552
Stroke Incidence and Case
Fatality In Australasia
A Comparison of the Auckland and Perth
Population-Based Stroke Registers
Ruth Bonita, MPH, PhD; Craig S. Anderson, MBBS, BMedSc, FRACP;
Joanna B. Broad, BA; Konrad D. Jamrozik, MBBS, DPhil, FAFPHM;
Edward G. Stewart-Wynne, BM, ChB, FCP(SA), FRACP; Neil E. Anderson, MB, ChB, FRACP
Background and Purpose Population-based studies are cru-
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
cial for identifying explanations for the decline in mortality
from stroke and for generating strategies for public health
policy. However, they present particular methodological diffi-
culties, and comparability between them is generally poor. In
this article we compare the incidence and case fatality of
stroke as assessed by two independent well-designed incidence
studies.
Methods Two registers of acute cerebrovascular events
were compiled in the geographically defined metropolitan
areas of Auckland, New Zealand (population 945 369), during
1991-1992 for 12 months and Perth, Australia (population
138 708), during 1989-1990 for 18 months. The protocols for
each register included prospective ascertainment of cases
using multiple overlapping sources and the application of
standardized definitions and criteria for stroke and case
fatality.
Results In Auckland, 1803 events occurred in 1761 resi-
dents, 73% of which were first-ever strokes. The corresponding
figures for Perth were 536 events in 492 residents, 69% of
which were first-ever strokes. Both studies identified a sub-
A fter a long period of neglect, there is now fresh
/ \ interest in the epidemiology of stroke. In part
J. V . this reflects a desire to explain the sustained
decrease in mortality from stroke seen in many devel-
oped countries over recent decades as well as the
increases in several Eastern European countries.1 It also
reflects concern that rapid increases in both relative and
absolute numbers of the elderly will overwhelm eco-
nomic, medical, and social resources providing care for
the disabled aged, many of whom have experienced a
stroke.2 Despite advances in noninvasive neuroimaging
techniques, the recognition of various risk factors for
stroke, and improved epidemiologic methods, it is un-
Received September 30, 1993; final revision received November
29, 1993; accepted November 29, 1993.
From the Section of Geriatric Medicine, University of Auckland
(New Zealand) (R.B., J.B.B.); the Department of Medicine (Neu-
rology), Flinders Medical Centre, South Australia (C.S.A.); the
Department of Public Health, University of Western Australia,
Nedlands (K.DJ.); the Department of Neurology, Royal Perth
(Australia) Hospital (E.G.S.-W.); and the Department of Neurol-
ogy, Auckland (New Zealand) Hospital (N.E.A.).
Correspondence to Ruth Bonita, MPH PhD, Section of Geriat-
ric Medicine, University of Auckland, Private Bag 93-503, Auck-
land 9, New Zealand.
stantial proportion of nonfatal strokes managed solely outside
the hospital system: 28% in Auckland and 22% in Perth of all
patients registered. The age-standardized annual incidence of
stroke (all events) was 27% higher among men in Perth
compared with Auckland (odds ratio, 1.27; P=.O16); women
tended to have higher rates in Auckland, although these
differences were not statistically significant. In both centers
approximately a quarter of all patients died within the first
month after a stroke. There were significant differences in the
prevalence of hypertension among first-ever strokes.
Conclusions These two studies emphasize the importance
of identifying all patients with stroke, both hospitalized and
nonhospitalized, in order to measure the incidence of stroke
accurately. The incidence and case fatality of stroke were
remarkably similar in Auckland and Perth in the early 1990s.
However, there are differences in the sex-specific rates that
correspond to differences in the pattern of risk factors. (Stroke.
1994^5^52-557.)
Key Words • Australia • cerebrovascular disorders •
epidemiology • incidence • New Zealand
clear whether there has been a change in the incidence
of stroke or a change in natural history, with the average
stroke becoming less severe.3 However, in the absence
of definitive breakthroughs in the treatment of acute
stroke, it seems likely that much of the fall in mortality
from stroke can be attributed to changes in risk factors
rather than to improvements in the medical manage-
ment of stroke.
Largely because of methodological difficulties in mea-
suring the incidence of stroke, there is a scarcity of
well-designed epidemiologic studies of its trends and
determinants.4 Although hospital-based studies are a
popular source of data on stroke, they are complicated
by selection bias and therefore are of limited use either
for monitoring the effects of primary prevention or for
planning of health services. Population-based studies,
on the other hand, can provide the necessary data, but
only if they adhere to several "ideal" criteria for defi-
nitions, design, and presentation of data.4 However,
these studies are costly and demanding, particularly in
regard to identifying nonfatal events managed entirely
in the community. Clearly, incomplete case-finding
would lead to an underestimation of the true incidence
of stroke. In this article we present data from two stroke
registers that met all the criteria for well-designed
 Bonita et al Stroke in Australasia 553

