Subtypes and case-fatality rates of stroke:

Subtypes and case-fatality rates of stroke:

Journal of Neurological Sciences 156 (1998) 220–226 Subtypes and case-fatality rates of stroke: A hospital-based stroke registry in Taiwan (SCAN-IV) ...

70KB Sizes 0 Downloads 49 Views

Journal of Neurological Sciences 156 (1998) 220–226

Subtypes and case-fatality rates of stroke: A hospital-based stroke registry in Taiwan (SCAN-IV) Jiann-Shing Jeng a , Ti-Kai Lee b ,c , Yang-Chyuan Chang a , Zei-Shung Huang b , Sien-Kiat Ng b , a ,c a, Rong-Chi Chen , Ping-Keung Yip * a

b

Department of Neurology, National Taiwan University Hospital, Taipei 100, Taiwan Department of Internal Medicine, National Taiwan University Hospital, Taipei 100, Taiwan c En Chu Kong Hospital, Taipei 100, Taiwan

Received 5 June 1997; received in revised form 4 September 1997; accepted 5 November 1997

Abstract Stroke data bank can afford important information regarding risk factors, pathogenesis, prognosis, etc. By means of hospital-based stroke registry, we investigated the risk factors and case-fatality rates in different types of stroke and transient ischemic attack (TIA) patients at the National Taiwan University Hospital in 1995. After excluding ineligible patients, 995 patients aged 1–98 years (575 men and 420 women) were recruited. Men predominated in all age groups for stroke and TIA in general except for cerebral hemorrhage (CH) in patients aged ,35 years and subarachnoid hemorrhage (SAH) in patients aged $45 years. Of these, 676 (67.9%), 41 (4.1%), 228 (22.9%) and 50 (5%) patients were classified in the categories of cerebral infarction (CI), TIA, CH and SAH, respectively. The CI / CH ratio was 2.96. Hypertension remained one of the most important risk factors for CI, CH and TIA patients. Severe extracranial carotid artery stenosis ($50%) was found in 12% of the CI patients and 27% of the TIA patients, but not found in the CH and SAH patients. Of these patients, the 30-day case-fatality rate was 10.9%, highest in SAH (30%), followed by CH (24.1%) and CI (5.6%). There were 41 in-hospital stroke patients who had significantly higher case-fatality rates than the other stroke patients (P,0.001 for all stroke, CI and CH patients by x 2 test). As compared to the previous stroke registries in Taiwan, there is a secular trend of increasing CI / CH ratios. These findings in Taiwan were compared with those in other populations, including other Asian, Caucasian and black populations. The CI / CH ratios in Asian populations, including Chinese, Japanese and Korean, were much lower than those in Caucasian and black populations. Dietary, environmental and genetic factors probably play important roles in these differences.  1998 Elsevier Science B.V. Keywords: Cerebrovascular disorders; Stroke classification; Stroke registry

1. Introduction Stroke data banks have proven to be valuable in assessment of many aspects of cerebrovascular disease (Mohr, 1986; Brainin, 1994). The systemic collection of data from each stroke patient can provide insights into the symptomatology, pathophysiology, etiology, and prognosis of stroke. Although population-based stroke data banks can give predefined clinical questions, such as incidence, prevalence, time trends, etc. (Brainin, 1994), they have *Corresponding author. Tel.: 1886 2 3970800 ext. 5338; Fax: 1886 2 3418395; e-mail: [email protected] 0022-510X / 98 / $19.00  1998 Elsevier Science B.V. All rights reserved. PII S0022-510X( 98 )00046-X

appeared to be less effective when employed to categorize the stroke subtypes reliably, evaluate the risk factor profiles extensively and assess the clinical details ubiquitously (Bogousslavsky et al., 1988; Ricci et al., 1994). Besides, brain imaging studies [computed tomography (CT) or magnetic resonance imaging (MRI)], ultrasonographic studies and delicate hematological survey are more feasible in hospitals than in the general population. Moreover, stroke occurring during hospitalization is only attainable in hospitals. In the past 17 years, there were some similar stroke registry studies in Taiwan (Ryu, 1986; Hung et al., 1987, 1987, 1992; Lie, 1989; Chang et al., 1995; Hsu et al.,

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

1995). However, most of these hospital-based studies recruited stroke patients from only one department of a hospital (Ryu, 1986; Lie, 1989; Hsu et al., 1995). Stroke patients may be missed if patients were admitted to the other departments, or the stroke occurred during hospitalization, or patients expired at the emergency room without further hospitalization. Therefore, we designed this study to determine each type of stroke and transient ischemic attack (TIA), and their risk factors and case-fatality rates in a hospital-based stroke registry. This study was a part of the Stroke and Cerebral Atherosclerosis Study of National Taiwan University Hospital (SCAN). The aims of the SCAN project are to assess the relation of various cardiovascular risk factors to cerebrovascular disease and carotid atherosclerosis (Chung et al., 1994; Jeng et al., 1994) and to probe the extent and severity of atherosclerosis in the carotid arteries (Chen et al., 1995; Jeng et al., 1995).

