Stroke incidence and case fatality in Shiga, Japan

Transcription

1 International Epidemiological Association 1999 Printed in Great Britain International Journal of Epidemiology 1999;28: Stroke incidence and case fatality in Shiga, Japan Yoshikuni Kita, a Akira Okayama, a Hirotsugu Ueshima, a Mitsumasa Wada, a Akihiko Nozaki, a Sohel Reza Choudhury, a Ruth Bonita, b Yasuhiko Inamoto c and Takayuki Kasamatsu d Background This paper describes incidence rates and case-fatality for sub-types of stroke using data collected in Takashima, Shiga, Japan, from 1989 to 1993 and compares these with similar registers in other parts of Japan. Methods Registered patients included all residents of the county who experienced a firstever stroke. Stroke was defined as sudden onset of neurological symptoms which continued for a minimum of 24 hours or led to death. Almost all such patients are hospitalized in this country. Early case fatality was defined as patients who died within 28 days of stroke onset. Diagnosis of stroke type was based on clinical symptoms as well as computed tomography (CT) scans. Results Age-adjusted incidence rates for stroke per population aged 35 years and older were for men and for women. The age-specific incidence rate of both cerebral infarction and cerebral haemorrhage increased with advancing age. The occurrence of cerebral infarction in men was twice as high as in women. The 28-day case fatality for all sub-types of stroke was 16.1% in men and 15.8% in women. Case fatality for cerebral infarction, cerebral haemorrhage, and subarachnoid haemorrhage was 10.7%, 22.4% and 28.6% respectively. Conclusions Takashima County has a moderately high stroke incidence rate and case fatality compared with other similar studies in Japan. The incidence rate of cerebral infarction in men is twice that in women, while other sub-types of stroke showed smaller differences. In order to decrease the incidence of stroke in Japan, greater efforts at primary prevention will be necessary, in particular, it is important to prevent cerebral infarction in men. Keywords Stroke, incidence, case fatality, subtype of stroke, Japan Accepted 29 June 1999 In Japan, both the crude and age-adjusted mortality rates from cerebrovascular disease (stroke) have been declining rapidly. 1 3 This has contributed to the increased longevity of the Japanese population, which is now the highest in the world. Stroke as a single cause of death, however, is still the third leading cause of death in Japan. The decreasing stroke mortality rate could be a result of a decline in incidence or a change in the natural history of the disease with stroke becoming less severe. Several studies have investigated trends in the incidence of stroke in Japan. 4 9 However, few are community-based studies and many have only a small number of annual events making it difficult to a Department of Health Science, Shiga University of Medical Science, Tsukinowa-cho, Seta, Otsu, Shiga, Zip , Japan. kita@sums. shig-med.ac.jp b Faculty of Medicine and Health Science, University of Auckland, Auckland, New Zealand. c Ohmiairinen Foundation Imazu Hospital, Imazu-cho, Shiga, Japan. d Makino Hospital, Makino-cho, Shiga, Japan. determine whether a decline in incidence or improvement of stroke prognosis has contributed to the decline in stroke mortality. Stroke is also an important cause of disability. Because of the ageing of the population, an increasing prevalence of stroke resulting from improved case fatality will lead to a greater demand for support of patients with ongoing disability. To prevent stroke and to provide the social support systems for stroke patients, it is important to understand current incidence rates and also the reasons for the decline in stroke mortality. Ongoing stroke registration was established in Takashima County in Shiga Prefecture, Japan, in 1988 in order to measure trends in incidence and case-fatality of stroke. All patients admitted to two community hospitals in Takashima County and three tertiary hospitals outside the county with a first-ever stroke have been registered since 1988 and have been followed up to investigate their vital and functional prognoses post stroke for planning purposes and to predict survival. This article describes the methodology of this stroke registration system, the 1059

2 1060 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY incidence rate and the case-fatality for men and women and sub-types of strokes using data collected for the 5-year period 1989 to Material and Methods Geographical conditions of investigated area Shiga prefecture borders on Kyoto municipal area to the northeast, and is located at about the geographical centre of Japan. Lake Biwa is located to the east of Takashima County and the Hira-san Ganges runs north south to the west. Only one primary road runs through the area and movement is therefore restricted to this extent. Takashima County has a stable population of around with 17.6% of the population (around ) aged 65 years or over. Case finding and diagnostic criteria Takashima County contains two community hospitals, one in the south which is a public facility providing 261 beds, and another in the north which is a private facility providing 72 beds. It has been estimated that approximately 98% of all hospital admissions are seen at the two community hospitals. Registered patients included all residents of the county who were hospitalized with a first-ever stroke in the two community hospitals and one geriatric hospital which was the only care facility for elderly people in Takashima. The remaining patients are seen at three tertiary hospitals outside the county which have more sophisticated facilities for advanced treatment. Internist and specialist investigative personnel trained by neurologists carried out both the case finding and registration of patients who met the criteria. Registration procedures were investigated once every 3 months at the six facilities. Stroke diagnostic criteria employed in this study are the criteria established for the Monitoring System for Cardiovascular Disease commissioned by the Ministry of Health and Welfare. 