incidence studies, with the important objective of com- TABLE 1. Comparison of Selected Baseline Variables
paring the incidence and case fatality of stroke between Among Stroke Events Registered In the Auckland
two defined Australasian populations. (1991) and Perth (1989-1990) Stroke Studies*

Auckland Perth
Subjects and Methods (N=18O3) (N=536)
Study Populations
The study populations were all residents of Auckland, New Variable n % n %
Zealand, and of a geographically defined segment of metro- Age, y
politan Perth, Western Australia. The population of Auckland
15-64 433 24.0 107 20.0
was 945 369 according to the 1991 Census. The study popula-
tion in Perth was estimated at 138 708 (at June 30, 1989) by 65-74 512 28.4 144 26.9
linear extrapolation of the relative change in the population of 75-84 611 33.9 203 37.8
the study area for the intercensal period 1981 through 1986.
The proportion of people 65 years and older was greater in 85 + 247 13.7 82 15.3
Perth (12.3%) than Auckland (9.4%), but the structure and Mean age, y
availability of hospital and other health services for stroke
share certain similarities. Both cities have experienced consid- Men 68 71
erable postwar immigration flows and now have distinctive Women 74 75
multicultural populations.
Total 72 73
Ascertainment of Stroke Events
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

Country of birth
The Auckland register operated for 12 months from March Australia/New Zealand 1239 68.7 286 53.4
1, 1991; the study period for the Perth register was 18 months
beginning February 20, 1989. Details of the case-finding United Kingdom 299 16.6 99 18.5
procedures adopted by each stroke register are described Pacific Islands 125 6.9 0 0.0
elsewhere.5-6 In brief, each study included a wide range of
overlapping sources, including daily checks of attendances at Europe 34 1.9 103 19.2
and admissions to all acute hospitals, routine searches for Other 106 5.8 48 9.0
in-hospital events, radiology requests for axial computed to-
mographic (CT) scans, perusal of death certificates and au- Married 875 48.5 260 48.5
topsy reports, and in the Perth study radiology requests for Residence
carotid duplex ultrasound. Hospital morbidity data were also
scrutinized for any mention of cerebrovascular disease (the Home 1512 83.9 435 81.2
International Classification of Diseases [9th revision, ICD-9] Hostel/lodge 103 5.7 44 8.2
codes 430-438) as an additional check for completeness of
ascertainment.7 Institution 188 10.4 57 10.6
Both studies paid particular attention to identifying nonfatal History of stroke 498 27.6 166 31.0
events occurring outside the hospital setting. To enlist the Management in hospital 1305 72.4 418 78.0
active support of the two communities, considerable emphasis
was given to widespread publicity. Before the Perth study *Missing and unknown values are not Included.
began, all medical practitioners serving the study population
were telephoned and sent a letter. The objectives of these
contacts were to clarify the aims of the study and invite each recurrent event during the study period was considered to have
doctor to refer all cases of suspected stroke or transient occurred if more than 28 days had elapsed from the initial
ischemic attack, irrespective of medical condition, to a central episode that brought the patient into the study. Total stroke
office. Access to this office was facilitated by the use of a events therefore refer to all events, rather than patients, and
24-hour answering machine. Regular follow-up by telephone include patients both with and without a history of stroke. A
and progress reports ensured continuing cooperation from case "managed in hospital" is one in which admission occurred
medical practitioners. Similarly, the Auckland study included within 28 days of stroke onset. Twenty-eight-day case fatality
regular phone calls to 170 general practitioners who repre- is defined as the proportion of all events that resulted in death
sented a 25% random sample of all primary care physicians in within 28 days of onset.
the city. In Auckland, domiciliary visits by physiotherapists
and speech therapists and by stroke field officers to stroke Study Variables
patients being cared for in the community were reviewed on a Once it was established that the patient met the clinical
monthly basis as an additional check to the referrals by the criteria for inclusion, a semistructured personal interview
participating doctors. covering the onset of the stroke and other medical and social
factors was completed as soon as possible after the event. In
Study Definitions the Perth study, a neurology registrar (C.S.A.) interviewed all
Stroke was defined according to the World Health Organi- patients, whereas in Auckland, trained nurses in liaison with a
zation definition as "rapidly developing signs of focal (or neurologist (N.E.A.) interviewed the patients. If the patient
global) disturbance of cerebral function lasting 24 hours or was dead or unable to communicate, an appointment was
longer, or leading to death, with no apparent cause other than made with a close relative or friend to whom the same
vascular."8 This definition includes spontaneous subarachnoid questions were directed. Both studies were also characterized
hemorrhage but excludes subdural and extradural hematomas by follow-up interviews at selected intervals in the year after
and transient ischemic attacks. First-ever strokes refer to the index event: 1, 6, and 12 months in the Auckland study and
patients who had no previous history of stroke; that is, they 4 and 12 months in the Perth study.
suffered their "first-ever-in-a-lifetime" stroke during the study With regard to risk factors, diabetes mellitus was accepted
period. Patients who gave a history of stroke before the study on the basis of either a history of that condition or when a
period were classified as having suffered a recurrent event only random blood glucose level was found to exceed 11.1 mmol/L
after careful review of all the available medical information. A poststroke. Smoking status was classified as "current smoker"
 554 Stroke Vol 25, No 3 March 1994

Men Women
rate per 100 000, log scale rate per 100 000, log scale
10000q 10000q

1000 = 1000

100: 100:

-4- " • '

10:
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

10:

+ P«rth -"t- Auckland Auckland -+• Ptrth

15-24 25-34 35-44 45-54 55-64 65-74 75-84 85* 15-24 25-34 35-44 45-54 55-64 65-74 75-84 85*
Age group Age group
Line graphs show age-specific annual stroke incidence (first-ever) rates (per 100 000 population) for men and women in Auckland
(1991) and Perth (1989-1990).