2. Patients and methods The National Taiwan University Hospital Stroke Registry is a hospital-based registry initiated in January 1995. The purposes are to study the etiologies, clinical course, prognosis and complications of stroke / TIA. Prospectively and retrospectively, we included and evaluated all patients who had presumed stroke or TIA before or during hospitalization in 1995. Not only patients from the neurological wards, but patients from the neurosurgical, internal medicine, pediatric and other wards were also included in the present study. We also reviewed all charts with International Classification of Disease, 9th Revision (ICD9) coded 430.0 to 437.9, 671.5 and 674.0. The diagnosis of stroke and TIA was based on the clinical features and the data collected by examinations such as brain imaging (CT / MRI), cardiac imaging (transthoracic and / or transesophageal echocardiography), ultrasonography of extracranial and / or intracranial arteries, angiography (MR or conventional angiography), and laboratory assessment of a prothrombotic state. Stroke was classified into the following categories: cerebral infarction (CI), cerebral hemorrhage (CH) and subarachnoid hemorrhage (SAH). The diagnosis of TIA was made when patients had transient focal neurologic dysfunction of ischemic origin that resolved within 24 h, leaving no deficit. The diagnosis of CI was made when patients had: (1) focal neurological deficits with sudden onset and lasting more than 24 h; and (2) a hypodense area on brain imaging corresponding to the clinical manifestations or no lesion seen on brain imaging. The diagnosis of CH was made when a brain parenchymal hemorrhage found on brain imaging corresponding to the clinical pictures. The diagnosis of SAH was made when patients had: (1) characteristic clinical manifestations of neurologic deficits, usually with sudden severe headache: and (2) either

221

subarachnoid blood on brain imaging or nontraumatic bloody (or xanthochromic) cerebrospinal fluid. Patients with both parenchymal and subarachnoid hemorrhage was classified into the CH group. Information on demographics, risk factors, clinical courses, complications and prognosis was also collected. Patients with old stroke and its complications, traumatic intracranial hemorrhage, subdural hematoma, no CT images, or subsequent testing to have noncerebrovascular causes for their presentations were excluded. The average time to death, cause of death, 30-day case-fatality rates and case-fatality rates during hospitalization were investigated in different types of stroke. The case-fatality rate of in-hospital stroke patients were compared to those of the other stroke patients.

3. Results A total of 1338 patients were assessed. According to the exclusion criteria, there were 270 patients excluded for the following reasons: old stroke without new insult (83), without head imaging studies (29), headache without evidence of stroke (23), brain tumors (19), dizziness / vertigo (19), subdural hematoma (16), seizures (14), metabolic encephalopathy (12), cardiogenic problems (8), Bell’s palsy (8), syncope (6), and others (33). After excluding 270 patients, 1068 patients (622 men and 446 women) were recruited. There were 772 patients admitted to different wards, 133 patients followed-up at our outpatient clinics after emergency room management, 42 patients transferred to other hospitals after emergency room management, 48 patients expired at our emergency room, 31 patients without outpatient clinic follow-up after emergency room management, and 42 patients having stroke or TIA during hospitalization. Excluding patients transferred to other hospitals and without further followup, 995 patients were analyzed. There were 575 men [mean age 62.0 years (S.D., 14.6), age range 1–94 years] and 420 women [mean age 63.0 years (S.D., 16.4), age range 1–98 years]. Of the 995 cases, patients with CI accounted for 67.9%, CH for 22.9%, SAH for 5%, and TIA for 4.1%. Table 1 shows the age distribution in males and females. Males predominated in all age groups of stroke and TIA in general except for CH in patients aged ,35 years and SAH in patients aged $45 years. The clinical characteristics and risk factors in different types of stroke and TIA are shown in Table 2. The average age for stroke / TIA occurrence was highest in CI (64.9 years; S.D., 13.8), followed by TIA (61.2 years; S.D., 10.7), CH (57.2 years; S.D., 18.2), and SAH (53.5 years; S.D., 11.9). Hypertension remained one of the most important risk factors, especially for CH (72%), TIA (71%) and CI (64%), and less for SAH (38%). The frequency of diabetes mellitus was higher in CI (31%) and

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

222

Table 1 Age and sex distribution of 995 patients with different types of cerebrovascular disease Age range (years) ,35 35–44 45–54 55–64 65–74 .74 Total

Sex

Stroke type CI

TIA

Total CH

SAH

Men 11 Women 11 Men 19 Women 15 Men 38 Women 28 Men 89 Women 66 Men 150 Women 98 Men 76 Women 75

0 0 3 1 5 1 7 6 10 4 1 1

11 15 18 5 25 7 45 14 32 23 17 16

3 3 5 2 4 6 4 12 0 7 0 4

25 29 45 23 72 42 145 98 192 132 94 96

Men 383 Women 293

28 13

148 80

16 34

575 420

CI, cerebral infarction; TIA, transient ischemic attack; CH, cerebral hemorrhage; SAH, subarachnoid hemorrhage.

