monitor-
ing of the anticoagulation status or blood pressure.”15 In our
country, patient access to general practitioners and other
medical institutions for outpatients may be easier and thus
Downloaded from http://ahajournals.org by on November 11, 2019
facilitate poststroke medical care. Hence, differences in
medical health care systems among countries may account for
the diversity of these results. Moreover, patient-physician
contact may be closer in sicker patients, and patients with
more severe or cardioembolic strokes are more likely to have
a recurrence of stroke than are those with, for example,
lacunes.14,25,26 A British survey demonstrated that primary
care is mainly focused on the management of patients at high
absolute risk of stroke recurrence.27 As a result, monitoring
conditions in patients with higher treatment benefits may
account for better compliance rates.
To date, reasons for changes in discharge medication have
merely been reported for primary stroke prevention in pa-
tients with atrial fibrillation.28,29 These were similar to our
findings for secondary prevention that changes were based on
the general practitioner’s decision in most instances. Remark-
ably, in the present study, 12.2% of changes in antithrombotic
therapy but only 2.5% of changes in stroke risk factor therapy
were initiated by the patient. This might be due to side
effects; the lack of biochemical antiaggregation control,
resulting in fewer physician contacts; and a lack of motiva-
tion to continue medication in patients with minor or no
disabilities.30 Thus, our results stress the importance of
providing detailed and repeated information about the cause
and prognosis of stroke and the possibilities of reducing the
risk of recurrence to increase long-term patient compliance.
The present study has potential limitations. Patients who
were lost to follow-up or died were not evaluated for
compliance, which may have led to an overestimation of
compliance, because no follow-up data on the actual treat-
ment in these subgroups were available. Moreover, because
we gained information mainly from the patients via telephone
interviews, the results of the study are limited by the
credibility of patients. However, the statements about cessa-
tion of smoking are within the range reported from the
literature, suggesting that the information obtained was quite
reliable.14,23,24 Due to the study design, patients who were
treated insufficiently with regard to anticoagulation were not
identified. Because insufficiency of treatment is a common
phenomenon in clinical practice,21,23 our study does not allow
statements to be made on the quality of secondary stroke
prevention in general.
In conclusion, long-term compliance with preventive
stroke therapy is surprisingly good in the setting of the
present study. Higher age, a more severe neurological deficit,
and cardioembolic stroke cause predict higher compliance
rates, whereas treatment with thienopyridines decreases the
adherence to treatment recommendations. Consequently, spe-
cial attention should be paid to younger patients with less
severe deficits to improve their long-term compliance with
secondary stroke prevention.
References
1. Murray CJL, Lopez AD. Mortality by cause for eight regions of the
world: Global Burden of Disease Study. Lancet. 1997;349:1269 –1276.
2. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy
to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann
Intern Med. 1999;131:492–501.
3. Albers GW, Easton JD, Sacco RL, Teal P. Antithrombotic and
thrombolytic therapy for ischemic stroke. Chest 1998;114:683S– 698S.
4. Stroke prevention by the practitioner. Cerebrovasc Dis. 1999;9(suppl
4):1– 68.
5. Sacco RL, Wolf PA, Gorelick PB. Risk factors and their management for
stroke prevention: outlook for 1999 and beyond. Neurology. 1999;
53(suppl 4):15–24.
6. Gueyffier F, Boissel J-P, Boutitie F, Pocock S, Coope J, Cutler J, Ekbom
T, Fagard R, Friedman L, Kerlikowske K, Perry M, Prineas R, Schron E,
the INDANA Project Collaborators. Effect of antihypertensive treatment
in patients having already suffered from stroke. Stroke. 1997;28:
2557–2562.
7. Du X, Cruickshank K, McNamee R, Saraee M, Sourbutts J, Summers A,
Roberts N, Walton E, Holmes S. Case-control study of stroke and the
quality of hypertension control in North West England. BMJ. 1997;314:
272–276.
8. PATS Collaborating Group. Epidemiology Survey: Post-Stroke Antihy-
pertensive Treatment Study: a preliminary result. Chin Med J. 1995;108:
710 –717.
9. Wannamethee SG, Shaper AG, Whincup PH, Walker M. Smoking ces-
sation and the risk of stroke in middle-aged men. JAMA. 1995;274:
155–160.