or "ex-smoker" (defined as a person who had given up Results

smoking for at least 12 months). "Ever smoker" comprises all
current smokers and ex-smokers. All patients were asked
about any history of hypertension, myocardial infarction, or
stroke.
Calculation of Age-Standardized Rates
In the Auckland study, the numerator for the cases includes
a weighting by a factor of four for the 25% of nonfatai episodes
of stroke treated at home (or in an institution) in patients
under the care of the randomly selected doctors. In the
absence of counts (for each age/sex group) of people unaf-
fected by a previous stroke, normal census data were used for
the denominators. Age-standardized rates were calculated
using the direct method and weights from Segi's "world"
population, 15 years and over.9 Rates are reported with 95%
confidence intervals (CI); in the Auckland study, this included
the sampling variance.5 Case fatality is presented for first-ever,
recurrent, and total events. The x2 test, unadjusted for sam-
pling variance, was used for comparisons of proportions.
TABLE 2. Age-Standardized Stroke Annual Incidence
(First-Ever) Rates* In Men and Women In Auckland
(1991) and Perth (1989-1990) Per 100 000 Population
Auckland Perth
(n=1305) (n=370) Rate RaUot
Men 153.3 (136-170) 191.5(158-225) 1.27(1 •05, 1.53)
Women 128.6(115-142) 108.3(86-131) 0.88(0 •72, 1.07)
Total 141.5 (131-152) 145.6(126-165) 1.05 (0 .92, 1.21)
Values in parentheses are 95% confidence intervals.
*Age standardized to the world population 15 years and over
by the direct method.
tMantel-Haenszel age-adjusted.
During the 12-month study period, 1803 events oc-
curred in 1735 Auckland residents: 835 (46%) men and
968 (54%) women. The proportion of first-ever strokes
was 72%. Almost 70% (1089 cases) were first recruited
to the study from prospective monitoring; admissions to
the hospital accounted for 60% and death records for
9%. An additional 21% were first recruited from a
general practice. The pathological basis of 26% of the
cases of stroke was confirmed by CT or magnetic
resonance imaging scan or by necropsy.
During the 18-month period of the Perth study, 536
events occurred in 492 patients, of which 69% were
first-ever strokes. These events occurred in 281 (52%)
men and 255 (48%) women. Eighty-five percent (452
cases) were first recruited to the study either from a
medical practitioner (75) or from prospective monitor-
ing of admissions and discharges from hospitals (377).
The pathological basis of 86% of the cases of stroke was
confirmed by CT or magnetic resonance imaging scan or
by necropsy.6 The incidence of defined pathological
subtypes of stroke in Perth appear in an earlier article.10
Table 1 shows the distribution of selected baseline
variables for the events registered in each of the two
studies. The age distribution was somewhat younger in
the Auckland study, in which 24% were aged under 65
years and only 47% over 75 years, compared with 20%
and 53%, respectively, for Perth. However, if Maoris
and Pacific Islanders are omitted, age distribution was
remarkably similar, with approximately one fifth of all
registered patients being less than 65 years of age and
just over half (51% in Auckland and 53% in Perth) aged
75 years or more. At the time of stroke, more than four
 Bonita et al Stroke in Australasia 555

TABLE 3. Comparison of 28-Day Case Fatality of Stroke by Age, Sex, Management, and Sequence In
Auckland (1991) and Perth (1989-1990), All Events

Auckland (N=1803) Perth (N=536)

n % (95% Cl) n % (95% Cl)

Age
<75 years 181/945 19 (16,22) 44/251 18 (12,23)
a75 years 254/858 30 (26,33) 84/285 29 (23,36)
Sex
Men 178/835 21 (18,24) 61/281 22 (16,27)
Women 257/968 27 (23,30) 67/255 26 (20,33)
Management
Hospital 321/1305 25 (22,27) 94/428 22 (17,26)
Non hospital 114/498 23 (17,27) 34/108 31 (21, 42)
Histcxy
First-ever stroke 305/1305 23 (21,26) 84/370 23 (18,28)
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

Recurrent stroke 130/498 26 (22,31) 44/166 27 (19,34)

Total 435/1803 24 (22,26) 128/536 24 (20, 28)