10 These criteria are in accord with the World Health Organization 11 and define a stroke as sudden onset of neurological symptoms, which continue for a minimum of 24 hours or result in death. Transient ischaemic attack (TIA) was not registered under this system. Early case fatality was defined as patients who died within 28 days of the onset of the stroke event. Diagnosis of stroke type was based on clinical symptoms as well as computed tomography (CT) scans. A cerebral infarction was defined as a region of low-density absorption by the CT scan within a few days of the attack. An intracerebral haemorrhage was defined on the basis of a region of high-density absorption in the causal region due to a haematoma as shown on the CT scan. A subarachnoid haemorrhage was defined on the basis of a region of high-density absorption in the cerebrospinal fluid due to a haemorrhage as shown on the CT scan. Patients who satisfied clinical symptoms of acute stroke but type of stroke could not be determined from clinical signs and/or CT scan were classified as unclassified stroke. Cerebral infarction was classified as either cerebral thrombosis or cerebral embolism. Patients were diagnosed with cerebral thrombosis if, in addition to CT scan observations, they exhibited gradual (within a few days) and localized progression of neurological symptoms, underlying disorders (hypertension, diabetes, abnormal lipid metabolism) accompanied by arteriosclerosis. Patients were diagnosed with a brain embolism if, in addition to CT scan observations, they exhibited localized progression of neurological symptoms, or underlying disorder such as arrhythmia, valvular disease, myocardial infarction or occlusive pathological findings by angiography. Items recorded at registration of a stroke were the date and time of stroke event, the situation and symptoms at the event, the extent of neurological symptoms at the event, clinical observations at the event (blood pressure, presence of fibrillation, level of consciousness, impairment of neurological function), past history, family history, smoking history, consumption of alcohol, early (within one week) rehabilitation, fatality (within 28 days), cause of death, recurrence in acute stage, and CT scan observations. Past history included other disorders as well as those first diagnosed at the time of the event. On admission to hospital patients were also examined for abnormal lipid metabolism and kidney function disorders simultaneously with the diagnosis of the event. Abnormal lipid metabolism was defined as a serum total cholesterol level of at least 220 mg/dl. A serum total creatinine level of at least 2.0 mg/dl over two samples taken at intervals was used as the criteria for diagnosis of kidney function disorders. Items investigated in CT scan were the size of regions of lowdensity absorption in association with cerebral infarction, and the size of regions of high-density adsorption in association with cerebral haemorrhage. Cerebral angiography was employed in the investigation of cerebral aneurysms and anomalies in the cerebral venous system, and secondary cerebral infarctions in association with subarachnoid haemorrhage. Registered stroke patients were monitored annually by death certifications. Death certificates were seen with the approval of the General Affairs Office of Japan in order to establish cause of death. Data analysis Age-specific rates were calculated in 10-year age groups (35 44, 45 54, 55 64, 65 74, and 85 ) as described below. Three patients less than 35 years (two men and one woman) were registered during the study period, but have been excluded from these analyses because of small numbers. The age-specific incidence rates were determined by dividing the average annual number of cases over the 5-year period by the population of the relevant age group in Takashima County based on the 1990 census. 12 Age-adjusted incidence rates were calculated using the Japanese population aged 35 years and older from the 1980 census 13 as the standard population to allow comparison with other published Japanese studies. For age-specific incidence rates, age-adjusted incidence rates, and age-specific case fatality, 95% confidence intervals (95% CI) were calculated. 14 SPSS 15 was used for these statistical analyses. Results A total of 401 patients (217 in men and 184 in women) were registered as first-ever stroke cases in the 5-year period between 1 January 1989 and 31 December Among these patients, 373 (93.0%) underwent CT scans. The remaining 28 patients who did not receive CT scan included 10 patients who met the clinical criteria for acute stroke but who died within 24 hours of stroke onset.