10. Antiplatelet Trialists‘ Collaboration. Collaborative overview of ran-
domized trials of antiplatelet therapy, I: prevention of death, myocardial
infarction, and stroke by prolonged antiplatelet therapy in various cate-
gories of patients. BMJ. 1994;308:81–106.
11. WHO Task Force on Stroke and Other Cerebrovascular Disorders. Rec-
ommendations on stroke prevention, diagnosis, and therapy. Stroke 1989;
20:1407–1431.
12. Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL,
Marsh EE. Classification of subtype of acute ischemic stroke: definitions
for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute
Stroke Treatment. Stroke. 1993;24:35– 41.
13. Easton JD, Diener H-C, Bornstein NM, Einhäupl K, Gent M, Kaste M,
Sacco RL, Tijssen JGP, van Gijn J. Antiplatelet therapy: views from the
experts. Neurology. 1999;53(suppl 4):32–37.
14. Marini C, Totaro R, Carolei A. Long-term prognosis of cerebral ischemia
in young adults. Stroke 1999;30:2320 –2325.
15. Hillen T, Dundas R, Lawrence E, Stewart JA, Rudd A, Wolfe CDA.
Antithrombotic and antihypertensive management 3 months after ische-
mic stroke. Stroke. 2000;31:469 – 475.
 Sappok et al Compliance With Secondary Prevention of Ischemic Stroke
16. Masuhr F, Busch M, Einhäupl KM. Differences in medical and surgical
therapy for stroke prevention between leading experts in North America
and Western Europe. Stroke. 1998;29:339 –345.
17. Kalra L, Perez I, Melbourn A. Stroke risk management. Stroke. 1998;29:
53–57.
18. Van der Meulen JHP, Limburg M, von Straten A, Habbema JDF.
Computed tomographic brain scans and antiplatelet therapy after stroke:
a study of the quality of care in Dutch hospitals. Stroke. 1996;27:
633– 638.
19. O’Connell JE, Gray CS. Atrial fibrillation and stroke prevention in the
community. Age Aging. 1996;25:307–309.
20. Deplanque D, Corea F, Arquizan C, Parnetti L, Mas JL, Gallai V, Leys D,
and the SAFE I Study Investigators. Stroke and atrial fibrillation: is stroke
prevention treatment appropriate beforehand? Heart. 1999;82:563–569.
21. Brass LM, Krumholz HM, Scinto JM, Radford M. Warfarin use among
patients with atrial fibrillation. Stroke. 1997;28:2382–2389.
22. Bradley F, Morgan S, Smith H, Mant D. Preventive care for patients
following myocardial infarction. Fam Practice. 1997;14:220 –226.
23. Joseph LN, Babikian VL, Allen NC, Winter MR. Risk factor modification
in stroke prevention: the experience of a stroke clinic. Stroke. 1999;30:
16 –20.
Downloaded from http://ahajournals.org by on November 11, 2019
1889
24. Dachenhausen A, Brainin M, Steiner M. Modification of risk factors after
cerebral infarct: results of the Klosterneuburg Stroke Databank. Wien
Med Wochenschr. 1997;147:41– 45.
25. Anderson CS, Jamrozik KD, Broadhurst RJ, Stewart-Wynne EG. Pre-
dicting survival for 1 year among different subtypes of stroke. Stroke.
1994;25:1935–1944.
26. Moroney JT, Bagiella E, Paik MC, Sacco RL, Desmond DW. Risk factors
for early recurrence after ischemic stroke. Stroke. 1998;29:2118 –2124.
27. Coppola WGT, Whincup PH, Walker M, Ebrahim S. Identification and
management of stroke risk in older people: a national survey of current
practice in primary care. J Hum Hypertens. 1997;11:185–191.
28. Man-Son-Hing M, Laupacis A, O’Connor A, Wells G, Lemelin J, Wood
W, Dermer M. Warfarin for atrial fibrillation. Arch Intern Med. 1996;
156:1841–1848.
29. Mc Crory DC, Matchar DB, Samsa G, Sanders LL, Pritchett ELC.
Physician attitudes about anticoagulation for nonvalvular atrial fibril-
lation in the elderly. Arch Intern Med. 1995;155:277–281.
30. Hackett ML, Duncan JR, Anderson CS, Broad JB, Bonita R. Health-
related quality of life among long-term survivors of stroke: results from
the Auckland Stroke Study, 1991–1992. Stroke. 2000;31:440 – 447.