TIA (20%) than in CH (14%) and SAH (14%). A history of prior stroke was conspicuous in CI (26%). Atrial fibrillation was identified in 12% of all patients and 16% of CI patients. Ischemic heart disease (history of ischemic heart disease or ischemic changes on electrocardiogram) was found in 24% of all patients. Left ventricular hypertrophy on electrocardiogram was found in 33% of patients, especially for CH (42%). The frequencies of cigarette smoking or alcohol drinking habits were higher in the CI, TIA and CH than those in the

SAH group. Hypercholesterolemia (serum level$240 mg / dl) was discovered in 126 patients (12.7%), highest in TIA (37%) and lowest in CH (4%). Hypertriglyceridemia (serum level$200 mg / dl) was found in 119 patients (12%). Of these patients, 599 had received extracranial duplex ultrasound examinations. Sixty-eight percent of the CI patients, 65% of the TIA patients, 57% of the CH patients and 30% of the SAH patients had extracranial carotid artery (ECCA) atherosclerosis. Severe ECCA atherosclerosis (luminal stenosis$50%) was found in 12% of the CI patients and 27% of the TIA patients, but not found in the CH and SAH patients. The case-fatality rates within 30 days after stroke and during hospitalization in different types of stroke and TIA are shown in Table 3. Overall, the 30-day case-fatality rate was 10.9% and the case-fatality rate during hospitalization was 13.2%. The 30-day case-fatality rates was highest in the SAH (30%), followed by CH (24.1%) and CI (5.6%). Most of the fatality occurred within 1 week of the stroke, especially for SAH and CH. The average time to death was shorter in CH and SAH than in CI. In analysis of the presumed causes of death, most of the CH and SAH patients were directly related to stroke. However, 42% of the expired CI patients were not directly related to stroke. Of the 41 in-hospital stroke patients, the 30-day casefatality rate was 36.6% (n515). Nine of 23 (39.1%) CI patients, 10 of 14 (71.4%) CH patients, and 2 of 4 (50%) SAH patients expired during hospitalization. As compared to the other stroke patients, in-hospital stroke patients had significantly higher case-fatality rates (P,0.001 for all stroke, CI and CH patients by x 2 test).

Table 2 Clinical characteristics of 995 patients with different types of cerebrovascular disease Characteristics

Cerebral infarction (n5676)

Transient ischemic attack (n541)

Cerebral hemorrhage (n5228)

Subarachnoid hemorrhage (n550)

Overall (n5995)

Men Age (years) Range Mean (S.D.) Hypertension Diabetes mellitus Prior stroke Atrial fibrillation Ischemic heart disease Left ventricular hypertrophy Smoking Alcohol drinking Serum cholesterol level$240 mg / dl Serum triglyceride level$200 mg / dl ECCA atherosclerosis a ECCA stenosis$50%

383 (57)

28 (68)

148 (65)

16 (32)

575 (58)

1–98 64.9 (13.8) 434 (64) 208 (31) 176 (26) 106 (16) 174 (26) 218 (32) 220 (33) 123 (18) 97 (14)

37–79 61.2 (10.7) 29 (71) 8 (20) 6 (15) 4 (10) 10 (24) 6 (15) 16 (39) 10 (24) 15 (37)

1–92 57.2 (18.2) 165 (72) 33 (14) 39 (17) 8 (4) 45 (20) 96 (42) 64 (28) 49 (21) 10 (4)

3–81 53.5 (11.9) 19 (38) 7 (14) 4 (8) 1 (2) 5 (10) 10 (20) 5 (10) 4 (8) 4 (8)

1–98 62.4 (15.4) 647 (65) 256 (26) 189 (19) 119 (12) 234 (24) 330 (33) 305 (31) 186 (19) 126 (13)

85 (13)

8 (20)

21 (9)

5 (10)

109 (11)

355 (68) 63 (12)

24 (65) 10 (27)

17 (57) 0

3 (30) 0

399 (67) 73 (12)

ECCA indicates extracranial carotid arteries. Values are number (percentage). a 599 patients receiving duplex ultrasonographic study.