3 STROKE INCIDENCE RATE IN TAKASHIMA, JAPAN 1061 Table 1 Age-specific and age-adjusted annual stroke incidence (per population) of stroke sub-type for men and women by age group Takasima, Shiga Prefecture, Cerebral Cerebral Subarachnoid Population no. infarction haemorrhage haemorrhage Total cases b Age group of Takashima n Rate 95% CI a n Rate 95% CI n Rate 95% CI n Rate 95% CI Men Total Age-adjusted rate c Women Total Age-adjusted rate Total Total Age-adjusted rate a 95% confidence interval. b Six cases which could not been classified have been included in Total cases. c Adjusted to 1980 Japanese population aged 35 years and over. Average age at the time of the stroke event for the overall stroke patients was 69.1 (median: 69.7) years in men and 72.6 (median: 74.0) years in women. The average age for patients with cerebral infarction was 70.6 (median: 71.0) years in men and 73.4 (median: 74.4) years in women, for cerebral haemorrhage 66.9 (median: 64.3) years in men and 72.6 (median: 74.6) years in women, and for subarachnoid haemorrhage 58.6 (median: 63.1) years in men and 68.0 (median: 68.4) years in women. Table 1 shows age-specific rates and age-adjusted incidence rates for all strokes and for all sub-types of stroke. Age-adjusted incidence rates for stroke per population aged 35 years and older was (95% CI : ) for men and (95% CI : ) for women, and the rate in men was significantly higher (1.6 times) than in women. The age-specific incidence rate of both cerebral infarction and cerebral haemorrhage increased with advancing age. The age-specific incidence rates of cerebral infarction were higher in men than in women for all age groups. Ratios of age-adjusted incidence rate of cerebral infarction versus cerebral haemorrhage were 3.3:1.0 for men, 2.0:1.0 for women and 2.6:1.0 for men and women together. The occurrence of cerebral infarction in men was twice that in women, especially in the age groups under 64 years old. There were no differences in the incidence of cerebral haemorrhage or subarachnoid haemorrhage between men and women. Case fatality at 28 days is shown in Table 2 for the three subtypes of stroke. Overall, 64 people (35 men and 29 women) died within 28 days. The average age for fatal cases was 71.8 (median: 71.2) years in men and 74.3 (median: 77.9) years in women. Overall (men and women combined) case fatality rate for all sub-types of stroke was 16.0% (95% CI : ). Case fatality for all sub-types of stroke was 16.1% (95% CI : ) in men and 15.8% (95% CI : ) in women. Agespecific case-fatality rates were higher in men than in women except for patients aged and years. Overall casefatality rate for cerebral infarction, cerebral haemorrhage and subarachnoid haemorrhage were 10.7% (95% CI : ), 22.4% (95% CI : ) and 16.0% (95% CI : )

4 1062 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Table 2 Case fatality at 28 days by stroke sub-type, age group, and sex in Takashima, Shiga Prefecture, Cerebral Cerebral Subarachnoid infarction haemorrhage haemorrhage Total cases b Age Rate Rate Rate Rate group Patient Fatality (%) 95% CI a Patient Fatality (%) 95% CI Patient Fatality (%) 95% CI Patient Fatality (%) 95% CI Men Total Women Total Total Total a 95% confidence interval. b Four fatal cases of six unclassified cases have been included in Total cases.