fifths of all patients registered in both studies were
living in a private residence, the remainder being in
institutional care; the proportions of patients married,
living alone, and managed in hospital were similar.
The age- and sex-specific annual incidence (first-ever)
rates for the two populations are presented in the
Figure. These increase rapidly with age and are consis-
tently and significantly higher in men than women in
both studies. Although not shown in the Figure, in each
age group the 95% Cl values overlap around the rates
(shown elsewhere56), suggesting no statistically signifi-
cant difference in age-specific rates between the two
populations. There was no significant difference be-
tween Auckland and Perth in overall annual age-stan-
dardized incidence rates (141.5 and 145.6 per 100 000,
respectively) (Table 2). However, age-standardized
rates in men in Perth were higher (191.5 per 100 000)
than in men in Auckland (153.3 per 100 000); this
difference was statistically significant (odds ratio, 1.27;
95% Cl, 1.05-1.53; P= .016). The age-standardized rates
in women were similar (128.6 and 108.3 per 100 000 in
Auckland and Perth, respectively).
Table 3 shows there was no significant difference
between the two populations in the 28-day case fatality
for stroke. In both studies, the overall case fatality was
24% and was higher in women and in older people (75
years and older).
Table 4 compares selected characteristics of first-ever
strokes for patients from the two populations. Among
the men who suffered a stroke, the proportion with
either a history of hypertension (P=.O14) or on antihy-
pertensive treatment (P=.O07) was significantly higher
in Perth. For women who suffered a stroke, a similar
difference was found in the proportion with a history of
hypertension (P<.0001) or on antihypertensive drugs
(i'=.011). Slight differences in data collection methods
used in the two studies are thought unlikely to account
for this difference.
The women in the Auckland study were more likely to
have smoked, whereas men in Auckland were more
likely to have a history of diabetes mellitus, although
these differences did not reach conventional levels of
significance.

TABLE 4. Prevalence of Selected Characteristics Between Auckland and Perth, By Sex, Flrst-Ever Events

Men Women

Auckland Perth Auckland Perth

(n=835), (n=281), (n=968), (n=255),
Variable % % P % % P
History of hypertension 48 58 .014 50 67 .0001
On treatment for hypertension 30 41 .007 34 44 .011
Ex-smoker 52* 49 NS 22t 17 NS
Current smoker 24* 28 NS 21t 17 NS
Ever smoker 76* 78 NS 43t 36 .072
History of myocardial infarction 19 23 NS 12 14 NS
History of diabetes mellitus 15 11 NS 14 15 NS
*Fifty-nlne men with missing data were excluded from analysis.
tForty-five women with missing data were excluded from analysis.
 556 Stroke Vol 25, No 3 March 1994
Discussion
Strict criteria have been established to ensure that
geographical and secular trends in the incidence of
stroke are not affected by changes in diagnostic fashion,
incomplete ascertainment of cases, or under-enumera-
tion of the denominator of reference.4 This article
presents data from two population-based studies that
adhered to these criteria and show that the total event
rates and case fatality of stroke display little geograph-
ical variation within Australasia. However, the data also
suggest there are different sex-specific rates between
Auckland and Perth, which may reflect differences in
the prevalence and management of factors such as
hypertension and smoking.
Selection bias due to misclassification or incomplete
ascertainment of cases, the two major pitfalls of popu-
lation-based studies, is unlikely to be a problem here.
Both registers were compiled prospectively using com-
prehensive sources and internationally accepted defini-
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
tions for stroke. Objective evidence for the pathological
basis of the stroke was obtained in a high proportion of
cases, particularly in Perth. Although the diagnosis of
stroke is essentially and certainly initially a clinical
diagnosis made on the characteristic profile of the
event, problems can occur among patients who present
with impairment and have no reliable witness of the
event or have symptoms and signs that rapidly resolve.
Ideally, therefore, studies should include diagnostic
assessment of patients by neurologists. This creates
obvious logistic and cost difficulties for registers cover-
ing entire populations such as that in Auckland. How-
ever, we have no reason to suspect diagnostic errors in
the Auckland study, because trained nurse interviewers
used standardized questionnaires, with all doubtful di-
agnoses reviewed by the principal investigator (R.B.)
and neurologist (N.E.A.).
Few community-based studies report in detail system-
atic efforts at identifying patients with acute stroke who
are not subsequently admitted to the hospital. The
proportion in this category varies widely around the
world11 and attests to the limitations of hospital-based
studies in measuring the true incidence of stroke. Al-
though close cooperation and ease of referral of pa-
tients from medical practitioners are essential in this
regard, additional methods of case-finding in the com-
munity, often overlapping with other sources, must be
used. Although we cannot rigorously exclude a degree
of under-ascertainment of events in either study consid-
ered here, the high degree of concordance in the 28-day
case fatality, one useful test for the completeness of
ascertainment, is an important index of the validity of
these data.
In these two studies we found that, had those patients
who died before reaching a hospital or who were
managed at home or in long-term care facilities been
omitted, the true incidence of stroke would have been
underestimated by between one fifth and one quarter.
These sources of error are particularly important in
studies that include elderly people because many are
already in institutional care.
Overall incidence and case fatality of stroke were
similar in these two populations, although the incidence
for men was higher in Perth than in Auckland. Because
the inequalities in rates are not likely to be explained by
incomplete case ascertainment or misdiagnosis of
stroke, one would expect different risk factor profiles
between men in Auckland and Perth. In fact the data do
show that men in the Perth study were more likely to
have a history of hypertension than men in the Auck-
land study, although there were no other significant
differences. For women, incidence rates appeared sim-
ilar despite lower rates of hypertension in the Auckland
women.
A difference in the risk of stroke among men is of
particular interest because the trend is opposite that
seen for coronary heart disease, in which mortality rates
are known to be lower among men in Perth compared
with Auckland.12 This supports data indicating that
different patterns of atherosclerosis, and presumably
also of risk factors, exist for cerebrovascular disease and
coronary heart disease.8'13 However, it also raises the
question of whether a difference between the two
centers in the natural history of coronary heart disease
or its management could have contributed to differences
in the incidence of stroke. It has been suggested that the
decline in mortality rates from coronary heart disease,
together with the aging of the population, may be one
explanation for the increase in the incidence of stroke in
Rochester, Minnesota, over the last decade.14 If correct,
this hypothesis has important implications for the future
prevention and management of stroke.
Although the importance of hypertension as a risk
factor for stroke (both ischemic and hemorrhagic) is
well established,15 controversy exists regarding the con-
tribution of the treatment of hypertension to the decline
in mortality from stroke.16 Despite concern that it is
difficult to extrapolate the results of clinical trials to the
wider community, epidemiologic studies have consis-
tently shown that most strokes occur in people in the
average range of blood pressure who have not been
exposed to large-scale detection and management pro-
grams.17 Cigarette smoking is recognized as an impor-
tant modifiable risk factor for stroke,18 but other aspects
of lifestyle, and particularly dietary factors, could be
responsible for such changes in a whole population.1319
Moreover, the large number of factors implicated in the
etiology of stroke suggests that a large proportion of
strokes are preventable.17
In summary, ongoing monitoring of disorders such as
stroke provides a measure of the effectiveness of pri-
mary and secondary efforts aimed at reducing the
burden of cerebrovascular disease. The methods for
such monitoring have been refined, as evidenced by the
ability to identify comparable groups of patients in two
independent populations. Both studies were character-
ized by a sizable proportion of nonfatal events managed
outside the hospital setting. The comparable overall and
hospital 28 -day case fatality also suggest similarities in
the natural history of cerebrovascular disease in the two
populations. Finally, the tendency toward a greater risk
of stroke among men in Perth supports public health
measures aimed at changes in lifestyle factors relating
to hypertension. Were they to be successful, such
changes might result in additional sizable reductions in
the absolute number of fatal and disabling strokes.
Acknowledgments
We thank the Health Research Council of New Zealand, the
National Health and Medical Research Council of Australia,