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

223

Table 3 Case fatality ratios in each type of cerebrovascular disease in 1995 NTUH stroke registry Case fatality ratios

In-hospital 30-days Average time to death (days) Mean (S.D.) Median (25%, 75%) Presumed causes of death Directly related to stroke Indirectly related to stroke

Cerebral infarction (n5676)

Transient ischemic attack (n541)

Cerebral hemorrhage (n5228)

Subarachnoid hemorrhage (n550)

Overall (n5995)

53 (7.8) 38 (5.6)

0 0

62 (27.2) 55 (24.1)

16 (32) 15 (30)

131 (13.2) 108 (10.9)

38.6 (69.2) 7 (5, 36)

0 0

13.6 (40) 3 (2, 8.8)

12.3 (14.1) 7 (2.3, 18.5)

31 (4.6) 22 (3.3)

0 0

55 (24.1) 7 (3.1)

14 (28) 2 (4)

24.5 (53.2) 5 (3, 17) 100 (10.1) 31 (3.1)

Values are number (percentage) except average time to death.

4. Discussion Since the Harvard Cooperative Stroke Registry (Mohr et al., 1978) stimulating worldwide clinical stroke research, there have been many stroke data banks established for various purposes. Although there is selection bias (Ellenberg, 1994), a hospital-based stroke registry still has many advantages for clinical research. The stroke data bank can clarify the diagnosis, classify the diseases, assess the risk factors related to the etiology easily, observe the natural courses and disease outcome, and be used in the planning of clinical trials (Mohr et al., 1978; Bogousslavsky et al., 1988; Foulkes et al., 1988; Sacco et al., 1989; Bamford et al., 1990; Kay et al., 1992; Hier, 1995). The National Taiwan University Hospital is a tertiary referral center which may be biased toward selecting more serious patients. For minimizing the selection bias, we recruited cases not only from the neurological ward, but also from any patients who had stroke / TIA before being brought to our emergency room, and in-hospital stroke patients. In addition to patients with mild stroke or TIA who were only managed at our outpatient clinics, the present stroke registry included all the other patients ever treated at our hospital. Besides, all the patients’ records were reviewed by the authors, and by the same diagnostic criteria and structured form. Therefore, we have tried to reduce the bias due to selection and process of analysis. In the past two decades, several hospital-based stroke registry studies were carried out in Taiwan (Ryu, 1986; Hung et al., 1987, 1992; Hung and the Study Group on Stroke, 1987; Lie, 1989; Chang et al., 1995; Hsu et al., 1995). The first one was conducted by Ryu (1986) at Chang Gang Memorial Hospital in 1980. He made a descriptive analysis of 854 acute stroke patients with further classification. In the following years, changing pattern of CI / CH ratio from 1.5 in 1981–82 to 2.7 in 1987–88 was disclosed at the Cathay General Hospital (Taipei) (Lie, 1989). Thereafter, Hung et al. (1987), (1992) conducted some large retrospective and prospective collaborative studies at many hospitals during the period of

1984–1992. A hospital-based community study was also done in 1991 with the aim of exploring the stroke incidence in Ilan, the northern part of Taiwan (Chang et al., 1995). The second Chang Gang Stroke Registry was carried out in 1994–95 and a total of 885 hospitalized patients was recruited (Hsu et al., 1995). Of the abovementioned studies, accompanied with the present study, the relative frequencies of each type of stroke and CI / CH ratios are shown in Table 4. A trend of increasing CI / CH ratios was found in these studies. The secular trend of CI / CH ratios from a mortality analysis during 1974–88 in Taiwan also revealed a similar finding (Chang and Chen, 1993). These results of stroke registries in Taiwan were comparable with those in other Asian countries (Table 5). In Japan, the CI / CH ratio varied from 1.89 to 4.27 (Tanaka et al., 1981; Suzuki et al., 1987; Ueda et al., 1988). In Korea, the CI / CH ratio was 1.46 from 17 hospital registries (Lee et al., 1993). For the Chinese, the CI / CH ratios varied between 1.16 and 1.46 in mainland China (Li et al., 1985; Chen et al., 1992) were lower than those in Hong Kong, 1.64 and 2.52 (Huang et al., 1990; Kay et al., 1992) and in Taiwan, .2.0 in the 1990s studies (Table 4). As compared to the Asian populations, the Caucasian and black populations have much higher CI / CH ratios (Table 6). Most of these studies showed that the CI / CH. 5.0 in Caucasian [8.26 in Harvard Cooperative Stoke Registry (Mohr et al., 1978); 14.56 in South Alabama (Gross et al., 1984); 9.74 in Pilot SDB (Kunitz et al., 1984); 5.85 in NINCDS SDB (Foulkes et al., 1988); 6.72 in Lausanne Stroke Registry (Bogousslavsky et al., 1988), 10.1 in Lehigh Valley (Friday et al., 1989); 8.26 in Oxfordshire Community Stroke Project (Bamford et al., 1990); 5.63 in Perth Community (Anderson et al., 1993) and 12.86 in Framingham Study (Wolf, 1994)]. For black populations, the CI / CH ratios were relatively lower than those in Caucasian, but still higher than those in Asians [8.24 in South Alabama (Gross et al., 1984); 5.52 in Pilot SDB (Kunitz et al., 1984); 5.78 in NINCDS SDB (Foulkes et al., 1988); 6.35 in Lehigh Valley (Friday et al., 1989)