5 STROKE INCIDENCE RATE IN TAKASHIMA, JAPAN 1063 respectively. In men case-fatality rates for cerebral infarction, cerebral haemorrhage and subarachnoid haemorrhage were 9.7% (95% CI : ), 29.8% (95% CI : ) and 30.8% (95% CI : ) respectively. In women the rates for cerebral infarction, cerebral haemorrhage and subarachnoid haemorrhage were 12.1% (95% CI : ), 15.7% (95% CI : ) and 27.3% (95% CI : ) respectively. The case-fatality rate combined men and women for subarachnoid haemorrhage is the highest among sub-types of stroke, and the rate in men was higher than in women. Discussion Takashima County in Shiga Prefecture is located at the geographical centre of Japan. Crude mortality rates from all causes of death and from cerebrovascular disease in Shiga Prefecture are and 97.5 per respectively, very close to the national figures of and 96.2 per respectively. 16 Based on these observations of geographical characteristics and population statistics, we may assume that the estimated incidence rate of stroke in our study represents the overall stroke incidence rate in Japan, and on this basis we are continuing the registration of stroke events in this prefecture to fulfil our prime aim of measuring secular trends. The necessity for comprehensiveness of the registration system for diseases like stroke is a significant problem in determining incidence in a particular area. A system to capture all patients in the study area, together with very accurate diagnosis, is required to ensure comprehensive registration. 17,18 Factors which reduce the comprehensiveness of registration are the non-registration of sudden death for which no confirmed diagnosis is available, non-registration of stroke patients admitted to hospitals outside the registration area, and nonregistration due to stroke patients being cared for at home and in nursing homes. We used comprehensive and multiple case-finding sources including identification through the emergency system which operates in Takashima County. Overall 91% of 123 patients admitted via the emergency system in Takashima County with stroke-like symptoms during a 2-year period were admitted to the two community hospitals inside Takashima County. Therefore it is considered that only a very small number of patients with acute stroke, who required emergency care, would have been missed, since in Japan stroke is seen as an emergency requiring immediate attention in hospital. To ensure that eligible patients hospitalized outside the county through other systems were not excluded, registration procedures were also conducted at the three high-level medical facilities outside the county. This allowed us to estimate that 96% of all admissions with stroke-like symptoms via the emergency system in the county were registered at the hospitals. The major limitation of our case-finding methods related to fatal and non-fatal strokes managed outside the three main medical facilities. However, because the practice in Takashima County is to admit almost all acute stroke cases to hospital, we estimate that there would be few such cases. Almost 100% of residents of Japan have health insurance offered by sources such as Ministry of Health and Welfare and there are many types of mutual benefit societies. 19 Health insurance is not expensive and policies cover all disease excluding injury by road traffic accident. Therefore, people with mild stroke who attended general physicians in the community are almost always referred to secondary or tertiary level hospitals for extensive investigations. This is the usual practice in Japan and so we believe that only a few cases with mild stroke were not registered in our system. As in most population-based studies of stroke in Japan, access to death certificates is limited, and we do not know how many fatal events occurred before emergency help was sought. The proportion of such cases is likely to be fewer than the 6% quoted in other populationbased studies where stroke is not viewed as a medical emergency. 