the National Heart Foundation of New Zealand, the Austra-
lian Brain Foundation, and the Medical Research Foundation
of Royal Perth Hospital for financial support. We are indebted
to numerous patients, doctors, and hospitals and grateful for
cooperation from the Health Department of Western Austra-
lia, the Auckland and Perth Coroners, the Registrars-General
in New Zealand and Western Australia, and the Perth Office
of the Australian Bureau of Statistics. Dr Bonita is the
Masonic Senior Research Fellow and Joanna Broad the Ma-
sonic Research Analyst in the Section of Geriatric Medicine,
School of Medicine, Auckland.
References
1. Bonita R, Stewart AW, Beaglehole R. International trends in
stroke mortality: 1970-1985. Stroke. 1990;21:989-992.
2. Heikkinen E. Health implications of population ageing in Europe.
World Health Slat Q. 1987;40:22-40.
3. Scheinberg P. Controversies in the management of cerebrovascular
disease. Neurology. 1988;38:1609-1616.
4. Malmgren R, Warlow C, Bamford J, Sandercock P. Geographical
and secular trends in stroke. Lancet. 1987;2:1196-1199.
5. Bonita R, Broad JB, Beaglehole R. Changes in stroke incidence
and case-fatality Auckland, New Zealand in 1981-1991. Lancet.
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017
1993;342:1470-1472.
6. Anderson CS, Jamrozik KD, Burvill PW, Chakera TMH, Johnson
GA, Stewart-Wynne EG. Ascertaining the true incidence of stroke:
experience from the Perth Community Stroke Study, 1989-1990.
MedJAust. 1993;158:80-84.
7. World Health Organization. International Classification of Diseases.
9th rev, Volume 1. Geneva, Switzerland: World Health Organi-
zation; 1977.
8. WHO MONICA Project principal investigators. The World Health
Organisation MONICA Project (monitoring trends and deter-
minants in cardiovascular disease): a major international collabo-
Bonita et al Stroke in Australasia 557
ration (prepared by H Tunstall-Pedoe). J Clin Epidemiol. 1988;41:
105-114.
9. Doll R, Muir C, Waterhouse J, eds. Cancer Incidence in Five
Continents. New York, NY: Springer-Verlag; 1970.
10. Anderson CS, Jamrozik KD, Burvill PW, Chakera TMH, Johnson
GA, Stewart-Wynne EG. Determining the incidence of different
sub-types of stroke: results from the Perth Community Stroke
Study, 1989-1990. MedJAust. 1993;158:85-89.
11. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Why are
patients admitted to hospital? experience of the Oxfordshire Com-
munity Stroke Project. BrMedJ. 1986;292:1369-1372.
12. Hobbs MST, Jamrozik KD, Hockey RL, Alexander H, Beaglehole
R, Dobson AJ, Heller RF, Jackson R, Stewart AW. Mortality from
coronary heart disease and incidence of acute myocardial
infarction in Auckland, Newcastle and Perth. Med J Aust. 1991;
155:436-442.
13. Marmot MG, Poulter NR. Primary prevention of stroke. Lancet.
1992;339:344-347.
14. Kuller LH. Incidence rates of stroke in the eighties: the end of the
decline? Stroke. 1989;20:841-843.
15. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J,
Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke
and coronary heart disease, Part I: prolonged differences in blood
pressure, prospective observational studies corrected for the
regression of dilution bias. Lancet. 1990;335:765-774.
16. Bonita R, Beaglehole R. Increased treatment of hypertension does
not explain the decline in stroke mortality in the United States,
1970-1980. Hypertension. 1989;13(suppl I):I-69-I-73.
17. Shaper AG, Phillips AN, Pocock SJ, Walker M, Macfarlane PW.
Risk factors for stroke in middle aged British men. Br Med J.
1991;302:llll-1115.
18. Shinton R, Beevers G. Meta-analysis of relation between cigarette
smoking and stroke. Br Med J. 1989;298:789-794.
19. Jamrozik K, Broadhurst RJ, Anderson CS, Stewart-Wynne EG.
The role of lifestyle factors in the etiology of stroke: a population-
based case-control study in Perth, Western Australia. Stroke. 1994;
25:51-59.
 Stroke incidence and case fatality in Australasia. A comparison of the Auckland and Perth
population-based stroke registers.
R Bonita, C S Anderson, J B Broad, K D Jamrozik, E G Stewart-Wynne and N E Anderson

Stroke. 1994;25:552-557
doi: 10.1161/01.STR.25.3.552
Downloaded from http://stroke.ahajournals.org/ by guest on December 24, 2017

Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1994 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/25/3/552

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click Request
Permissions in the middle column of the Web page under Services. Further information about this process is
available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/