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

224

Table 4 Proportion of stroke subtypes and case fatality ratios in previous Taiwan stroke registries Study year

Study sites

Author published year

No. of patients

1980 1981–2 1985 1984–7 1987–8 1991 1987–92 1994–5 1995

CGMH CGH 26 hospitals 3 hospitals CGH Northern Ilan PSRST CGMH NTUH

Ryu, 1986 Lie, 1989 Hung et al., 1987 Hung et al., 1987 Lie, 1989 Chang et al., 1995 Hung et al., 1992 Hsu et al., 1995 Jeng et al., present study

854 817 7355 3982 901 258 21 631 885 995

Stroke subtypes (%) CI

CH

SAH

59.2 60 50.1 57.5 73.7 62.4 60.9 69.3 67.9

30.2 40 32.6 35.2 26.3 31 30.3 26.9 22.9

7.8 2 5.8 6.6 2 5.4 4.9 2 5

CI / CH ratios

30-days case fatality ratios (%)

1.96 1.5 1.54 1.63 2.8 2.01 2.01 2.58 2.96

CI

CH

SAH

Overall

8.9 2 8.1 6.7 2 2 8.2 2 5.6

43.3 2 30.7 22.4 2 2 27.1 2 24.1

32.8 2 29.6 26.6 2 2 35 2 30

22.3 2 16.8 13.7 2 2 15.4 2 10.9

CGMH, Chang Gung Memorial Hospital; CGH, Cathay General Hospital; PSRST, Prospective Survey and Registry of Stroke in Taiwan; NTUH, National Taiwan University Hospital; CI, cerebral infarction; CH, cerebral hemorrhage; SAH, subarachnoid hemorrhage.

Table 5 Proportion of stroke subtypes in Asian countries City, country

Shibata, Japan Akita, Japan Hisayama, Japan 17 hospitals in Korea Six cities, China Beijing, China Hong Kong Hong Kong 26 hospitals in Taiwan Taipei, Taiwan

Reference

Tanaka et al., 1981 Suzuki et al., 1987 Ueda et al., 1988 Lee et al., 1993 Li et al., 1985 Chen et al., 1992 Huang et al., 1990 Kay et al., 1992 Hung et al., 1992 Jeng et al., present study

No. of patients

Stroke subtypes (%) CI

CH

SAH

415 1765 292 3021 115 2593 540 777 21 631 995

57.8 55.9 71.6 48.1 51 56.1 50 68.4 60.9 67.9

23.4 29.5 16.8 32.9 44 38.5 30.6 27.1 30.3 22.9

7.5 14.6 8.9 18 2 2.6 3.7 2 4.9 5

CI / CH ratios 2.47 1.89 4.27 1.46 1.16 1.46 1.64 2.52 2.01 2.96

Table 6 Proportion of stroke subtypes in Caucasian and black populations Study

Reference

Harvard cooperative (n5694) South Alabama (n5160)

Mohr et al., 1978 Gross et al., 1984

Pilot SDB (n51158)

Kunitz et al., 1984

CHSPs (n54125)a NINCDS SDB (n51805)

Becker et al., 1986 Foulkes et al., 1988

Laussanne (n51000) Lehigh Valley (n52675)

Bogousslavsky et al., 1988 Friday et al., 1989

Oxfordshire community (n5675) Northern Manhattan (n51008)

Bamford et al., 1990 Sacco et al., 1991

Luis Vernaza Hospital (n5500) Perth community (n5370) Framingham (n5688) SCAN (n5995)

Del Brutto et al., 1993 Anderson et al., 1993 Wolf, 1994 Jeng et al., present study

Ethnicity

Caucasian Caucasian Black Caucasian Black Caucasian Black Caucasian Caucasian Black Caucasian White Hispanic Black Hispanic Caucasian Caucasian Chinese

Stroke subtypes (%) CI

CH

SAH

83.4 83 76.6 58.2 66.3 60 69.6 74.5 73.2 89.9 84.5 80.7 92 82.8 86.4 62.6 70 65.4 67.9

10.1 5.7 9.3 6 12 4.9 11.9 12.9 10.9 8.9 13.3 9.8 7 11 10.6 37.4 12.4 5.1 22.9

6.5 5.7 5.6 9.3 11.8 2.2 14.9 10.7 2 2 2 4.9 1 6.2 3 2 4.9 6.7 5

CI / CH ratios 8.26 14.56 8.24 9.74 5.52 12.24 5.85 5.78 6.72 10.1 6.35 8.26 13.14 7.53 8.15 1.67 5.63 12.86 2.96

SDB, stroke data bank; NINCDS, National Institute of Neurological and Communicative Disorders and Stroke; CHSPs, Community hospital-based programs; SCAN, stroke and cerebral atherosclerosis study at National Taiwan University Hospital. a Includes 15.7% of nonwhite patients.

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

and 8.15 in Northern Manhattan (Sacco et al., 1991)]. For Hispanics, there were two inconsistent study results. The CI / CH ratio was 7.53 in Northern Manhattan study, considerably higher than one study from Ecuador, 1.67 (Del Brutto et al., 1993). The reasons for the significant race–ethnic differences of CI / CH ratios are poorly understood. In a review of stroke incidence in the People’s Republic of China, the excess stroke rate in the PRC over the Caucasian population is largely attributed to CH and some hypotheses have been proposed to explain these East–West racial differences (Shi et al., 1989). Hypertension, genetic, environmental and dietary factors may account for the high percentage of CH occurrence in the Asian populations. The Chinese and Japanese diets are generally higher in carbohydrates and salts and lower in fats than the Western diets. The inverse association between baseline serum total cholesterol level was documented in a 6-year follow-up study of 350 977 US men (Iso et al., 1989). One Japanese population study inferred that the significant and consistent fall in the incidence of CH in Japan could be attributed to the combination of an elevation of total cholesterol and a reduction of blood pressure (Shimamoto et al., 1989). Another study also disclosed that declining trends in blood pressure level in Japan paralleled the decrease in salt consumption (Ueshima et al., 1987). Furthermore, effective control of hypertension in recent decades have largely contributed to the decreasing incidence and mortality of CH both in the East and the West (Furlan et al., 1979; Garraway and Whisnant, 1987; Ueda et al., 1988). A trend of increasing CI / CH ratios in Taiwan, probably reflecting decrease in CH occurrence, may be explained by dietary changes and hypertension control. In addition, the prevalence of hypertension in Asian and Caucasian populations does not differ significantly. The predisposing factors to CH for the Asian people deserves further investigations. In the present study, the 30-day case-fatality rate was 10.8% and the case-fatality rate during hospitalization was 13.1%. The 30-day case-fatality rates was highest in the SAH (28%), followed by the CH (24.1%) and the CI (5.6%). As compared with the previous stroke registries in Taiwan, the overall 30-day case-fatality rates was decreasing gradually (Table 4). The reasons of decreasing stroke fatality may be largely due to increasing CI / CH ratios and reduction in CH fatality. These results were consistent with the Taiwan vital statistics data in 1974–1988 (Chang and Chen, 1993). We found that in-hospital stroke patients had higher case-fatality rate than the other stroke patients had. Of these 41 in-hospital stroke patients, 30-day case-fatality rate was 36.6% which was similar to that of one Italian in-hospital stroke study (Azzimondi et al., 1994). We supposed that the higher in-hospital stroke case-fatality rate was partly explained by: (1) a higher percentage of massive lobar CH due to coagulopathy after chemotherapy or anticoagulation therapy; (2) a higher percentage of

225

cardiac-related stroke; and (3) probable underestimation of the number of mild strokes in this study. In conclusion, the present stroke registry study has disclosed that there were some differences between similar studies in Taiwan and a temporal change of different types of stroke existed. The significant East–West differences in the proportions of different stroke types deserve further investigations.