20 The strength of this study is the comprehensiveness of diagnostic investigations, which has allowed almost complete categorization of stroke sub-types. This allows a comparison of stroke incidence rates with a similar study in Akita, Japan 21 (Table 3). Age-specific and age-adjusted incidence rates for the different sub-types of stroke and total stroke in Akita and Takashima are shown in Table 3. Rates for cerebral infarction were higher in Takashima than in Akita. In contrast, incidence of cerebral haemorrhage and subarachnoid haemorrhage in Takashima were lower than in Akita. Overall (all sub-types combined), the rates appear higher in Akita than in Takashima (177 per ). A comparison of rates with a study in Okinawa ( ) 8 suggested higher rates in Takashima (165 per population) than in Okinawa (137 per population). These patterns are in keeping with the geographical distribution of stroke in Japan where stroke mortality rates are higher in the northern regions of Japan than in the southern region. 22 We believe that stroke incidence rates in Takashima are close to the national average in Japan. In comparison with results from hospital-based studies in other Far East countries, stroke incidence rates in Takashima were similar to the rates in Taiwan. 23 We compared our data to those of MONICA Projects, although the case-finding methods are different in some respects, and our data showed that the stroke incidence rates for both of men and women in our study were similar to the middle rate group of MONICA Project participating countries. 24 In Takashima, incidence rates of cerebral infarction in men were more than twice as high as the rates in women. The differences were particularly great in the middle-aged group. These patterns of incidence of stroke in men and women have also been noted in a large stroke registry in Akita 21 and Okinawa. 8 In small community and cohort studies the differences in incidence rates for cerebral infarction between men and women were smaller Also a population-based study in Europe reported that incidence rate of cerebral infarction in men was higher than in women, but the differences were smaller than in Takashima. 28 In this study, we observed that overall 28-day case fatality (all stroke sub-types combined) was higher than that reported in seven area studies in Japan (Akita, Nagano, Shiga, Osaka, Ehime, Nagasaki and Okinawa). 8,21,29 Severity of stroke, type of stroke, the age of onset of stroke event and sex are important factors which influence early case fatality. 30 Incidence of haemorrhagic stroke which has a poor prognosis 26 was lower in Takashima than in Akita and Okinawa. Furthermore, average age of stroke onset in both men and women in Takashima was higher than in Okinawa (5.2 years for men and 3.6 years for

6 1064 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Table 3 Age-specific and age-adjusted annual stroke incidence (per population) of stroke sub-type for total men and women by age group, Takasima ( ) and Akita ( ) Cerebral Cerebral Subarachnoid infarction haemorrhage haemorrhage Total cases b Age group Rate 95% CI Rate 95% CI Rate 95% CI Rate 95% CI Takashima Total Age-adjusted rate a Akita Total Age-adjusted rate a a Adjusted to 1980 Japanese population aged years and over. b In Takashima six cases which could not been classified have been included in Total cases. women), 8 and the age-rank distribution of stroke patients in both men and women in Takashima had shifted to higher age groups than that in Akita. 21 Therefore, we conclude that case fatality in this study is high as a result of higher age at stroke onset. Our study showed that Takashima County has moderately high incidence rates and case fatality compared with other centres in Japan. It also confirms the higher incidence rate of cerebral infarction in men than in women. In order to decrease the incidence of stroke in Japan, greater efforts at primary prevention will be necessary. Ongoing monitoring of stroke in Takashima County is feasible and will in the future develop understanding of the reasons for the continuing decline in stroke mortality which has been noted in all regions of Japan. Acknowledgements This study was supported in part by grants (3A-1, 6A-5 and 7A-2) from The Research on Cardiovascular Diseases of Ministry of Health and Welfare. The authors acknowledge, in particular, Teruo Omae, MD, Kouji Isomura, MD and Osamu Iimura, MD for their support of this research. This study was carried out in cooperation with Department of Health Science, Shiga University of Medical Science, Takashima General Hospital, Makino Hospital, Otsu Red-Cross Hospital, Shiga Medical Center for Adult Diseases and Imazu Hospital. We wish to thank Minoru Ohkuma, MD, Yutaka Morita, MD, Atsushi Fujita, MD, Masanobu Yamakawa, PhD, Yasuhiro Yoshida, MD and Takashi Kadowaki, MD for their cooperation in this research. We also wish to thank Professor Robert Beaglehole, Professor Craig Anderson, and John Duncan of the Faculty of Medicine and Health Science, University of Auckland, for their advice. References 1 Omae T, Oita J, Ueda K. The Japanese experience in hemorrhegic stroke. J Hypertens 1994;12(Suppl.10):S19 S23. 