References Anderson, C.S., Jamrozik, K.D., Burvill, P.W., Chakera, T.M.H., Johnson, G.A., Stewart-Wynne, E.G., 1993. Determining the incidence of different subtypes of stroke: results from the Perth Community Stroke Study, 1989–1990. Med. J. Aust. 158, 85–89. Azzimondi, G., Nonino, F., Fiorani, L., Vignatelli, L., Stracciari, A., Pazzaglia, P., D’Alessandro, R., 1994. Incidence of stroke among inpatients in a large Italian hospital. Stroke 25, 1752–1754. Bamford, J., Sandercock, P., Dennis, M., Warlow, C., 1990. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire Community Stroke Project 1981–6, 2. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid hemorrhage. J. Neurol. Neurosurg. Psychiatry 53, 16–22. Becker, C., Howard, G., McLeroy, K.R., Yatsu, F.M., Toole, J.F., Coull, B., Feibel, J., Walker, M.D., 1986. Community hospital-based stroke programs: North Carolina, Oregon, and New York. II. Description of study population. Stroke 17, 285–293. Bogousslavsky, J., Melle, G.V., Regli, F., 1988. The Lausanne Stroke Registry: Analysis of 1,000 consecutive patients with first stroke. Stroke 19, 1083–1092. Brainin, M., 1994. Overview of stroke data banks. Neuroepidemiology 13, 250–258. Chang, C.C., Chen, C.J., 1993. Secular trend of mortality from cerebral infarction and cerebral hemorrhage in Taiwan, 1974–1988. Stroke 24, 212–218. Chang, S.F., Su, C.L., Chen, Z.Y., Hung, T.P., 1995. Stroke incidence in Ilan, Taiwan. J. Formos. Med. Assoc. 94, 30–36. Chen, C.C., Chung, M.Y., Jeng, J.S., Yip, P.K., Hwang, B.S., Chang, Y.C., Chen, R.C., 1995. A scoring system for evaluation of the extent of extracranial carotid atherosclerosis with B-mode imaging. Acta Neurol. Sin. 4, 29–33. Chen, D., Roman, G.C., Wu, G.X., Wu, Z.S., Yao, C.H., Zhang, M., Hirsch, R.P., 1992. Stroke in China (Sino-MONICA-Beijing study) 1984–1986. Neuroepidemiology 11, 15–23. Chung, M.Y., Jeng, J.S., Yip, P.K., Hsu, H.C., Lee, Y.T., 1994. Study on blood lipids, lipoproteins and apolipoproteins in different types of ischemic stroke. Acta Neurol. Sin. 3, 194–199. Del Brutto, O.H., Mosquera, A., Sanchez, X., Santos, J., Noboa, C.A., 1993. Stroke subtypes among Hispanics living in Guayaquil, Ecuador. Results from the Luis Vernaza Hospital Stroke Registry. Stroke 24, 1833–1836. Ellenberg, J.H., 1994. Observational data bases in neurological disorders: selection bias and generalization of results. Neuroepidemiology 13, 268–274. Foulkes, M.A., Wolf, P.A., Price, T.R., Mohr, J.P., Hier, D.B., 1988. The Stroke Data Bank: design, methods, and baseline characteristics. Stroke 19, 547–554. Friday, G., Lai, S.M., Alter, M., Sobel, E., La Rue, L., Gil-Peralta, A., McCoy, R.L., Levitt, L.P., Isack, T., 1989. Stroke in the Lehigh Valley: racial / ethnic differences. Neurology 39, 1165–1168. Furlan, A.J., Whisnant, J.P., Elveback, L.R., 1979. The decreasing incidence of primary intracerebral hemorrhage: a population study. Ann. Neurol. 5, 367–373.

226

J.-S. Jeng et al. / Journal of Neurological Sciences 156 (1998) 220 – 226

Garraway, W.M., Whisnant, J.P., 1987. The changing pattern of hypertension and the declining incidence of stroke. J. Am. Med. Assoc. 258, 214–217. Gross, C.R., Kase, C.S., Mohr, J.P., Cunningham, S.C., Baker, W.E., 1984. Stroke in South Alabama: incidence and diagnostic features – a population based study. Stroke 15, 249–255. Hier, D.B., 1995. Stroke data banks and stroke registries. In: Caplan, L.R. (Eds.), Brain ischemia: Basic Concepts and Clinical Relevance. Springer-Verlag, London, pp. 355–363. Hsu, W.C., Chen, S.T., Chang, H.S., Hsieh, F.Y., Wu, L.L., 1995. The Chang Gung Stroke Registry: 1994. Acta Neurol. Sin. 4, 171–179. Huang, C.Y., Chan, F.L., Yu, Y.L., Woo, E., Chin, D., 1990. Cerebrovascular disease in Hong Kong Chinese. Stroke 21, 230–235. Hung, T.P., Study Group on Stroke, 1987. Epidemiology and Clinical Profiles of Stroke in Taiwan: Report of a Stroke Registry of 26 Teaching Hospitals in Taiwan, 1985. Taipei: Department of Health, R.O.C., 1–107 (In Chinese with English abstract). Hung, T.P., Chen, S.T., Lie, S.K., 1987. Final Report of Collaborative Study Group on Stroke in Taipei Region, 1984–1987. Taipei: Department of Health, R.O.C. (In Chinese with English abstract). Hung, T.P., Study Group on Stroke, 1992. Prospective Survey and Registry of Stroke in Taiwan Area, Taipei: Department of Health, R.O.C. (In Chinese with English abstract). Iso, H., Jacobs, D.R., Wenworth, D., Neaton, J.D., Cohen, J.D., 1989. Serum cholesterol levels and six-year mortality from stroke in 350 977 men screened for the Multiple Risk Factor Intervention Trial. New Engl. J. Med. 320, 904–920. Jeng, J.S., Chung, M.Y., Yip, P.K., Huang, B.S., Chang, Y.C., 1994. Extracranial carotid atherosclerosis and vascular risk factors in different types of ischemic stroke in Taiwan. Stroke 25, 1989–1993. Jeng, J.S., Yip, P.K., Chung, M.Y., Chen, C.C., Hwang, B.S., Lin, W.H., Chang, Y.C., Chen, R.C., 1995. Extracranial carotid artery atherosclerosis in 1014 physical check-up subjects (SCAN-III). J. Med. Ultrasound 3, 144–150. Kay, R., Woo, J., Kreel, L., Wong, W.Y., Teoh, R., Nicholls, M.G., 1992. Stroke subtypes among Chinese living in Hong Kong: The Shatin Stroke Registry. Neurology 42, 985–987. Kunitz, S.C., Gross, C.R., Heyman, A., Kase, C.S., Mohr, J.P., Price, T.R., Wolf, P.A., 1984. The Pilot Stroke Data Bank: definition, design, and data. Stroke 15, 740–746. Lee, S.B., Roh, J.K., Yoon, B.W., Hong, S.B., Lee, J.H., Korean Neurological Association, 1993. Epidemiology of cerebrovascular disease in Korea: a collaborative study, 1989–1990. J. Korean Med. Sci., 8, 281–289. Li, S.C., Schoenberg, B.S., Wang, C.C., Cheng, X.M., Bolis, C.L., Wang, K.J., 1985. Cerebrovascular disease in the People’s Republic of China: epidemiologic and clinical features. Neurology 35, 1708–1713.

Lie, S.K., 1989. Changes in the pattern of cerebral hemorrhage: 8-year experience at Cathay General Hospital. J. Formos. Med. Assoc. 88 (Suppl. 5), 546–549 (In Chinese). Mohr, J.P., Caplan, L.R., Melski, J.W., Goldstein, R., Duncan, G.W., Kistler, J.P., Pessin, M.S., Bleich, H.L., 1978. The Harvard Cooperative Stroke Registry: a prospective registry. Neurology 28, 754– 762. Mohr, J.P., 1986. Stroke data banks (editorial). Stroke 17, 171–172. Ricci, S., Celani, M.G., Righetti, E., 1994. Clinical methods for diagnostic confirmation of stroke subtypes. Neuroepidemiology 13, 290–295. Ryu, S.J., 1986. The Chang Gung Stroke Registry: 1980. Chang Gung Med. J. 9, 204–216. Sacco, R.L., Ellenberg, J.H., Mohr, J.P., Tatemichi, T.K., Hier, D.B., Price, T.R., Wolf, P.A., 1989. Infarcts of undetermined cause: the NINCDS Stroke Data Bank. Ann. Neurol. 25, 382–390. Sacco, R.L., Hauser, W.A., Mohr, J.P., 1991. Hospitalized stroke in blacks and Hispanics in Northern Manhattan. Stroke 22, 1491–1496. Shi, F., Hart, R.G., Sherman, D.G., Tegeler, C.H., 1989. Stroke in the People’s Republic of China. Stroke 20, 1581–1585. Shimamoto, T., Komachi, Y., Inada, H., Doi, M., Iso, H., Sato, S., Kitamura, A., Iida, M., Konishi, M., Nakanishi, N., Terao, A., Naito, Y., Kojima, S., 1989. Trends for coronary heart disease and stroke and their risk factors in Japan. Circulation 79, 503–515. Suzuki, K., Kutsuzawa, T., Takita, K., Ito, M., Sakamoto, T., Hirayama, A., Ito, T., Ishida, T., Ooishi, H., Kawakami, K., Hirota, K., Ogasawara, T., Yoshida, J., Tamura, T., Hattori, S., Iwabuchi, S., Karouji, Y., Waga, T., Oosato, Y., Yazaki, K., Saito, T., Oouchi, T., Kojima, S., 1987. Clinico-epidemiologic study of stroke in Akita. Japan. Stroke 16, 402–406. Tanaka, H., Ueda, Y., Date, C., Baba, T., Yamashita, H., Hayashi, M., Shoji, H., Owada, K., Baba, K.I., Shibuya, M., Kon, T., Detels, R., 1981. Incidence of stroke in Shibata, Japan, 1976–1978. Stroke 12, 460–466. Ueda, K., Hasuo, Y., Kiyohara, Y., Wada, J., Kawano, H., Kato, I., Fujii, I., Yanai, T., Omae, T., Fujishima, M., 1988. Intracerebral hemorrhage in a Japanese Community, Hisayama: Incidence, changing pattern during long-term follow-up, and related factors. Stroke 19, 48–52. Ueshima, H., Tatara, K., Asakura, S., Okamoto, M., 1987. Declining trends in blood pressure level and the prevalence of hypertension, and changes in related factors in Japan, 1956–1980. J. Chron. Dis. 40, 137–147. Wolf, P.A., 1994. Epidemiology of intracerebral hemorrhage. In: Kase, C.S., L.R. Caplan (Eds.), Intracerebral Hemorrhage. ButterworthHeinemann, Newton, pp. 21–30.