2 Ueshima H. Changes in dietary habits, cardiovascular risk factors and mortality in Japan. Acta Cardiol 1990;45: Tanaka H, Tanaka Y, Hayashi M et al. Secular trends in mortality for cerebrovascular diseases in Japan, 1960 to Stroke 1982;13: Suzuki K, Kutsuzawa T, Takita K et al. Clinico-epidemiologic study of stroke in Akita, Japan. Stroke 1987;18: Shimamoto T, Komachi Y, Inada H et al. Trends for coronary heart disease and stroke and their risk factors in Japan. Circulation 1989; 79: Tanaka H, Ueda Y, Date C et al. Incidence of stroke in Shibata, Japan Stroke 1981;12: Ueda K, Omae T, Hirota Y et al. Decreasing trends in incidence and mortality from stroke in Hisayama residents, Japan. Stroke 1981;12: Kinjo K, Kimura Y, Shinzato Y et al. An epidemiological analysis of cardiovascular diseases in Okinawa, Japan. Hypertens Res 1992;15: Takeya Y, Popper JS, Shimizu Y, Kato H, Rhoads GG, Kagan A. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California: incidence of stroke in Japan and Hawaii. Stroke 1984;15: Study Project of Monitoring System for Cardiovascular Disease commissioned by the Ministry of Health and Welfare. Manual for the Registry and Follow-up of Stroke (in Japanese). Osaka, Japan: National Cardiovascular Center, World Health Organization MONICA Project. Event Registration Data Component, MONICA Manual Version 1.1. Document for meeting of MONICA Principal Investigators. 1986;S-4: Shiga Prefectural Imazu Public Health Center Annual Report of Imazu Public Health Center. Shiga Prefectural Imazu Public Health Center, Takashima, Shiga, Japan, 1991.

7 STROKE INCIDENCE RATE IN TAKASHIMA, JAPAN Statistics and Information Department, Minister s Secretariat, Ministry of Health and Welfare. About age-adjusted death rate. Jpn J Public Health 1991;38: Clayton D, Hills M. Statistical Models in Epidemiology. New York: Oxford University Press, SPSS for Windows Release 7.5 Standard Version. SPSS Inc., Health and Welfare Statistics Association. Health service in Japan (Kokumin eisei no doukou). Indices of Health and Welfare (Kousei-no shihyou). 1993;40: Sudlow CLM, Warlow CP. Comparing stroke incidence worldwide. What makes studies comparable? Stroke 1996;27: Bonita R, Broad JB, Anderson NE, Beaglehole R. Approaches to the problems of measuring the incidence of stroke: The Auckland Stroke Study, Int J Epidemiol 1995;24: Health and Welfare Statistics Association. Health and Welfare Statistics in Japan 1992, pp.147, 1992, Tokyo. 20 Anderson NE, Bonita R, Broad JB. Early management and outcome of acute stroke in Auckland. Aust NZ J Med 1997;27: Suzuki K, Sarti C, Tuomilehto J et al. Stroke incidence and case fatality in Finland and in Akita, Japan: a comparative study. Neuroepidemiology 1994;13: Kagami M. Regional variance of cerebrovascular mortality in Japan. Ecology of Disease 1983;2: Chang SF, Su CL, Chen ZY, Hung T. Stroke incidence in Iian, Taiwan. J Formos Med Assoc 1995;94: Thorvaldsen P, Asplund K, Kuulasmaa K, Rajakangas AM, Schroll M. Stroke incidence, case fatality, and mortality in the WHO MONICA Project. Stroke 1995;26: Ueda K, Hasuo Y, Kiyohara Y et al. Intracerebral hemorrhage in Japanese community, Hisayama: incidence, changing pattern during long-term follow-up, and related factors. Stroke 1988;19: Nakayama T, Date C, Yokoyama T, Yoshiike T, Yamaguchi M, Tanaka H. A year follow-up study of stroke in a Japanese provincial city The Shibata Study. Stroke 1997;28: Kodama K. Stroke trends in Japan. Ann Epidemiol 1993;3: D Alssandro G, Giovanni DM, Roveyaz L et al. Incidence and prognosis of stroke in the Valle d Aosta, Italy: first-year results of a communitybased study. Stroke 1992;23: Isomura K. 3A-1 Study on the development of the community based long-term follow-up system for cardio-cerebrovascular diseases Annual Report of the Research on Cardiovascular Diseases (in Japanese with English abstract). Osaka, Japan: National Cardiovascular Center, Kotila M. Declining incidence and mortality of stroke? Stroke 1984;15: